List of possible variables for estimation of probability.
\r\n\t
",isbn:"978-1-83968-388-6",printIsbn:"978-1-83968-387-9",pdfIsbn:"978-1-83968-389-3",doi:null,price:0,priceEur:0,priceUsd:0,slug:null,numberOfPages:0,isOpenForSubmission:!1,hash:"61ec2bad4fc3f7060fd64b91fa12e82c",bookSignature:"Ph.D. Vicente Vanaclocha",publishedDate:null,coverURL:"https://cdn.intechopen.com/books/images_new/10574.jpg",keywords:"Prevalence, Incidence, Worldwide Differences, Red Flags, Moyamoya and School Performance, Medical Treatment, Surgical Treatment, Genetic Markers, Immunologic Factors, Recommended Anesthetic Agents, Source of Intraoperative Complications, Post-Operative ICU Management",numberOfDownloads:null,numberOfWosCitations:0,numberOfCrossrefCitations:null,numberOfDimensionsCitations:null,numberOfTotalCitations:null,isAvailableForWebshopOrdering:!0,dateEndFirstStepPublish:"September 23rd 2020",dateEndSecondStepPublish:"October 21st 2020",dateEndThirdStepPublish:"December 20th 2020",dateEndFourthStepPublish:"March 10th 2021",dateEndFifthStepPublish:"May 9th 2021",remainingDaysToSecondStep:"3 months",secondStepPassed:!0,currentStepOfPublishingProcess:4,editedByType:null,kuFlag:!1,biosketch:"Dr. Vicente Vanaclocha is a Chief of Neurosurgery at the University Hospital of Navarra and head of Neurosurgery Service of San Jaime Hospital in Torrevieja. He has over 25 years of experience in neuro-oncology and minimally invasive surgery techniques. He is a pioneer in many areas in neurosurgery (treatment of brain tumors, Chiari Malformation, and sacroiliac joint disorders).",coeditorOneBiosketch:null,coeditorTwoBiosketch:null,coeditorThreeBiosketch:null,coeditorFourBiosketch:null,coeditorFiveBiosketch:null,editors:[{id:"199099",title:"Ph.D.",name:"Vicente",middleName:null,surname:"Vanaclocha",slug:"vicente-vanaclocha",fullName:"Vicente Vanaclocha",profilePictureURL:"https://mts.intechopen.com/storage/users/199099/images/system/199099.jpeg",biography:"Vicente Vanaclocha is Chief of Neurosurgery. Doctor of Medicine from the University of Valencia, he has over 25 years experience in neuro-oncology, minimally invasive and minimally invasive surgery techniques. Specialist in neurosurgery both nationally and internationally (including the General Medical Register of England and stay at the Groote Schuur Hospital in Cape Town, South Africa) has been Chief of Neurosurgery at the University Hospital of Navarra and head of Neurosurgery Service of San Jaime Hospital in Torrevieja. He was also associate professor of neurosurgery at the Faculty of Medicine of the University of Navarra and is a professor of neuroanatomy at the Catholic University of Valencia also serving as an editorial board member of repute.\nCurrently he is Associate Professor at the University of Valencia.",institutionString:"University of Valencia",position:null,outsideEditionCount:0,totalCites:0,totalAuthoredChapters:"7",totalChapterViews:"0",totalEditedBooks:"1",institution:{name:"University of Valencia",institutionURL:null,country:{name:"Spain"}}}],coeditorOne:null,coeditorTwo:null,coeditorThree:null,coeditorFour:null,coeditorFive:null,topics:[{id:"16",title:"Medicine",slug:"medicine"}],chapters:null,productType:{id:"1",title:"Edited Volume",chapterContentType:"chapter",authoredCaption:"Edited by"},personalPublishingAssistant:{id:"297737",firstName:"Mateo",lastName:"Pulko",middleName:null,title:"Mr.",imageUrl:"https://mts.intechopen.com/storage/users/297737/images/8492_n.png",email:"mateo.p@intechopen.com",biography:"As an Author Service Manager my responsibilities include monitoring and facilitating all publishing activities for authors and editors. From chapter submission and review, to approval and revision, copyediting and design, until final publication, I work closely with authors and editors to ensure a simple and easy publishing process. I maintain constant and effective communication with authors, editors and reviewers, which allows for a level of personal support that enables contributors to fully commit and concentrate on the chapters they are writing, editing, or reviewing. I assist authors in the preparation of their full chapter submissions and track important deadlines and ensure they are met. I help to coordinate internal processes such as linguistic review, and monitor the technical aspects of the process. As an ASM I am also involved in the acquisition of editors. Whether that be identifying an exceptional author and proposing an editorship collaboration, or contacting researchers who would like the opportunity to work with IntechOpen, I establish and help manage author and editor acquisition and contact."}},relatedBooks:[{type:"book",id:"6546",title:"Treatment of Brachial Plexus Injuries",subtitle:null,isOpenForSubmission:!1,hash:"24a8e7c7430e86f76fb29df39582855a",slug:"treatment-of-brachial-plexus-injuries",bookSignature:"Vicente Vanaclocha and Nieves Sáiz-Sapena",coverURL:"https://cdn.intechopen.com/books/images_new/6546.jpg",editedByType:"Edited by",editors:[{id:"199099",title:"Ph.D.",name:"Vicente",surname:"Vanaclocha",slug:"vicente-vanaclocha",fullName:"Vicente Vanaclocha"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"6550",title:"Cohort Studies in Health Sciences",subtitle:null,isOpenForSubmission:!1,hash:"01df5aba4fff1a84b37a2fdafa809660",slug:"cohort-studies-in-health-sciences",bookSignature:"R. Mauricio Barría",coverURL:"https://cdn.intechopen.com/books/images_new/6550.jpg",editedByType:"Edited by",editors:[{id:"88861",title:"Dr.",name:"René Mauricio",surname:"Barría",slug:"rene-mauricio-barria",fullName:"René Mauricio Barría"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"1591",title:"Infrared Spectroscopy",subtitle:"Materials Science, Engineering and Technology",isOpenForSubmission:!1,hash:"99b4b7b71a8caeb693ed762b40b017f4",slug:"infrared-spectroscopy-materials-science-engineering-and-technology",bookSignature:"Theophile Theophanides",coverURL:"https://cdn.intechopen.com/books/images_new/1591.jpg",editedByType:"Edited by",editors:[{id:"37194",title:"Dr.",name:"Theophanides",surname:"Theophile",slug:"theophanides-theophile",fullName:"Theophanides Theophile"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"3092",title:"Anopheles mosquitoes",subtitle:"New insights into malaria vectors",isOpenForSubmission:!1,hash:"c9e622485316d5e296288bf24d2b0d64",slug:"anopheles-mosquitoes-new-insights-into-malaria-vectors",bookSignature:"Sylvie Manguin",coverURL:"https://cdn.intechopen.com/books/images_new/3092.jpg",editedByType:"Edited by",editors:[{id:"50017",title:"Prof.",name:"Sylvie",surname:"Manguin",slug:"sylvie-manguin",fullName:"Sylvie Manguin"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"3161",title:"Frontiers in Guided Wave Optics and Optoelectronics",subtitle:null,isOpenForSubmission:!1,hash:"deb44e9c99f82bbce1083abea743146c",slug:"frontiers-in-guided-wave-optics-and-optoelectronics",bookSignature:"Bishnu Pal",coverURL:"https://cdn.intechopen.com/books/images_new/3161.jpg",editedByType:"Edited by",editors:[{id:"4782",title:"Prof.",name:"Bishnu",surname:"Pal",slug:"bishnu-pal",fullName:"Bishnu Pal"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"72",title:"Ionic Liquids",subtitle:"Theory, Properties, New Approaches",isOpenForSubmission:!1,hash:"d94ffa3cfa10505e3b1d676d46fcd3f5",slug:"ionic-liquids-theory-properties-new-approaches",bookSignature:"Alexander Kokorin",coverURL:"https://cdn.intechopen.com/books/images_new/72.jpg",editedByType:"Edited by",editors:[{id:"19816",title:"Prof.",name:"Alexander",surname:"Kokorin",slug:"alexander-kokorin",fullName:"Alexander Kokorin"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"1373",title:"Ionic Liquids",subtitle:"Applications and Perspectives",isOpenForSubmission:!1,hash:"5e9ae5ae9167cde4b344e499a792c41c",slug:"ionic-liquids-applications-and-perspectives",bookSignature:"Alexander Kokorin",coverURL:"https://cdn.intechopen.com/books/images_new/1373.jpg",editedByType:"Edited by",editors:[{id:"19816",title:"Prof.",name:"Alexander",surname:"Kokorin",slug:"alexander-kokorin",fullName:"Alexander Kokorin"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"57",title:"Physics and Applications of Graphene",subtitle:"Experiments",isOpenForSubmission:!1,hash:"0e6622a71cf4f02f45bfdd5691e1189a",slug:"physics-and-applications-of-graphene-experiments",bookSignature:"Sergey Mikhailov",coverURL:"https://cdn.intechopen.com/books/images_new/57.jpg",editedByType:"Edited by",editors:[{id:"16042",title:"Dr.",name:"Sergey",surname:"Mikhailov",slug:"sergey-mikhailov",fullName:"Sergey Mikhailov"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"371",title:"Abiotic Stress in Plants",subtitle:"Mechanisms and Adaptations",isOpenForSubmission:!1,hash:"588466f487e307619849d72389178a74",slug:"abiotic-stress-in-plants-mechanisms-and-adaptations",bookSignature:"Arun Shanker and B. Venkateswarlu",coverURL:"https://cdn.intechopen.com/books/images_new/371.jpg",editedByType:"Edited by",editors:[{id:"58592",title:"Dr.",name:"Arun",surname:"Shanker",slug:"arun-shanker",fullName:"Arun Shanker"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"878",title:"Phytochemicals",subtitle:"A Global Perspective of Their Role in Nutrition and Health",isOpenForSubmission:!1,hash:"ec77671f63975ef2d16192897deb6835",slug:"phytochemicals-a-global-perspective-of-their-role-in-nutrition-and-health",bookSignature:"Venketeshwer Rao",coverURL:"https://cdn.intechopen.com/books/images_new/878.jpg",editedByType:"Edited by",editors:[{id:"82663",title:"Dr.",name:"Venketeshwer",surname:"Rao",slug:"venketeshwer-rao",fullName:"Venketeshwer Rao"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}}]},chapter:{item:{type:"chapter",id:"18021",title:"Clinical Engineering",doi:"10.5772/19763",slug:"clinical-engineering",body:'\n\t\t\n\t\t\t\tClinical Engineering (CE) represents the part of Biomedical Engineering focused on the applications of theories and methodologies of the broad biomedical engineering field to improve the quality of health services. Its activities especially concern the appropriate management of biomedical technologies (from purchasing to risk controlling) and the development and the adjustment of hospital informative systems and telemedicine networks. CE combines with the medicine knowledge for conducing of healthcare activities by providing expertise in a wide spectrum of topics, from human physiology and biomechanics to electronics and computer science.
\n\t\t\tAs biomedical technology developed towards ever more complex systems and spread in every clinical practice, so the field of CE grew. Such growth has been accompanied by an analogous expansion of biomedical and clinical engineering studies at the University and development of skills and tasks of CE professionals.
\n\t\t\tThe main aim of CE is to support the use of biomedical technology by health professionals and hospital organizations with appropriate skills in order to reach the best compromise between clinical efficacy/efficiency, patient and operators safety, care quality and innovation, and management and equipment costs.
\n\t\t\t\n\t\t\t\tCE techniques and methodologies are mainly focused on safe, appropriate and economical management of technologies, as well as on governance and management (limited to specific responsibilities) of healthcare facility. Thus, CE covers all those knowledge and methods applied to the management of biomedical technologies, ranging from their early evaluation and assessment, to their technical conduct, to their dismissing. Thus the chapter will highlight different aspects of technology management by exploring technical and/or clinical, and/or economic issues related to the individuation and acquisition of appropriate equipment (i.e., Health Technology Assessment), acceptance testing, management of preventive and corrective maintenance, risk management, planning of quality testing, ICT management, management of maintenance contracts, equipments replacement planning, and so on.
\n\t\tBecause of the strong pressure on the health structures to optimize the services provided while lowering the associated costs and reducing the likelihood of adverse events, an organizational approach, in which a Healthcare Risk Management program plays a central role, becomes important.
\n\t\t\tMistakes can be minimized, in fact, by creating organizational systems and using technologies to make it easier to do the right thing. It is clear that patient safety can be increased by means of appropriate procedures aimed at avoiding possible mistakes or correcting those that do happen.
\n\t\t\tIn particular, the potential for biomedical equipment related adverse events needs to be analyzed in order to prevent their occurrence: healthcare structures have to use systematic analytical methods and instruments to manage technological risks to both patients and operators.
\n\t\t\tThe aim of the health organizations is to take care of patients, by providing effective, appropriate and, in particular, safe treatments. The healthcare institutions (such as the clinicians themselves) have to ensure the care, as adequate as possible, of patients, avoiding or at least containing damage caused by human and system errors. Healthcare service activities connote, in fact, with the presence of several hazards that have the potential to harm patients and health operators.
\n\t\t\tCurrently the best known approach is the Healthcare Risk Management program, with which it is possible to identify, assess, mitigate and control healthcare facilities risks, and thus realize the concept of “systemic safety”.
\n\t\t\tOriginally such approaches focused mainly, if not entirely, on the problem of reducing the “Clinical Risk” (Clinical Risk Management, CRM) with the aim of limiting enterprise liability-costs. In fact, over the course of the last several years healthcare institutions and practitioners have experienced a "malpractice crisis" that has led to the increase in jury verdicts, settlement amounts and insurance premiums, as well as dwindling insurance availability due to carrier withdrawals from the medical malpractice market (McCaffrey & Hagg-Rickert, 2010), and consequent increase of risk retention cost.
\n\t\t\tGradually, the focus shifted to clinical problems and thus the term CRM now encompasses strategies to reduce the incidence and magnitude of harm and improve the quality of care (Taylor-Adams, et al., 1999) by focusing on patient safety and patient care related issues, including information gathering systems, loss control efforts, professional liability, risk financing and claims management activities.
\n\t\t\tDealing with clinical risk and patient safety means also dealing with biomedical technologies. In fact, as medical treatments have greatly progressed along with the analogous technological advances in medical equipment (ME), all medical procedures depend, to some extent, on technology to achieve their goals. Despite the (presupposed) inherent safety of MEs (also guaranteed by a plethora of laws and technical standards), device-related adverse events occur every day in hospitals around the world. Some can be very dangerous and occasionally even deadly.
\n\t\t\t\tAn adverse event is (as defined by Medicines and Healthcare Products Regulatory Agency, MHRA) “an event that causes, or has the potential to cause, unexpected or unwanted effects involving the safety of device users (including patients) or other persons”. ME related adverse events can occur for several reasons, ranging from incorrect choice and acquisition of the device, wrong installation, and poor maintenance, to use error and device obsolescence.
\n\t\t\t\tAs stated before, a systemic approach is needed. Such an approach, identified as Medical Equipments Risk Management (MERM), is part of the global Technology Management (Wilkins & Holley, 1998) as practised by the Clinical Engineering Department (CED) within the hospital. The specific activities of the MERM process are, as coded by several international standards (AS/NZS 4360:2004\n\t\t\t\t\t; ISO 14971:2007; ISO 31000:2009) regarding risk management applied to general production processes and specifically to the design and production of medical devices (but addressed to manufacturers, not the users, of medical equipment) as follows:
\n\t\t\t\trisk identification
risk analysis and assessment (including risk prioritization);
planning of actions to mitigate the risk;
tracking of information about the implemented actions;
control and follow up.
All the standards stress that the task of risk assessment, along with risk identification,. is the most important element. This is mainly because all the measures the CED (as well as the healthcare organization as a whole) will take to reduce the level of risk will depend on the results of these two phases: an error in assessment would probably lead to several mistakes (and therefore waste of economical and human resources) in the subsequent phases.
\n\t\t\t\tGiven below is a brief description of the methods available for addressing risk analysis and assessment. However, a thorough analysis of the remaining phases is left to the reader, since they require the active involvement of several lines of professionals, and thus are strongly dependent on the organizational and operational arrangements of the specific healthcare facility.
\n\t\t\t\tRisk identification and risk analysis are processes aimed at identifying the type of hazard and determining the potential severity associated with an identified risk and the probability that a harmful event will occur. Together, these factors establish the “seriousness of a risk” and guide the clinical engineer’s choice of an appropriate “risk treatment” strategy (including preventive maintenance, user training, definition of a renewal plan, etc.).
\n\t\t\t\t\tTechniques for risk identification and assessment are various and dependent on the specific kind of hazard under assessment. In the healthcare sector, two techniques are widely and commonly used: Failure Mode and Effects Analysis and Root Cause Analysis.
\n\t\t\t\t\tFailure Mode and Effect Analysis (FMEA) is a systematic process for identifying potential process and technical failures, with the intent to eliminate them or minimize their likelihood, before they occur, that is in advance of the occurrence of the adverse event related to the analyzed risk (American Society for Healthcare Risk Management [ASHRM], 2002). Initiated in the 1940s by the U.S. Defense Department, FMEA was further developed by the aerospace and automobile industries, but it was only in the late 1960s that it was first applied to healthcare processes. Since then, in the healthcare sector, Failure Mode and Effects Analysis has been developed as a systematic, proactive method for evaluating clinical processes to identify where and how they might fail, and to assess the relative impact (in terms of damage to patients, workers and facilities) of different failures in order to identify the parts of the process that are most in need of change.
\n\t\t\t\t\tThe rationale of FMEA is the acknowledgement that errors are inevitable and predictable, and thus can be anticipated and/or minimized by design.
\n\t\t\t\t\tAs suggested by the name, the focus is on the Failure Mode (defined as the incorrect behavior of a subsystem or component due to a physical or human reason), on the Effect (defined as the consequences of a failure on operation, function or functionality, or status of some item) and, potentially (in which case the acronym becomes FMECA) on Criticality (defined as the combination of the probability that a failure will occur and the severity of its effect on the system or subsystem). In other words FMEA (or FMECA) analysis aims to identify and analyze
\n\t\t\t\t\tAll potential failure modes of a system and components of the system;
The effects these failures may have on the system and parts of the system;
How to avoid or reduce the probability of the failures, or mitigate the effects of the failures on the system.
Depicted below is a schematic, step-by-step description of how to conduct the FMEA process:
\n\t\t\t\t\tDefine the FMEA topic.
Write a clear definition of the process to be studied.
Narrow the scope of the review so that it is manageable, and the actions are practical and able to be implemented.
Assemble the Team.
Guarantee the multidisciplinarity of the team by including expert representatives of all affected areas.
Identify the team leader/coordinator.
Prepare a graphic description of the process
Create and verify the flow chart.
Number each process step.
For complex processes, specify the area to focus on.
Identify and create a flow chart of the subprocesses.
Conduct a Hazard Analysis
List all possible/potential failure modes for each process/subprocess.
List all the possible causes of the failure mode (each failure mode may have multiple failure mode causes).
Determine the “severity (S)”, “probability (P)” and “detectability (D).”
Determine the Risk Priority Number (RPN = S x P x D).
Determine if the failure mode warrants further action (e.g. RPN > 32).
Actions and Outcome Measures
identify actions or strategies to reduce the Risk Priority Number for each failure mode
The other widely adopted methodology is Root Cause Analysis (RCA) that aims to assess risks affecting healthcare activities by investigating the adverse events which have occurred. RCA is an analytic tool for performing a comprehensive, system-based review of critical incidents. It includes the identification of the root cause and contributory factors, determination of risk reduction strategies, and development of action plans along with measurement strategies to evaluate the effectiveness of the plans (Canadian Patient Safety Institute [CPSI], 2006). Unlike FMEA, which is a proactive and preventive process, RCA is carried out retrospectively in response to a specific, harmful event.
\n\t\t\t\t\tThe main purpose of the RCA is to uncover the factor(s) that led to and caused the serious preventable adverse event. The preventable adverse event is very often the tip of the iceberg. Conducting and writing an RCA is an opportunity to examine how the systems for providing care function. The more areas investigated, the greater the possibility the system(s) will become better functioning and prevent the next event from occurring.
\n\t\t\t\t\tRCA focuses on the “how” and the “why”, not on the “who”. The goals of a root cause analysis are to determine:
\n\t\t\t\t\twhat happened;
why it happened;
what can be done to reduce the likelihood of recurrence.
A step by step description of the RCA may be depicted as follows:
\n\t\t\t\t\tPlan of action
strategies the organization intends to implement in order to reduce the risk of similar events occurring in the future.
responsibility for implementation, supervision, pilot testing as appropriate, time lines, and strategies for measuring the effectiveness of the actions.
What happened / Facts of the event
Information about the patient
Details of the event
Use of interviews, brain storming, or written description, etc.
Why it happened
Individuate the contributory factors
Identify root causes
Identification of the “Root Causes”
Minimize recurrence/monitoring
Implementation of each specific action that will be measured and communicated
The final goal of both methodologies is to address the commitment of healthcare organizations to reduce the likelihood or severity of adverse events. However, besides their technical, practical and philosophical differences, both present a major fault/drawback when applied to the specific case of medical equipment risk assessment. In fact, the methods themselves require some form of subjective assessment, mainly due to lack of quantitative data on which the assessment could be based. Moreover, to assess the risk related to the entire biomedical technological assets of healthcare facilities would certainly require a more systematic and structured method for collecting and processing data.
\n\t\t\t\t\tA possible solution to this problem could be an adapted implementation of the Risk Map or Risk Matrix (Ruge, 2004; Cox, 2008). A risk matrix (risk map) is a table (Cartesian diagram) that presents on its rows (y-axis), the category of probability (or likelihood or frequency) and on its columns (x-axis), the category of severity (or impact or consequences). Each cell of the table (or point in the Cartesian plane), which mathematically represents the product of the probability and severity values, is associated to a level of risk that eventually identifies the urgency or priority of the required mitigation actions.
\n\t\t\t\t\tThe figure 1 shows an example of risk matrix, where probability and severity have been split into a range of five values, whereas risk level is categorized into three classes.
\n\t\t\t\t\tThus, the risk assessment problem can be reduced to the estimate of probability and severity values. The estimate of severity does not present any particular concerns: by analyzing equipment design and features (such as, also, the FDA or CE risk classification), device user manual, clinical procedure and medical room in which the ME is used, it should be easy to determine the maximum possible damage the ME could do to the patient (or even to the operator). Moreover, such elements can be easily described by specifically defined numeric variables (for instance, all the considered aspects can be assigned values ranging from 1 to 5, in analogy with the main Risk Matrix axe values) and recorded in the equipment management system used by the CED. Lastly, by defining a computation method (whose complexity can vary from a very simple linear sum up to more complex fuzzy or neural network systems) the severity value can be associated to each ME owned by the healthcare facility.
\n\t\t\t\t\tHowever, the achievement of a robust, objective estimate of probability definitely presents more difficulties. In particular, it would be preferable to take into account only measurable characteristics, thus using easily quantifiable numeric variables.
\n\t\t\t\t\tExample of Risk Matrix
The complexity of estimating probability stems from the fact that probability is dependent on three main different but inter-influenced issues: human factor, medical device functional reliability, medical device design and environmental characteristics (Brueley, 1989; Anderson, 1990; Dillon, 2000; FDA, 1997; FDA, 2000; Samore, et al., 2004). So, estimating the probability value must take into account the evaluation of these three elements. In estimating the human factor element, one must take into account not only those characteristics of the ME, of the process and/or of the environment that may facilitate a human error leading to an adverse event, but also the factors that may make the operator take corrective action for a ME or system failure. The ME functional reliability refers to the potential for device (material and/or functional) failure, potentially leading to an adverse event. Aspects to be considered are those related to the device reliability assessment such as the execution of safety checks, assessment of device obsolescence, and respect of a preventive maintenance plan. Medical device design and environmental characteristics are those related respectively to the possibility of the ME having specific features that could lead to an adverse event without the occurrence of material or functional failure or human error, and to the presence of environmental factors that could cause the ME to fail.
\n\t\t\t\t\tWhen defining the elements to be analyzed on each ME owned by the hospital, two considerations apply:
\n\t\t\t\t\tDefine measurable variables more quantitatively.
Prefer elements (variables) already monitored by the organization and recorded in an information system (such as the ME management system used by the CED)
\n\t\t\t\t\t\tTable 1 shows an example of variables for probability estimation.
\n\t\t\t\t\t\n\t\t\t\t\t\t\t\t\tHuman factor\n\t\t\t\t\t\t\t\t | \n\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t\tMedical device functional reliability\n\t\t\t\t\t\t\t\t | \n\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t\tMedical device design and environmental characteristics \n\t\t\t\t\t\t\t\t | \n\t\t\t\t\t\t\t
Availability (at point of use) of complete written instructions (e.g., user manual) from the manufacturer | \n\t\t\t\t\t\t\t\tDevice obsolescence | \n\t\t\t\t\t\t\t\tAppropriateness of wiring according to clinical activities and devices | \n\t\t\t\t\t\t\t
Device ergonomics | \n\t\t\t\t\t\t\t\tExistence and respect of a preventive maintenance plan | \n\t\t\t\t\t\t\t\tEnvironmental conditions (noise, temperature, vibrations, electromagnetic interference, etc.) | \n\t\t\t\t\t\t\t
Difficult working conditions (staff shortage, staff shifts, etc.) | \n\t\t\t\t\t\t\t\tResults of safety checks (cfr. IEC or ISO or EN safety standards) | \n\t\t\t\t\t\t\t\tThe device is appropriate for the clinical needs for which it is intended | \n\t\t\t\t\t\t\t
Environmental conditions (noise, temperature, lighting, space, etc.) | \n\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t | \n\t\t\t\t\t\t\t |
Schedule and records of a training and education program on the use of specific ME and its related risks | \n\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t | \n\t\t\t\t\t\t\t |
List of possible variables for estimation of probability.
As is done for estimating severity, the last step consists of defining a computation method to elaborate the identified variables. Also, in this case the complexity of the method may vary from a very simple linear sum up to more complex fuzzy or neural network systems.
\n\t\t\t\tNowadays many factors, ranging from the aging of population to the continuous fast-paced technology innovation, as well as the even more critical scarcity of economic resources, emphasize the importance of correct resource allocation at every level of a national health care system. This background adds to the criticality and complexity of decision-making, rendering essential a thorough evaluation which takes into consideration all the areas (health benefits, risks, costs, etc.) where health technology may have an impact..
\n\t\t\tA variety of specific methods and tools are available to support health care and medical decision making, for example Health Technology Assessment (HTA), a standardized methodology that can help decision makers select the most appropriate choice for their specific context.
\n\t\t\tHTA is a multidisciplinary process that systematically examines the technical performance, safety, clinical efficacy, effectiveness, cost, cost-effectiveness ratio, organizational implications, social consequences and legal and ethical considerations of the application of a health technology (EUNEHTA).
\n\t\t\tAdvances in science and engineering | \n\t\t\t\t\t
Intellectual property, especially patent protection | \n\t\t\t\t\t
Aging population | \n\t\t\t\t\t
“Cascade” effects of unnecessary tests, unexpected results, patient or physician anxiety | \n\t\t\t\t\t
Emerging pathogens and other disease threats | \n\t\t\t\t\t
Third-party payment | \n\t\t\t\t\t
Inability of third-party payers to limit coverage | \n\t\t\t\t\t
Financial incentives of technology companies, clinicians, and others | \n\t\t\t\t\t
Clinician specialty training at academic medical centers | \n\t\t\t\t\t
Malpractice avoidance | \n\t\t\t\t\t
Provider competition to offer state-of-the-art technology | \n\t\t\t\t\t
Public demand driven by consumer awareness, direct-to-consumer advertising, and mass media reports | \n\t\t\t\t\t
Strong economies, high employment | \n\t\t\t\t\t
Factors that reinforce the market for health technology (Goodman 2004)
The term “health technology” is quite broad and includes the following categories: drugs, biologics, medical devices, equipment and supplies, medical and surgical procedures, support systems, organizational and managerial systems.
\n\t\t\tHTA may address the direct, intended consequences of technologies as well as their indirect, unintended consequences; its main purpose is to inform technology-related policy-making in health care.
\n\t\t\tHTA is increasingly used in American and European countries to inform decision- and policy-making in the health care sector and several countries have integrated HTA into policy, governance, reimbursement or regulatory processes.
\n\t\t\tAn HTA process is conducted by interdisciplinary groups using explicit analytical frameworks drawing from a variety of methods: given the variety of impacts addressed and the range of methods that may be used in an assessment, several types of experts are needed in HTA.
\n\t\t\t\tDepending upon the topic and scope of assessment, these may include a selection of the following (Goodman, 2004):
\n\t\t\t\tPhysicians, nurses, dentists, and other clinicians
Patients or patient representatives
Epidemiologists
Managers of hospitals, clinics, nursing homes, and other health care institutions
Biostatisticians
Economists
Lawyers
Radiology technicians, laboratory technicians and other health professionals
Social scientists
Ethicists
Decision scientists
Clinical and biomedical engineers
Computer scientists/programmers
Pharmacologists
Librarians/information specialists
According to a recent study there are also significant differences in the practical application of HTA. Whereas in some countries HTA merely studies the clinical effectiveness and perhaps safety and cost-effectiveness of technologies, agencies in other countries apply a broader perspective and also consider other issues, such as ethics, and organizational, social or legal aspects of technology.
\n\t\t\t\tIt is also known that the HTA activities can be carried out at different levels of health-care systems:
\n\t\t\t\tmacro level (international and national - i.e. decision-making within central government institutions)
meso level (administrative level - i.e. regional or provincial health authorities, agencies, primary health-care units or hospitals);
micro level (clinical practice)
At each of these levels, however, these activities should be carried out by a multidisciplinary staff, involving clinicians, clinical engineers, economists, epidemiologists, etc.) and, depending on the object of evaluation, also by specifically qualified professionals from the hospital departments.
\n\t\t\t\t\n\t\t\t\t\t\t\t\tAssessment reason\n\t\t\t\t\t\t\t | \n\t\t\t\t\t\t|
New technology | \n\t\t\t\t\t\t\tSafety concerns | \n\t\t\t\t\t\t
Changes in old technology | \n\t\t\t\t\t\t\tEthical concerns | \n\t\t\t\t\t\t
New indications for old technology | \n\t\t\t\t\t\t\tEconomic concerns | \n\t\t\t\t\t\t
New findings | \n\t\t\t\t\t\t\tInvestment decisions | \n\t\t\t\t\t\t
Structural/organizational changes | \n\t\t\t\t\t\t\t\n\t\t\t\t\t\t |
Reasons for performing an assessment (Velasco, et al., 2002)
As discussed in the previous paragraph, HTA now represents a multidimensional field of inquiry that increasingly responds to broad social forces such as citizen participation, accelerated technological innovation, and the allocation of scarce resources among competing priorities (Battista, 2006).
\n\t\t\t\tHowever,, this methodology was initially focused and applied on a small scale, concerning (clinical) engineering questions pertaining to a technology’s safety and technical performances, and involving the investigation of one or more properties, impacts, or other features of health technologies or applications.
\n\t\t\t\tThe technical evaluation represents, in fact, the core object of Clinical Engineering (CE) activity in HTA and is often conducted at a meso level. Many hospitals are increasingly developing HTA processes by means of HTA Commissions or structured HTA Unit, that include the CED.
\n\t\t\t\tIn the Health Technology, CED are typically involved in the technical evaluation of the medical electrical equipment (as defined by the IEC 60601-1-1 normative) and sometimes of medical devices.
\n\t\t\t\tThe main features characterizing these kinds of technologies can be summarized as follows:
\n\t\t\t\tfast-changing technologies: their development is characterized by a constant flow of incremental product improvements;
device impact on clinical and safety outcome depends on user training and experience that can vary and are hard to evaluate;
the life cycle of a device is often as short as 18–24 months, which is considerably less than, for example, pharmaceuticals;
the clinical application of the technology and potential utility for patients (accuracy or effectiveness) in comparison with the reference standard;
improvement in the operating principle;
state of development of technology (emerging, new, established);
impact on organization (implementation phase, change in the treatment, users’ qualification, IT requirements, etc.);
impact on patient and user safety;
economic aspects (acquisition, maintenance, spare parts, training, etc.);
devices cannot be evaluated by RCTs – hard to blind and randomize. Early evaluation not possible
Representation of Health Technology, Medical Device and Medical Electrical Equipment sets
A further classification of medical electrical equipment can be made according to their main characteristics or function. For instance, as can be found in the Italian CIVAB classification, medical electrical equipment can be grouped in three technological compartments:
\n\t\t\t\tFunctional explorations and therapeutic equipment;
Medical laboratory or clinical chemistry equipment;
Bio-imaging equipment.
The HTA process, while maintaining a uniform and systematic approach, may have to primarily focus on different characteristics because of the different weighting or different evaluation methodologies for the following aspects:
\n\t\t\t\tInnovation
Safety
Technology management
Efficacy
Investment (big ticket technology; high volume purchase; service)
Organization.
\n\t\t\t\t\tFunctional explorations and therapeutic equipment often undergo relevant innovation, such as that involving the change of the physical or biological operating principle, which is difficult to evaluate empirically (“impossibility” of randomized controlled trial (RCT), short Time To Market vs short mean life). As regards safety, electromedical equipment are regulated by directives and technical norms that constitute not only a fair guarantee of their safety but also a valid guide to evaluate it for the specific context of its intended utilization. Moreover, patient safety strongly depends on user education and training in equipment use. Equipment’s efficacy is often evaluated only by design data or in vitro or animal model tests. As such devices represent the greater part of an institute’s biomedical equipment assets, organizational, economical and management issues become fairly important: uniformity of equipment can facilitate technological management (including risk issues), rationalize maintenance, take advantage of scale factors (equipment acquisition and renewal, consumables/spare parts).
\n\t\t\t\tAssessment of innovation for Medical Laboratory equipment has to accommodate the continuous introduction of new reagents and controls as well as the presence of homebrew technology, particularly in the most advanced fields such as Proteomics and Metabolomics. As concerns the management of these technologies, uniformity of equipment is also important for better and easier use by the operators, and ensures the availability of backup equipment. The most common mode of acquisition is by rental or service, where the cost of the equipment is included in the cost of the reagents.
\n\t\t\t\t\n\t\t\t\t\tBio-Imaging\n\t\t\t\t\tequipment have been subject to innovations in virtually all aspects of their functioning, e.g. improvement of technical performances (e.g. spatial resolution), change in physical or biological operating principle (e.g. fMRI), safety for operators and patients (e.g. X-ray dose reduction). Their empirical evaluation is usually more practicable than for other kinds of equipment, particularly when testing no side-effects of technologies. Patient and operator safety relies on operational, technical and organizational issues (e.g. use of minimum dose setting for x-ray exams, implementing X-ray or magnetic shielding walls and ceilings, limiting access to exam room). As their complexity increases, so does the importance of user education and training to ensure a safe use of all the technological facilities. These kinds of technologies may have a very high cost both for their acquisition and for the necessary structural changes.
\n\t\t\t\tThe aim of the HTA process, developed within a healthcare facility, is to guide decision- makers on the “correct” acquisition or implementation of a health technology, from different viewpoints:
\n\t\t\t\t\n\t\t\t\t\t\t\tclinical : efficacy, risk/benefit rate, effect on current clinical procedures;
\n\t\t\t\t\t\t\ttechnological: technical and technological efficacy, technical specifications (technological and structural interfaces), management and maintenance activities;
\n\t\t\t\t\t\t\tenterprise : efficiency, productivity, impact on human (acceptability) and/or structural (e.g., need for building changes) and/or technological (e.g., need for HIS changes) resources.
The evaluation of the technical characteristics of a device can be performed in different ways. A technique based on the European network for Health Technology Assessment (EUnetHTA) model is described below.
\n\t\t\t\t\tThe EUnetHTA proposes an assessment scheme based on a basic unit, called assessment element. Each element defines a piece of information that describes the technology or the consequences or implications of its use, or the patients and the disease for which it is applied. An assessment element is composed of an evaluation area, a macro key performance indicator and a micro key performance indicator (see Figure 3a).
\n\t\t\t\t\tThe evaluation area (domain) represents a wide framework within which the technology is considered. It provides an angle of viewing the use, consequences and implications of any technology. The following domains are considered:
\n\t\t\t\t\tThe nature of the elements may vary across domains, since the consequences and implications are understood and studied differently in each domain. The following domains are considered:
\n\t\t\t\t\tHealth problem and current use of technology
Technical specifications
Safety
Clinical effectiveness
Costs and economic evaluation
Ethical analysis
Organizational aspects
Social aspects
Legal aspects.
A Macro Key Performance Indicator (Macro KPI or topic) represents a more specific area of consideration within any of the evaluation areas. One evaluation area is divided into several Macro KPIs. Similar Macro KPIs may be assigned to more than one evaluation area. A Micro Key Performance Indicator (Micro KPI or issue) is a specific area of consideration within any of the Macro KPI. One Macro KPI typically consists of several Micro KPIs, but it may also contain only one Micro KPI.
\n\t\t\t\t\tThe first task to accomplish in order to carry out the HTA process relates to the identification and definition of each KPI. To do this, the following steps are required:
\n\t\t\t\t\t\n\t\t\t\t\t\tStep 1 Literature search
\n\t\t\t\t\tA thorough literature analysis should be carried out by consulting the most important bibliographical sources such as clinical search engines (Pubmed, Medline, ISI Web of Knowledge, Cochrane Library, etc.), the national and international website of the HTA Agency (INAHTA, HTAi, EUnetHTA, Euroscan) or Institutes (ECRI, FDA, etc.), clinical practice guidelines, grey literature (technical reports from government agencies or scientific research groups, working papers from research groups or committees, white papers, or preprints). Other potential sources of data are manufacturers of the technology, clinicians, nurses, paramedics and patients.
\n\t\t\t\t\tThe search can be performed by using main keywords for the technology in question (for example limiting the research in “abstract/title” OR “topic” fields). The most interesting results of these searches are selected and details investigated in order to intensify and develop the assessment.
\n\t\t\t\t\t\n\t\t\t\t\t\tStep 2 Identify the assessment elements
\n\t\t\t\t\tThe analysis of the literature should therefore lead to the definition of the assessment elements, which are the core of the assessment. They are categorized into “evaluation area”, “macro KPI” and “micro KPI”. In order to make the assessment as objective as possible, the specific characteristics that support the assessment of a single area (and, subsequently, of the whole health technology) must be fully and measurably detailed, therefore objective and “instrumentally” measurable indicators are preferred. Moreover, those KPI that cannot be evaluated a priori should be excluded from the assessment.
\n\t\t\t\t\tTypically, the unit of measurement of KPIs may be:
\n\t\t\t\t\tmetric (e.g. spatial resolution, image uniformity, laser spot size, analytical specificity, etc.)
expressed as a percentage of coverage of the clinical/production/technical needs (e.g. percentage of coverage of analytical test panel; percentage of coverage of nominal “productivity”);
ON/OFF (presence/absence of a specific feature or functionality)
\n\t\t\t\t\t\t\t\t\tNumerical Value\n\t\t\t\t\t\t\t\t | \n\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t\tVerbal Scale\n\t\t\t\t\t\t\t\t | \n\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t\tExplanation\n\t\t\t\t\t\t\t\t | \n\t\t\t\t\t\t\t
1 | \n\t\t\t\t\t\t\t\tEqual importance of both elements | \n\t\t\t\t\t\t\t\tTwo elements contribute equally | \n\t\t\t\t\t\t\t
3 | \n\t\t\t\t\t\t\t\tModerate importance of one element compared to another | \n\t\t\t\t\t\t\t\tExperience and judgment favor one element over another | \n\t\t\t\t\t\t\t
5 | \n\t\t\t\t\t\t\t\tStrong importance of one element compared to another | \n\t\t\t\t\t\t\t\tAn element is strongly favored | \n\t\t\t\t\t\t\t
7 | \n\t\t\t\t\t\t\t\tVery strong importance of one element compared to another | \n\t\t\t\t\t\t\t\tAn element is very strongly dominant | \n\t\t\t\t\t\t\t
9 | \n\t\t\t\t\t\t\t\tExtreme importance of one element compared to another | \n\t\t\t\t\t\t\t\tAn element is favored by at least one order of magnitude | \n\t\t\t\t\t\t\t
2, 4, 6, 8 | \n\t\t\t\t\t\t\t\tIntermediate values | \n\t\t\t\t\t\t\t\tUsed to reach a compromise between two judgments | \n\t\t\t\t\t\t\t
Saaty scale
\n\t\t\t\t\t\tStep 3 Weight of the indicators
\n\t\t\t\t\tAfter the assessment elements have been identified, it is necessary to define the decision-making framework and in particular to estimate the value of the weight of each element: such activity must involve the whole multidisciplinary evaluation team.
\n\t\t\t\t\tThe definition of the weights, in fact, is a constituent part of the mathematical model of data processing, selected among those available in literature, such as the Analytic Hierarchy Process (AHP), expert systems based on Artificial Neural Network (ANN), and methodologies based on decision Fuzzy logic or Support Vector Machine (SVM).
\n\t\t\t\t\tWith reference to AHP, for example, a structured questionnaire with a series of “pairwise comparisons” between the assessment elements can be used: each team member will be required, therefore, to compare on a qualitative scale e.g., Saaty scale, see table 4) the relative importance of the two compared elements. Finally, after the comparison of all pairs, the weight of each indicator will be calculated.
\n\t\t\t\t\tStep 4 Value of the indicators
\n\t\t\t\t\tThe next step is to assess each technological alternative (the subject of the assessment) on the basis of the mathematical framework so far implemented. For this purpose, we assign values (quantitative or qualitative) to each lowest level KPI (usually a micro KPI, but also macro KPI and, rarely, even an evaluation area), on the basis of available literature data, and technical specifications or expert judgment. These values are then aggregated by the computational model to produce the value and rank of the single health technology.
\n\t\t\t\t\tA) The assessment element ; b) Combination of evaluation areas, macro KPI and micro KPI
\n\t\t\t\t\t\tStep 5 Results
\n\t\t\t\t\tThe results obtained by aggregating the values can be represented graphically or through numerical reports. In particular, results can be processed to allow, for example:
\n\t\t\t\t\tthe comparison between the technological alternatives in order to show the performance on each evaluation area and/or macro KPI and/or Micro KPI;
the comparison between weights of evaluation areas, macro and micro KPIs
analysis of the evaluation tree with evidence of weighted values for each technological alternative
etc.
Example of graphical representation of comparison of two health technologies
\n\t\t\t\t\t\tThe HTA report\n\t\t\t\t\t
\n\t\t\t\t\tThe final HTA report must provide the decision-makers with a clear, understandable summary of the information described above, in order to help them select the most appropriate technology. Moreover, it is essential to follow a standardized scheme, preferably one from a HTA agency or scientific community. However, it cannot be considered acceptable unless it contains the following sections:
\n\t\t\t\t\tdocument summary;
description of the technical characteristics and operating modalities of analyzed technologies;
summary of findings of literature search;
description of the criteria, indicators, macro and micro KPI;
definition of weights;
assigned values and mathematical processing method;
results (e.g., ranking, charts, graphs, etc.)
bibliography
Hundreds to many thousands of medical devices may need to be managed in a healthcare facility, with several million Euros being invested each year for the acquisition of new health technologies and for planned technology replacement, while thousands of maintenance processes per year are required in order to maintain the efficiency of these devices. As evident from the analysis of adverse events occurring during the last few years, serious incidents can often be related to the malfunctioning of medical devices. In particular, a high degree of obsolescence of the technologies, as well as missed, inadequate or improper maintenance, are among the possible causes of failure not attributable to the manufacturer. Therefore, in every healthcare facility, responsibility for the safe management of medical devices should be identified. The CED can provide a relevant contribution to the prevention of adverse events resulting from medical device failures by the technical and clinical assessment of the technologies to be acquired and proper management of maintenance. Different organizational models can be used to manage the above mentioned activities (Italian Ministry of Health, 2009): an internal service with employees of the healthcare facility; a mixed service, with internal control by clinical engineers as well as by means of maintenance contracts with manufacturers and technicians who may either be employees of the healthcare facility or of specialized companies; finally, an external service, with technical assistance entirely outsourced to a “global service” provider. Each of these three models has advantages and disadvantages. The first approach allows timely intervention and a better control of maintenance activities; however it is only justified when there is a sufficiently large quantity of technological equipment in the healthcare facility, and also requires the continuous training of the technical staff: Furthermore, maintenance contracts with manufacturers are still necessary for high-technology equipment. The second model permits flexibility as regards the organizational structure of the healthcare facility, internal control of processes, and a better integration of skills. The last organizational model is often preferred by healthcare facilities that do not yet have a CED; it allows organizational flexibility, but requires a careful selection of a qualified external company and authoritative supervision by the healthcare facility staff, otherwise control of the processes will be progressively lost and the quality of service will deteriorate.
\n\t\t\tMaintenance of medical devices has gradually evolved from the operational repair of out of order equipment to a management function aimed at preventing breakdown and failures, thus reducing risks associated with the use of medical devices, decreasing downtime and contributing to the improvement of diagnostic and therapeutic pathways, where technology is a key determinant. Healthcare facilities should identify responsibilities for maintenance and plan maintenance activities based on a detailed definition of methods, resources (i.e., operators, laboratories, measuring equipment, and maintenance contracts with external suppliers) and tools for supervision of the activities (e.g. dedicated software for the maintenance data management). To ensure adequate quality and safety standards and the rationalization of maintenance activities, a plan for maintenance and safety tests must be implemented, taking into account, for each device, the risks for patients and operators, degree of criticality and function of the device (e.g., therapeutic, diagnostic, or analytical). Within the European Community, preventive maintenance must be planned by the manufacturer prior to marketing the device. The 2007/47/EC Directive states that “the instructions for use must contain... details of the nature and frequency of the maintenance and calibration needed to ensure that the devices operate properly and safely at all times”. Preventive maintenance is of critical importance for ensuring the safe use of devices. Therefore, a preventive maintenance plan for each device must be defined, well documented and available at all operational levels to personnel responsible for maintenance tasks, including daily maintenance. Documentation should include informative documents and specific operating instructions which take into account both mandatory technical regulations and the service and user manuals provided by the manufacturer. Preventive maintenance is particularly relevant for life support devices, equipment for diagnosis and treatment, and devices identified as critical in relation to specific aspects such as the intended use of the device, class of risk, clinical features, type of location in which it is installed (e.g,. operating room, intensive care unit, ward), and presence of backup units. In carrying out the maintenance, the responsible technician must take into account all the maintenance instructions provided mandatorily by the manufacturer. Without affecting the liability of the manufacturer for any original product defects or faults, the person(s) performing maintenance will assume direct responsibility for all events deriving from this action. It is therefore essential that technicians, whether internal or external (see par. 5), have specific and proven experience. Training programs should be planned and preferably technicians should be trained by the manufacturers of the technologies which they maintain. Software for medical use deserves special consideration. Due to the complexity of systems and interactions, software behaviour may not be completely deterministic even when principles of good design practice are respected. Thus, software maintenance, which is usually performed by the manufacturer, should be supervised by the healthcare facility. Safety and performance tests must be periodically performed in order to ensure compliance with the essential safety requirements set by technical standards. The frequency of tests should be established taking into account criticality of device and according to reference guidelines. Particular attention is required in testing devices that can be used for critical applications (e.g., ventilators, anesthesia machines, infusion pumps, defibrillators, electrosurgical units) and for devices emitting or detecting ionizing radiations. Specific procedures and forms for different types of devices should be adopted to examine, measure, and verify the conformity of the device with the current mandatory technical standards and the instructions contained in the user manuals provided by the manufacturer. Dedicated equipment, for which calibration must be regularly performed and documented, should be used to measure parameters specific to each type of technology. Strategies for improving maintenance will only succeed if supervised effectively by external maintenance technicians in order to ensure their compliance with the agreed conditions (see par. 5). All relevant data relating to the life cycle of each device (from acceptance testing to disposal) must be recorded and made available at different operational levels. In order to ensure full traceability of the maintenance processes, preventive and corrective maintenance activities must be documented by detailed technical reports. In particular, preventive maintenance notice should be used to document the regularity of activities. Forms for maintenance requests to the CED must be defined and corrective maintenance notice should contain data useful for the identification of appropriate indicators (e.g. frequency of failures, time of first intervention, time to resolution, average downtime, distribution of failure types, maintenance costs, cost of spare parts), through which the condition of installed medical equipment can be analyzed.
\n\t\t\tEstablishing a complete and reliable inventory of medical equipment and ensuring the quality of the data is a complex task. Several different kinds of events, although rare, can lead to discrepancies between the inventory database and the technologies actually being used in a healthcare facility. These mismatches can be significantly reduced by establishing appropriate procedures and ensuring their strict observance. However, the large number of operators, devices and suppliers, the need to give priority to emergency care and the difficulty in directly and continuously monitoring the use of all devices in the healthcare facility, may inevitably produce such discrepancies. Failure to follow correct procedures for new equipment commissioning, for equipment transfer between departments, or for equipment disposal, are among the many possible events that could cause these mismatches. One possible solution is the use of Radio Frequency IDentification (RFID) tags and asset tracking systems. However, the use of this approach is limited because of ongoing debate about electromagnetic compatibility issues, and because the considerable cost of installation and management of these systems makes them still out of reach of most healthcare facilities. Until an advanced asset tracking solution is lacking in a healthcare facility, alternative strategies need to be implemented to keep the inventory data up-to-date. One way to monitor and update inventory data is through preventive and corrective maintenance or safety tests performed by CED technicians or by external service providers. Finally, it may be necessary to plan periodic inventory checks, which will be carried out independently or collaboratively by the CED and/or by the assets management office. Such controls may also provide an opportunity to remove devices that are no longer in use but are kept in stock and which may represent a source of risk.
\n\t\t\tDuring the last decades, planning health technology acquisitions has become of strategic importance for healthcare organizations, both at the national and at facility level. Such planning is also essential task for the reduction of clinical risk associated with the use of medical devices. The importance of acquisition planning is also determined by the considerable increase in technology investments, which is due to the increase in number and rapid technological evolution of medical devices and systems.
\n\t\t\t\tTherefore, healthcare organizations should define specific methods for planning the acquisition of health technology. Such methods should take into account the obsolescence of devices, the evolution of technical standards, the possibility of improving safety for patients and healthcare operators, the possible availability of innovative technologies for improving clinical performance, as well as considerations about actual or expected clinical needs, economic or technical feasibility, organizational changes, and investment priorities (e.g., innovative technologies vs device renewal). Moreover, the availability of adequate infrastructure, staff and consumables for the equipment must be foreseen in order to ensure full use of the benefits provided by the new technology. The decision to proceed with the acquisition should be conditional on the presence of a detailed clinical, economic and technical assessment with well defined comparative criteria, carried out by qualified and multidisciplinary staff and inspired by the principles of HTA (see par. 2). An equipment replacement plan is aimed at better identification of investment priorities for device renewal and may be based on the definition of a replacement priority value (RPV). RPV is an index which represents synthetically the level of urgency for the replacement of each device, permitting determination of a replacement priority ranking and planning of a progressive replacement of technologies (Fennigkoh, 1992). Variables considered by the RPV computational algorithm may come from different sources, principally the CED database and clinical activities records. Variables must be carefully chosen, according to the organization of the healthcare facility and based on data availability. In fact, the effort needed for collecting new data and keeping it up-to-date must be considered in order to limit the amount of new data to be collected and to make the best use of the data already available. A typical model for computing the RPV is based on the use of component indexes, with each index highlighting the impact on a specific aspect of the device replacement. A coefficient must be assigned to each component index in order to weight its contribution to the RPV. Possible aspects that might be taken into account, by defining specific numeric variables, are obsolescence of the device, maintainability (e.g. cost and availability of spare parts), reliability (e.g. downtime or number of failures), criticality, strategic impact, clinical efficacy, efficiency, clinical risk, potential for performance improvement. For example, the cost of replaced spare parts, the number of technical activities performed by the technicians of the CED, the annual cost of contracts and the cost of technical assistance by external suppliers will be taken into account in the computation of the component index for maintenance costs.
\n\t\t\tA quality assurance requirement for clinical assistance is the implementation of related processes based on the principles of best/good practice standards. In the field of management of medical devices, this concept is fundamental for meeting the need of retaining costs and providing effectiveness in patient care.
\n\t\t\tCEDs are also evaluated as to their ability to implement a policy of Good Management Practice of biomedical technologies (Cheng & Dyro, 2004). Related economic aspects, such as medical equipment maintenance costs, are a critical issue of such management (Table 5).
\n\t\t\t\n\t\t\t\t\t\t\tElement \n\t\t\t\t\t\t | \n\t\t\t\t\t\tFinancial | \n\t\t\t\t\t\tInternal processes | \n\t\t\t\t\t\tCustomer satisfaction | \n\t\t\t\t\t\tTraining and continuing education for CE staff | \n\t\t\t\t\t
\n\t\t\t\t\t\t\tMeasure\n\t\t\t\t\t\t | \n\t\t\t\t\t\tStaffing Beds per full-time equivalent employee Service/Acquisition ratio | \n\t\t\t\t\t\tPercent of IPM Complete IPM interval IPM time Repair time | \n\t\t\t\t\t\tAnnual survey | \n\t\t\t\t\t\tTime spent on these activities Certifications obtained | \n\t\t\t\t\t
A balanced performance scorecard for Benchmarking CE departments (Gaev, 2010a)
Clinical engineers play a fundamental role in determining the proper strategy for medical equipment maintenance and in recognizing the best available option for supporting these activities. More specifically, the CED is in charge of setting the expected level of performance, monitoring the quality and integrity of the delivered services, dividing activities between internal and external BMETs, and pursuing the goal of an expense reduction policy. For this reason, before maintaining biomedical technologies, CEDs should plan rational acquisitions, allotting part of the organization budget for service contracts.
\n\t\t\tA service contract is an agreement between a company and a user for the maintenance, in this case, of medical equipment during a specific period of time, usually for a fixed price which may be subject to changes if maintenance activities are performed outside the user’s location. The term “maintenance” typically includes inspection, preventive maintenance and repair. The terms and conditions of the contract usually stipulate the days and hours of service, the types of service, the response time, and which parts to be replaced are replaced free of charge” (Gaev, 2010b). This sort of contract can be extended to include the free loan of biomedical technologies. In this case, prices stated in the service contract are for consumables used for the equipment’s functions, and are increased to include maintenance costs.
\n\t\t\tReasons for having a service contract for a biomedical device are several. The first reason is the impossibility to provide a cost-effective service through in-house CED because of the lack of human and logistical resources. This is particularly common in hospitals where the problem of cost containment is approached with the sole objective of cost cutting and with no other financial or economic performance policy.
\n\t\t\tThe second main cause is that healthcare governance is particularly reluctant to assume responsibility for equipment maintenance, and the belief that original equipment manufacturer (OEM) service contracts represent the “gold standard” is difficult to remove. On the other hand, for certain classes of medical devices (those characterized by high-technological complexity or high consumable costs, such as clinical chemistry analyzers), service contracts seem to be the only realistic solution for accommodating their management costs. The main issues which have to be discussed and negotiated in the drawing up of a service contract are: inspection and preventive maintenance, repair, spare parts, legal and financial aspects.
\n\t\t\tThe term “Inspection and Preventive Maintenance (IPM)” covers all the activities involved in cleaning, lubricating, adjusting, checking for wear, and perhaps replacing components that could cause total breakdown or serious functional impairment of the equipment before the next scheduled inspection (Subhan, 2006).
\n\t\t\tThese activities are well-described in the manufacturer’s service manual and are aimed at avoiding the breakdown of a medical device in use, without any apparent warning of failure. Manufacturers are obliged to explicate preventive maintenance actions to healthcare operators or BMETs, and to suggest the minimum inspection frequency. The definition and respect of a timetable for IPM of all medical equipment is fundamental for reducing risk for patients and users, and preventing excessive repair costs by providing timely interventions; and it should be the CED’s first priority, and should be decided before carrying out preventive maintenance activities.
\n\t\t\tContracts should clearly explain the necessity of making known to all concerned the timetable for the maintenance by external technicians at the beginning of the year, in agreement with the CED and the healthcare personnel. This will allow the organization of clinical activities for healthcare operators and the possibility to enter the whole agenda into the biomedical technology maintenance management system. One other particular observation relates to the availability (at the charge of the contractor) of software update if required for the correct operation of the biomedical instrumentation. The last consideration relates to the possibility for CEDs (according to their competence) to evaluate the IPM requirements of medical equipment (Table 6) and to modify the service intervals recommended by the manufacturer, to obtain a more cost-effective maintenance without adversely affecting patient safety.
\n\t\t\tRepair (corrective maintenance) is a process to restore the physical integrity, safety and/or performance of a device after a failure. Aspects to be considered pertain to economic, safety and logistic concerns. Contracts should explain who can call for technical support: this aspect is fundamental for organizing the internal maintenance process. One possible solution would be for the healthcare personnel to first of all attempt to resolve the problem by telephone (with proper manufacturer’s customer support), and to define an internal procedure for advising the CED of the failure. In this way, the CED can monitor failure resolution time by the manufacturer’s technicians by means of its maintenance management system.
\n\t\t\t\n\t\t\t\t\t\t\tDevice \n\t\t\t\t\t\t | \n\t\t\t\t\t\t\n\t\t\t\t\t\t\tShortest IPM Interval\n\t\t\t\t\t\t | \n\t\t\t\t\t\t\n\t\t\t\t\t\t\tLongest IPM Interval\n\t\t\t\t\t\t | \n\t\t\t\t\t
Electrosurgical unit | \n\t\t\t\t\t\t6 months | \n\t\t\t\t\t\t12 months | \n\t\t\t\t\t
Exam light | \n\t\t\t\t\t\t12 months | \n\t\t\t\t\t\tNo IPM performer | \n\t\t\t\t\t
Physiologic monitor | \n\t\t\t\t\t\t12 months | \n\t\t\t\t\t\t24 months | \n\t\t\t\t\t
Pulse oximeter | \n\t\t\t\t\t\t12 months | \n\t\t\t\t\t\tNo IPM performer | \n\t\t\t\t\t
Variations in IPM intervals for selected equipment, proposed by ECRI Inst. (2010)
Another significant aspect related to maintenance contracts is the definition of “bad-management” of biomedical technologies by healthcare personnel which may cause failure of the equipment. Some manufacturers are reluctant (or do not agree) to repair equipment under contract if abuse or improper use by hospital staff caused the failure. It is essential that the internal training of healthcare staff makes them aware of their responsibility for the correct use of biomedical equipment.
\n\t\t\tMoreover, in the contract, clinical engineers should define a way to evaluate the performance of OEM technicians, and stipulate the right to suspend the service contract in the event of low-quality maintenance work.
\n\t\t\tA common aspect of IPM and repair contracts is the possibility of a partnership for maintenance activities between the OEM technical support and the BMETs (internal or outsourced). Some manufacturers only permit maintenance activities by qualified (and certified by the OEM itself) technicians. Positive results of partnership contracts were showed just a few years ago. A first Italian joint project between OEMs and in-house service was started in 2002 (De Vivo et al., 2004): in-house personnel received adequate training, both generic (basic principles on which devices work) and specific (how to use, repair and maintain a particular model), for maintaining 90 medical devices (mostly monitoring equipment, ventilators and anesthesia units) in shared OEM/internal BMETs maintenance contracts.
\n\t\t\t\n\t\t\t\tFigure 5 summarizes the success of this program. One important effect was the increased awareness of the OEMs about the need for a rational selection of an effective preventive maintenance program in which service procedures and frequencies are based on real world feedback, efficacy of activities are measured and areas needing improvement are identified.
\n\t\t\tClinical Engineers are also in charge of compiling technical reports related to maintenance activities (for instance, by means of an appropriate software system, see par. 6). These data are essential for monitoring the quality of OEM services, and claiming economic and legal penalties. Service contracts should also clearly explain the accuracy level of report writing, to avoid possible future disputes.
\n\t\t\ta) Percentage of in-house repairs (July 2002-March 2004). The number of in-house repairs reached 90% and more after one year and continue to grow as in-house personnel sharpen their required basic skills. b) Number of OEM and in-house repairs (years 1999-2004). The decrease in OEM corrective maintenance was soon significant: as a consequence, OEMs were able to focus their attention on accurate preventive maintenance in order to prevent certain predictable failures. c) Percentage reduction of annual maintenance fees. Significant discounts were obtained based on the percentage of in-house corrective maintenance, justifying the cost related to internal technicians and the energies needed to set up the whole system.
Service contracts should include a specific paragraph on spare parts. OEM contracts usually lack the inclusion of them or any specification of the condition (e.g. new, refurbished) of parts used for maintenance and repair (Gaev, 2010b). It should be the duty of clinical engineers to assess the need for spare parts and include them in the contract, in dedicated annexes.
\n\t\t\tThe economic assessment of service contracts is done using the definition of financial performance indexes. The most common index is the service cost/acquisition cost (S/A) ratio, i.e. the total cost to deliver a service, including parts and labor, divided by the acquisition cost of the equipment. Services delivered by OEMs (or third-party service suppliers) under a full-service contract usually include IPM and repair. The cost of same service delivered by an in-house CED is computed from the amount spent on parts and CED labor (labor hours) multiplied by the “loaded” rate including salary, benefits and other over-head expenses. In-house service is generally less expensive (50 percent less) than full-service OEM contract, even if this estimation varies significantly according to the equipment category. A recent ECRI review shows that imaging and high-tech laboratory equipment has a higher S/A ratio and is thus more costly to maintain than general biomedical equipment, even if this ratio may vary greatly due to institutional (e.g., teaching vs non teaching institution), logistics (e.g., urban vs rural hospital) as well as operational (e.g., low vs high negotiated acquisition price) differences (Gaev, 2010a).
\n\t\t\tParticular consideration should be given to the drawing-up of penalty clauses for the possibility of non-compliant service, the latter defined in terms of technical response time and equipment uptime/downtime. Moreover, competitive benchmarking for service contracts should also take into account fees for service outside of contract work hours, and any minimum charges required for travel time, service time, and work performed outside of the usual contract provisions. However, to make effective the use of penalty clauses, essential tools have to be set in place such as the computerized management of processes, implementation of a contact center (phone or online) for maintenance requests, systematic review of the quality of maintenance activities, failure analyses, and strict control of performance indicators and maintenance costs.
\n\t\tAny action undertaken to improve the management/control of medical devices in a healthcare facility through an efficient and effective organization of maintenance and technology assessment activities, requires the implementation of operating procedures that enable the standardization of CED processes. However, the rapid evolution of health technologies during the last decades and the spread of heterogeneous technologies, besides bringing undeniable clinical improvements, have resulted in a considerable increase in technology investments, with the subsequent need for tools that can aid decision making in acquiring new technologies and managing the existing ones. To achieve the double goal of correctly applying and automating procedures and of implementing a model for the appropriate management of available resources and the proper definition of priorities, a comprehensive and reliable dataset for health technologies as well as an appropriate software tool to support data management will be required. Electronic archives are thus essential for storing all data and all events in the life of the medical devices managed by the CED, from the technology assessment that should always precede their acquisition until their disposal. Such a tool will permit safer documentation and reporting of the maintenance and management activities, sharing of information between the CED and other hospital staff, a dramatic improvement in data search, provision of summary statistics, and the definition of indicators that may contribute to the proper management of health technologies. The organization of this database may vary markedly depending on heterogeneous factors such as healthcare facility organization, technical and administrative management policies, number of devices, and resources dedicated to data management. The opportunity to support and significantly improve the management of medical equipment makes it advisable to implement a solution that can be configured and easily updated according to the evolution of specific needs. The configurable features of the system should include database design, user interfaces, queries, reports and statistics. The possibility to configure the database is useful not only for adding tables and fields, but also for the development of new features and adaptation of the software to the organizational structure of the healthcare facility. Configurable user interfaces should include at least the appropriate forms for inventory, acceptance testing, safety and performance tests, maintenance processes, preventive maintenance plan, maintenance contracts, disposal of devices, and administrative data management. Customizable configurations for different users should be guaranteed, in order to adapt the software according to the role and responsibilities of each user, with different data visualization and operating permissions. System users should be allowed to extract and export data in convenient formats (e.g., spreadsheets) for offline processing. Templates for standard documents (e.g. acceptance testing reports, maintenance reports) must be available and it should be easy to obtain automatically filled in and ready-to-print documents. It should be possible to analyze data with a configurable statistics dashboard. Such a system architecture would be suitable for developing methods for health technologies management and for defining indicators for the implementation of a technology replacement plan, the identification of maintenance priorities, and the optimization of resources allocation. Ultimately, being able to customize the software makes it possible to update the structure and configuration of the system according to the organization and evolution of operational requirements specific to a particular healthcare facility, and also makes it a suitable tool to support the development of processes. This feature is also particularly relevant for the purpose of satisfying the requirements for certification and the standards for national and international accreditation. The configuration should be performed or at least supervised by the CED staff, who best know the specific needs of the organizational context in which the software is to be used. Another advisable solution is to adopt systems that are accessible via the facility’s intranet. Web-based systems that do not require any client-side software installation are useful for sharing information between the different actors involved and can improve the automation of processes for Health Technology Management (HTM). Moreover, with web-based systems it is possible for health operators to access many support features for the management of technologies. They can submit online requests for corrective maintenance, monitor the real-time evolution of submitted requests, search the database for devices, preventive maintenance plan or safety tests, and receive automatic e-mail notifications when certain events occur (e.g., maintenance processes closed by biomed technicians, reminders for scheduled maintenance). This approach also has the advantage that only one data entry is needed (e.g., biomed technicians no longer have to re-enter data that have already been entered by the health operators on the maintenance request form). Obviously, all users should be trained in at least the basic principles of the system. The use of such a system for the management of medical devices can be extended to (or integrated with) the management of other technological facilities, ICT equipment, and other hospital assets.
\n\t\t\tA number of advantages for budget management can be gained by using computerized procedures for online submission of requests from heads of hospital units for the acquisition of new medical devices. Specification of medical device type according to a standard nomenclature system could be required, which would avoid the use of disparate terms for the same equipment. Also, the use of required fields in the electronic request form (e.g., reason for the acquisition, expected benefits, consumables needed) would ensure that all requests contain the essential information for their proper assessment. The medical board, with the support of the CED, would then have the right tools to manage the submitted requests in a uniform manner and make an objective analysis. assign a priority ranking to each request, and finally decide which ones to approve and which to reject. This approach could also be useful for the activation of hospital-based HTA (HB-HTA) processes (see par. 2). Furthermore, the authorization process (i.e., approval by department directors and medical board or medical devices committee) can be automated and differentiated according to the type of acquisition (e.g. property, loan, service, rental, clinical trial). Approval of the request will be automatically notified and immediately available online. The technology renewal plan managed by the CED may be integrated and partially automated in the software by implementing an algorithm for calculating the replacement priority value (see par. 4). Following the approval of requests for new acquisitions and replacement of medical devices, the automation of CED processes would provide valuable support for the management of data and documents relating to the assessment and acquisition of technologies. Information concerning single budget items (e.g., type of acquisition, number of requested devices, allocated budget) and on acquisition progress (i.e., end of the market survey, drafting and issuance of the technical assessment, date of order by the administration, supplier name) can be shared between the CED and the healthcare facility administration, with automatic update of acquisition progress and online availability of documents for each budget item. At all stages, starting from submission of the requests, only a single data entry is needed.
\n\t\t\tIn a computerized system for managing CED’s processes, each medical device has its own inventory record containing the data relevant to its management (e.g., device model, accessories, system configuration, owner hospital unit, location, administrative data). Each device in the inventory must be uniquely identified, and the CED must place an identifying label on it. As stated above, the adoption of a standard medical device nomenclature for model identification is also strongly recommended. If a web-based system is used, health operators will be able to search for inventory records and obtain lists of devices that can be exported onto spreadsheets. For each device in the inventory, the acceptance testing must be registered in the system. The status of the device can be updated automatically and an e-mail notification sent upon completion of testing.. In order to keep the inventory data up-to-date, in addition to routine administrative tasks, periodic inventory checks must also be made. In this regard, mobile units (e.g., PDA) equipped with a tag (e.g. barcode, RFiD) reader, properly configured and synchronized with the CED software system, can be a useful tool. This approach allows easier tracking of devices and verification of equipment location and condition, as well as updating of system components. Another useful feature is the online availability of documents. These could include pre-acquisition documents (e.g., market survey, technical assessment, order form), user and service manuals, acceptance testing documents and training course forms, as well as pictures of system configurations and accessories.
\n\t\t\tMaintenance processes management could exploit the availability of an appropriate software tool. As stated above, a useful feature is the possibility for health operators, in case of failure of a medical device, to request corrective maintenance online. Maintenance activities should then be recorded in the system by CED biomed technicians. CED can enter and update the maintenance plan (i.e., the preventive maintenance activities for which both internal technicians and external maintenance personnel will be appointed) and share it, as well as related information (e.g. maintenance progress, e-mail notification of upcoming preventive maintenance), with all hospital units involved. Health operators should be allowed to retrieve and export lists of maintenance requests. Thousands of safety and performance tests are performed on medical devices each year by the CED. Thus the availability of test reports to health operators is only possible by implementing an automatic upload system. Radiology equipment deserves a particular mention in that it is usually managed by both CED and the Medical Physics Unit. This requires sharing of information on preventive and corrective maintenance and quality controls. Finally, the software tool can also be used to facilitate the management of spare parts. Online access to maintenance documents (i.e., preventive and corrective maintenance activities, safety and performance test reports, administrative documents) is another desirable feature. The availability of such electronic information enables the CED to analyze the history of maintenance processes for each device, to improve monitoring of maintenance activities performed both by CED technicians and by external maintenance personnel, to verify the compliance of suppliers with maintenance contracts, to gather downtime statistics, and to generate summaries of maintenance costs. Finally, algorithms can be defined and implemented to combine device replacement priority value (see par. 4) and maintenance priority rank for immediate identification of the most urgent corrective actions. Automated information sharing can also be helpful for the disposal of devices. The way this feature can be configured depends on the specific organization. For example, CED could be in charge of notifying the hospital unit of device disposal, while the physical removal of the device would be the responsibility of the facility handling service. An automatic e-mail notification of disposal confirmation to the CED would allow an easier tracking of out of order devices, thus reducing inconvenience and risk for patients and health operators.
\nTime series is a series of data points which are collected by recording a set of observations chronologically. Examples of time series include speech, human activities, electrocardiogram (ECG), etc. Recently, time series classification has attracted great interests and initiated various researches. However, the nature of time series data, including the large size of data, the high dimensionality and the continuously updating scheme of time series, makes time series classification a more challenging task.
\nTime series classification is widely applied in different fields such as in astronomy [1] to classify the brightness of a target star, in medical science to diagnose cardiac disorders [2] or to recognize human activities [3, 4], and in computer science for speech recognition [5, 6]. To handle time series classification, several techniques were proposed, which can be aggregated into three categories: model based, distance based and feature based.
\nThe first category of time series classification approaches consists of building a model for each class by fitting its parameters to that class. Examples of such approaches are the autoregressive (AR) model [7] and the hidden Markov model (HMM) [8] which are limited to stationary and symbolic non-stationary time series respectively.
\nThe second category relies on developing distance functions to measure the similarity (or dissimilarity) between two time series and on selecting a good classifier, such as dynamic time warping (DTW) distance [9, 10]. But these approaches are computationally expensive.
\nThe third category consists of extracting meaningful features from the time series. Examples of such approaches include the discrete Fourier transform (DFT) [11], the Short-time Fourier transform (STFT) [12], the discrete wavelet transform (DWT), principal component analysis (PCA), singular value decomposition (SVD), sparse coding [13], and shapelets [14].
\nMeanwhile, automatic feature-based approaches using deep learning models rely have been successfully applied to time series classification, classification problems, especially convolutional neural networks (CNNs) which are regarded as the most successful and commonly used deep learning model. In [5, 6], authors address the problematic of speech recognition whereby speech signals have similar patterns within different frequency band locations which convey a different meaning. A solution to this problem is to employ a limited weight sharing CNN [6] where weight sharing is limited only to local filters which are close to each other and which are pooled together in the subsampling layer. Another approach based on tiled CNN architecture with a pre-training stage (an unsupervised learning algorithm named topographic ICA) was proposed by [15], which showed its superiority over traditional CNN on small time series datasets. A tiled CNN [16] is a CNN which unties weights locally and uses a regular “tiled” pattern of tied weights that requires only that hidden units k steps away from each other to have tied weights. Another relevant CNN architecture for time series classification named multi-scale convolutional neural network (MCNN) [17] was introduced where each of the three transformed versions of the input (which will be discussed in Section 3.1) is fed into a branch i.e., a set of consecutive convolutional and pooling layers, resulting in three outputs which are concatenated and further fed into more convolutional and pooling layers, fully connected layers and a softmax layer to generate the final output. Training all parameters is done jointly using back-propagation. Another attempt to enhance time series classification was proposed in [18], which employs the same idea of multiple branches within the CNN architecture, except that the input is not a different transformed version of the time series signal fed into each branch, but rather a duplicate of the same time series signal fed into all the branches (three branches). However, different convolutional filter sizes are applied per branch in order to capture the multi-scale characteristics of the time series. Two other CNN proposals to time series classification were suggested in [19], namely fully convolutional networks (FCN) without subsampling layers, and residual networks (ResNet). FCNs [20] are defined as networks which have convolutional layers only and no fully-connected layers, whereas ResNet [21] is a type of specialized neural network that solves the “vanishing gradient” problem when having many layers within the network, by using residual blocks which take advantage of residual mapping to preserve inputs. By adding batch normalization layers into FCN and ResNet, and by replacing the fully connected layers with a global pooling layer in the FCN, these two deep learning models seem to yield comparable or better results than MCNN [17]. An ensemble method of deep learning networks named LSTM-FCN is proposed in [22] is proposed and consists of feeding the same time series input into two branches: an FCN and Long Short Term Recurrent Neural Network (LSTM) block [23], producing two outputs which are concatenated and then passed onto a softmax classification layer. Another attempt to helps the CNN converge faster and better to the minima was made by Guennec et al. [24] who propose to perform data-augmentation techniques (further described in Section 3.1) and pre-train each layer in an unsupervised manner (using an auto-encoder) using unlabeled training time series from different datasets. For multivariate time series, only few research papers based on CNNs were published (such as [3, 4, 25, 26]). Zheng et al. [25] proposed a multi-channels deep convolution neural network (MC-DCNN), each branch of which takes a single dimension of the multivariate time series as input and learns features individually. Then the MC-DCNN model combines the learnt features of each branch and feeds them into a fully connected layer to perform classification. And, to further improve the performance, authors also suggested to pre-train the MC-DCNN first by applying an unsupervised initialization via the convolutional auto-encoder method. Meanwhile, a different CNN architecture for multivariate time series classification was introduced in [3, 4, 26], which treats the 3-, 12-, and 9-variate time series inputs (in [3, 4, 26] respectively) as a 3-, 12-, and 9-channel inputs and convolves them as a whole instead of convolving each channel of the input separately as performed in [25]. Authors of this architecture argue that, by separating multivariate time series into univariate ones just as in [25], the interrelationship between different univariate time series may be lost and thus will not be mined/extracted.
\nIn this paper, we aim at presenting a review on our CNN approaches for time series classification. Our review discusses our CNN contributions at the data-level and at the algorithm-level. Our paper is organized as follows. In Section 2, some preliminaries about time series are introduced. In Section 3 reviews existent data-level techniques are presented, our data-level technique is reviewed, and experiments as well as results of our technique are laid out. Section 4 describes our algorithm-level approaches for time series classification, with experiments conducted and results analyzed. Section 4.3.3 concludes our paper with future perspectives.
\nUnivariate and multivariate time series data. Time series inputs can be categorized into: (i) Univariate Time series which have only a single variable observed at each time and thus resulting in one channel per time series input, and (ii) Multivariate Time series which have two or more variables observed at each time, ending up with multiple channels per time series input. Most time series analysis methods focus on univariate data as it is the simplest to work with. Multivariate time series analysis considers simultaneously multiple time series, which, in general is much more complicated than univariate time series analysis as it is harder to model and often many of the classical methods do not perform well.
\nRaw data or extracted signals. A raw time series is a series of data points indexed in time order i.e., a sequence of discrete-time data taken at successive equally spaced points in time. In time series classification tasks, some authors choose to evaluate the performance of their approaches using raw time series data taken from a specific field/domain while some others prefer to use public datasets in which the raw time series is already segmented and converted into a set of fixed-length signals. Indeed, several research papers using CNNs [17, 18, 19, 22, 24, 26, 27] build their experimental studies on the UCR time series classification archive [28] which consists of extracted short signals. Nonetheless, this benchmark is composed of relatively small datasets (with a small number of instances), which makes the CNN less efficient knowing that CNNs require large training sets for training. Furthermore, in most of the cases, fixed-length signals cannot be further encoded into new representations (which are discussed in Section 3.1), as opposed to raw time series. These issues have led authors of [3, 4, 5, 6, 15, 25, 29, 30] to use raw time series data instead.
\nThroughout this section, we show several approaches used in the literature to pre-process time series by re-framing them into new representations for a further CNN implementation. Indeed, a raw time series needs to be converted into a set of fixed-length vector or matrix inputs before being fed into the CNN. Then, we discuss our data-level approach (of previous works [3, 4]) based on the Stockwell transform method.
\nGiven a sequence of values for a time series dataset, values at multiple time steps can be grouped to form an input vector while the output corresponds to a specific label given to this vector (generally provided by an expert). The use of time steps to predict the label (e.g., the class) is called the Sliding Window method and is explained in the algorithm below. The width of the sliding window \n
Algorithm 1. Sliding window’s algorithm
procedure SlidingWindow(T, L, s)
\n
\n
while \n
\n
\n
end while
return \n
end procedure
Several research papers have focused mainly on applying some pre-processing to raw time series before being fed into the CNN. In this subsection, we present some important contributions which demonstrated that applying changes to the signals can further improve the CNN performance.
\nSeveral attempts have been made in order to encode raw time series as a matrix representation (e.g., 2D images) such as the Gramian Angular Field (GAF) [15], the Markov Transition Field (MTF) [15], Recurrence Plots (RP) [27], and stacked time series signals [29, 31], multivariate time series are treated as a 2D time-space input signals with one dimension denoting discrete time flows and the other corresponding to different channels of the multivariate time series.
\nAnother type of data pre-processing based on applying transformation to data is performed in order to augment the data, thereby ensuring a better CNN training and thus a higher performance. For instance, the window slicing method [24] trains the CNN using slices of the time series input, then at test time classifies each slice of the test time series using CNN, and performs majority voting to output the predicted label. The window warping method [24] consists of warping a randomly selected slice of a time series by speeding it up or down, producing a transformed raw time series. Then, this latter is further converted into fixed-length input signals/instances via window slicing. Another attempt of augmenting time series is suggested in [3] where either small noise or smoothing is applied to the raw time series. Other transformations were also considered in [17] such as down-sampling to generate versions of a time series at different time scales, and spectral transformations in the frequency domain by adopting low frequency to remove noise from time series inputs.
\nKnowing that random noise and high-frequency perturbations present in the time series data can interfere tremendously with the learning process and that it is hard to capture useful features with the presence of noise in raw time series data, some works [5, 30] proposed to apply the Fast Fourier transform (FFT) and convert the raw time series into a set of frequency domain signals which serve as inputs for the CNN training process.
\nInstead of employing the FFT which is restricted to a predefined fixed window length, we choose to adopt the Stockwell transform (ST) as our preprocessing method for CNN training [3, 4]. In this section, the ST method is defined, its implementation on real world applications is detailed, and its experimental results are analyzed.
\nThe advantage of the ST over the FFT is its ability to adaptively capture spectral changes over time without windowing of data, resulting in a better time-frequency resolution for non-stationary signals [32]. To illustrate the ST method, let \n
where \n
The discrete Stockwell transform (DST) is given by,
\nwhere \n
Given a \n
Apply an N-point DFT to calculate the Fourier spectrum of the signal \n
Multiply \n
For each fixed frequency shift \n
Note that there is a DST coefficient calculated for every pair of\n
A SMM is defined as a repetitive movement which is regarded as one of the most apparent and relevant atypical behaviors present within children on the Autism Spectrum. Thus, detecting SMM behaviors can play a major role in the screening and therapy of ASD, thus potentially improving the lives of children in the spectrum.
\nDataset. The SMM dataset used for training the CNN is derived from [33] and consists of raw time series of acceleration signals collected by three-axis wireless accelerometers (located at the torso, left wrist and right wrist) from six atypical (e.g., autistic) subjects in a longitudinal study. Activities including SMMs (body rocking, hand flapping, or simultaneous body rocking and hand flapping) and non-SMMs were engaged by subjects and were labeled (annotated) by an expert as SMM or non-SMM. Two to three sessions (9 to 39-min long) were recorded per participant, except subject 6 who was observed only once in Study 2.
\nPre-processing. The data collection called “Study1” and “Study2” were recorded at a sampling frequency of 60 and 90 Hz respectively. So, to equalize the data, the 60 Hz signals are resampled and interpolated to 90 Hz. Next, data of both sensors go through a high pass filter with a cut-off frequency of 0.1 Hz in order to get rid of noise.
\nAfterwards, data is turned into fixed-length vector samples either in time-domain (using the sliding window) or in frequency-domain (using ST). Time-domain samples are obtained by segmenting raw data using a one second window (e.g., \n
On the other hand, frequency-domain samples are obtained by deriving ST for every other 10th sample, and by selecting the proper best frequency range. Considering that 98% of the FFT amplitude of human activity frequencies is contained below 10 Hz, the ST frequencies are first chosen to be between 0 and 10 Hz, which yields bad CNN classification performance. And, after considering Goodwin’s observation that almost all SMMs are contained within a frequency range of 1–3 Hz, we chose a new frequency range of 0–3 Hz which produced higher CNN classification performance. So, computing the ST generates multiple input samples (vectors of length 50), each containing the power of 50 frequencies (\n
CNN training. The purpose is to analyze intersession variability of different SMMs by training one CNN per domain (time or frequency domain) per subject per study. In other words, feature learning that is performed is specific to one domain (time and frequency), one subject \n
Dataset. The dataset used for HAR is the PUC dataset [34] which consists of 8 hours of human activities collected at a sampling frequency of 8 Hz by 4 tri-axial ADXL335 accelerometers located at the waist, left thigh, right ankle, and right arm. The activities are: sitting, standing, sitting down, standing up, and walking.
\nPre-processing. The PUC data is further converted into time and frequency domain signals. In time-domain, a 1 s time window (e.g., \n
CNN training. In this experiment, one CNN is trained for each domain (time and frequency domain), with a 10-fold cross-validation. CNN architecture and parameters are set the same as in the SMM recognition task.
\nTable 1 summarizes accuracy and F1-score results of CNN (in both time and frequency domains) for both the SMM recognition and HAR tasks. For SMM recognition, as opposed to other subjects, Subject 6 within Study 2 had only one session recorded; thus, no CNN was trained for this subject. We observe that all frequency-domain CNNs (of all subjects in all studies) perform better than time-domain CNNs by 8.52% (in terms of the mean F1-score). This suggests that ST eliminates all noisy information and thus helps the CNN capture meaningful features.
\n\n | F1-scores SMM recognition | \nAccuracies HAR | \n|||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
\n | Study 1 | \nStudy 2 | \n|||||||||||
\n | S1 | \nS2 | \nS3 | \nS4 | \nS5 | \nS6 | \nS1 | \nS2 | \nS3 | \nS4 | \nS5 | \nMean | \n|
Time-domain CNN | \n91.23 | \n76.76 | \n84.95 | \n93.38 | \n86.41 | \n95.11 | \n95.97 | \n75.67 | \n60.17 | \n91.68 | \n82.55 | \n84.90 | \n99.90 | \n
Frequency-domain CNN | \n96.54 | \n78.41 | \n93.62 | \n96.46 | \n95.74 | \n98.58 | \n96.07 | \n95.27 | \n85.03 | \n98.03 | \n93.88 | \n93.42 | \n95.98 | \n
Performance rates of time-, and frequency-domain CNNs for the SMM recognition (in terms of F1-score) and Human Activity Recognition referred to as HAR (in terms of accuracy). Highest rates are in bold.
However, as opposed to these results, comparing results of time and frequency domain CNNs on the Human Activity Recognition (HAR) task demonstrates the efficiency of time over frequency by 3.92% in terms of accuracy (as shown in Table 1). These contradictory results can be explained by the difference in the chosen ST frequency range for SMM recognition and that of HAR. Indeed, in SMM recognition, the frequency range of the ST was carefully chosen to cover almost all SMMs (0–3 Hz), resulting in optimal frequency-domain samples (containing full and noise-free information) which produced better CNN parameters. Meanwhile, the ST frequency range for HAR (0–8 Hz) may be a short/small range which generated frequency-domain samples that may have lost relevant information. Indeed human activity frequencies fall between 0 and 20 Hz (with 98% of the FFT amplitude contained below 10 Hz). Thus, in order to train CNNs with frequency-domain signals, it is necessary to analyze raw time series to come up with the proper ST frequency range which covers all valuable information needed for the recognition task.
\nCNNs were developed with the idea of local connectivity. Each node is connected only to a local region in the input. The local connectivity is achieved by replacing the weighted sums from the neural network with convolutions. In each layer of the CNN, the input is convolved with the weight matrix (e.g., the filter) to create a feature map. As opposed to regular neural networks, all the values in the output feature map share the same weights so that all nodes in the output detect exactly the same pattern. The local connectivity and shared weights aspect of CNNs reduce the total number of learnable parameters, resulting in more efficient training and learning in each layer a weight matrix which is capable of capturing the necessary, translation-invariant features from the input.
\nThe input to a convolutional layer is usually taken to be three-dimensional: the height, weight and number of channels. In the first layer this input is convolved with a set of \n
The output feature map from the first convolutional layer is then given by convolving each filter \n
where \n
In each subsequent layer \n
This output is then passed through the non-linearity to give \n
The output is then fed into a pooling layer (usually a max-pooling layer) which acts as a subsampling layer. The output map \n
where \n
Multiple convolution, ReLU and pooling layers can be stacked on top of one another to form a deep CNN architecture. Then, the output of these layers is fed into a fully connected layer and an activation layer. The output of the network after \n
In previous CNN works, several attempts have been made to extract the most relevant/meaningful features using different CNN architectures. While works [17, 24] transformed the time series signals (by applying down-sampling, slicing, or warping) so as to help the convolutional filters (especially the 1st convolutional layer filters) capture entire peaks (i.e., whole peaks) and fluctuations within the signals, the work of [18] proposed to keep time series data unchanged and rather feed them into three branches, each having a different 1st convolutional filter size, in order to capture the whole fluctuations within signals. An alternative is to find an adaptive 1st convolutional layer filter which has the most optimal size and is able to capture most of entire peaks present in the input signals. By obtaining the most appropriate 1st convolutional layer filter, there will be no need to apply multiple branches with different 1st convolutional layer filter sizes, and no need to apply transformations such as down-sampling, slicing and warping, thus requiring less computational resources. The question of how to compute this adaptive 1st convolutional layer filter is addressed in [4]. In this section, we will discuss the approach based on the adaptive 1st convolutional layer filter. Next, to prove the efficiency of this/our approach, an application on SMM recognition is conducted and results are analyzed.
\nIn CNNs, multiple hyper-parameters are to be chosen carefully in order to retrieve the best classification rate, including model hyper-parameters which define the CNN architecture, and optimization hyper-parameters such as the loss function, learning rate, etc. Model Hyper-parameter values are generally chosen based on the literature and on the trial-and-error process (through running experiments with multiple values). A conventional approach is to start with a CNN architecture which has already been adopted in a similar domain to ours, and then update hyper-parameters by experimentation.
\nIn our study, we focus on the convolutional layer filter (also known as “Receptive field”). Conventionally, the 1st convolutional layer filter has one of the following sizes: \n
Given a population of signals with mean \n
For the sample median \n
Experiments are conducted on the SMM Recognition task. The dataset and experimental setup are the same as in Section 3.2.2. The inputs used are either time-domains acceleration signals of size \n
As explained in the methodology, the first step to determine the size of the 1st convolutional layer filter is to collect 30 random signals (for each of the time and frequency domain SMM signals) that contain at least one peak and to randomly pick 30 peaks from these signals. Histograms (a) and (b) of Figure 1 represent frequency distributions of the 30 peak lengths for time and frequency domain signals respectively. Afterwards, the computed time and frequency domain medians (9 and 10 respectively) are applied as the optimal size of the 1st convolutional layer filter for the time and frequency domain CNNs respectively.
\n(a) and (b) Histograms and box plots of the frequency distribution of 30 peak lengths present within 30 randomly selected time and frequency domain signals respectively.
In order to prove the efficiency of this adaptive 1st convolutional layer filter approach, we run experiments on different time and frequency domain CNN architectures by varying the size of the 1st convolutional layer filter between 7 and 11 across both architectures. Performance rates in terms of the F1-score metric are displayed in Figure 2. In time-domain, an increase in the size of the 1st convolutional layer filter from 7 (∼ a time span of 0.078 s) to 9 (∼ a time span of 0.1 s) results in an increase of 3.26%, while an increase of the filter size from 9 to 10 (∼ a time span of 0.11 s) and 11 (∼ a time span of 0.12 s) diminishes the performance of the network. Therefore, the most optimal size of the 1st convolutional filter is equal to the sample median of signal peak lengths, suggesting that 0.1 is the best time span of the 1st convolutional layer to retrieve the whole acceleration peaks and the best acceleration changes. Similarly, in frequency domain, the 1st convolutional layer kernel yielding the highest F1-score is the one with size 10, which is simply the sample median (\n
Effect of the size of 1st convolutional layer kernel on SMM recognition performance.
Furthermore, another way to show the efficiency of this adaptive 1st convolutional layer filter approach is to compare the performance of our time-domain CNN with the CNN of Rad et al. [30] which was trained on the same dataset as ours (in time-domain). Table 2 displays F1-score results of CNNs trained per subject and per study using the optimal 1st convolutional layer filter size (denoted as “Time-domain CNN”) and using the architecture of [30] (referred to as CNN-Rad). These results suggest that our time-domain CNN performs 20.17% higher in overall than the CNN of [30] and confirms the efficiency of the adaptive convolutional layer.
\n\n | Study 1 | \nStudy 2 | \nMean | \n|||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
\n | S1 | \nS2 | \nS3 | \nS4 | \nS5 | \nS6 | \nS1 | \nS2 | \nS3 | \nS4 | \nS5 | \n|
CNN-Rad [30] | \n71 | \n73 | \n70 | \n92 | \n68 | \n94 | \n68 | \n22 | \n2 | \n77 | \n75 | \n64.73 | \n
Time-domain CNN | \n91.23 | \n76.76 | \n84.95 | \n93.38 | \n86.41 | \n95.11 | \n95.97 | \n75.67 | \n60.17 | \n91.68 | \n82.55 | \n84.90 | \n
Comparative results (F1-scores) between the CNN using the adaptive 1st convolutional filter approach and the CNN of Rad et al. [30].
CNNs have so far yielded outstanding performance in several time series applications. However, this deep learning technique is a data driven approach i.e., a supervised machine learning algorithm that requires excessive amount of labeled (e.g., annotated) data for proper training and for a good convergence of parameters. Although in recent years several labeled datasets have become available, some fields such as medicine experience a lack of annotated data as manually annotating a large set requires human expertise and is time consuming. For instance, labeling acceleration signals of autistic children as SMMs or non-SMMs requires knowledge of a specialist. The conventional approach to deal with this kind of problem is to perform data augmentation by applying transformations to the existing data, as shown in Section 3.1.2. Data augmentation achieves slightly better time series classification rates but still the CNN is prone to overfitting. In this section, we present another solution to this problem, a “knowledge transfer” framework which is a global, fast and light-weight framework that combines the transfer learning technique with an SVM classifier. Afterwards, this technique is further implemented on another type of SMM recognition task, which consists of recognizing SMMs across different atypical subjects rather than recognizing SMMs across multiple sessions within one subject (as performed in experiments of Sections 3.2.2 and 4.2.2).
\nTransfer learning is a machine learning technique where a model trained on one task (a source domain) is re-purposed on a second related task (a target domain). Transfer learning is popular in deep learning, including Convolutional Neural Networks, given the enormous resources required to train deep learning models or the large and challenging datasets on which deep learning models are trained. For a CNN, given a source domain with a source learning task and a target domain with a target learning task (task of interest), transfer learning aims to improve learning of the target predictive function using the knowledge in the source domain which is the pre-trained CNN model containing features (parameters or weights) learned from the source domain task. This process works if these source domain features are general, meaningful and suitable to the target task. The pre-trained model can then be used as the starting point for a model on the target task. This may involve using all or parts of the pre-trained CNN model, depending on the modeling technique used. Accordingly, the questions that arise are: (i) which source learning task should be used for pre-training the CNN model given a target learning task, and (ii) which parts (e.g., learned features) of this model are common between the source and target learning tasks.
\nAn answer to the first question is to propose two source learning tasks. One source learning task is chosen to be very close and similar to the target learning task. And, if this source learning task lacks annotated data, then another source learning task is introduced which is chosen to be different but related to the target learning task.
\nA solution to the second problematic is to assume that features shared across the source and target tasks correspond to low- and mid-level information (e.g., fine details or local patterns) contained within inputs of both tasks, whereas the unshared features are the high-level information (e.g., global patterns) contained within inputs. And, knowing that training a CNNs produces learned low-, mid- and high-level features located at (contained within) the first, intermediate and last hidden layers respectively, we therefore assume that the features shared between the source and target tasks are contained within the first and intermediate CNN layers while features distinguishing one task from the other are contained within the last CNN layer. For instance, in image classification, as the CNN learns low-level features (Gabor filters such as edges, corners) through the first hidden layers, mid-level features (squares, circles, etc.) through intermediates hidden layers, and high-level features (faces, text, etc.) through last hidden layers, scene recognition (source learning task) and object recognition (target learning task) will have the same first and intermediate layers’ weights but different last layer weights. In time series, considering human activities where every activity is a combination of several basic continuous movements, with basic continuous movements corresponding to the smooth signals and the transitions or combinations among these movements causing the significant/salient changes of signal values, the purpose of the CNN will be to capture basic continuous movements through its low- and mid-level parameters (first and intermediate hidden layers) and the salience of the combination of basic movements through its high-level parameters (last hidden layers). Therefore, as an example, the CNN trained on recognizing basic human activities such as sitting, standing and walking (source learning task), and the one trained on recognizing SMMs (target learning task) will both have the same first and intermediate layer weights and different high layer weights. Another example is the CNN trained on SMMs of an atypical subject (source task) and the one trained on SMMs of another atypical subject (target task) which will have common first and intermediate hidden layers’ weights and different last hidden layer weights, due to the inter-subject variability across atypical subjects.
\nIn that sense, we propose a “Transfer learning with SVM read-out” framework which is composed of two parts: (i) the first part having first and intermediate layers’ weights of a CNN already pre-trained on a source learning task, (the last CNN layer being discarded), and (ii) the second part composed of a support vector machine (SVM) classifier with RBF kernel which is connected to the end of the first part. Then, we feed the entire training dataset of the target task into this framework in order to train the SVM parameters. As opposed to training a CNN on the target task which requires updating all hidden layers’ weights for several iterations using a large training set for all these weights to converge, our framework computes weights of the last layer(s) only, in one iteration only. Moreover the advantage of using SVM as the classifier is that it is fast and generally performs well on small training set since it only relies on the support vectors, which are the training samples that lay exactly on the hyperplanes used to define the margin. In addition, SVMs have the powerful RBF kernel, which allows to map the data to a very high dimension space in which the data can be separable by a hyperplane, hence guaranteeing convergence. Hence, our framework can be regarded as a global, fast and light-weight technique for time series classification where the target task has limited annotated/labeled data.
\nWe conduct this experiment again on the SMM recognition task. However, we will perform SMM recognition across multiple atypical subjects as opposed to SMM recognition within subjects which was developed in experiments of Sections 3.2. and 4.2. Indeed, due to the inter-subject variability of SMMs, a CNN trained on movements of an atypical subject \n
The target learning task will be the recognition of SMMs of subject \n
Datasets. The dataset used for the target learning task is the same SMM dataset used in Section 3.2.2. The dataset used for the source domain of the “TL SVM similar domains” experiment is also the SMM dataset, whereas the one used for the source domain of the “TL SVM across domains” experiment is the PUC dataset which is described in Section 3.2.2.
\nWhen using the SMM dataset in the target and source learning tasks, we do not take into consideration signals of all accelerometers/sensors (torso, right and left wrist) but rather signals of the torso sensor, resulting in input samples with 3 channels instead of 9. So, with torso measurements only, the only stereotypical movements that could be captured are the rock and flap-rock SMMs (and no flap SMMs). Accordingly, only rock and flap-rock SMM instances will be used as inputs in this experiment.
\nWhen using the PUC dataset for the source learning task, only the waist accelerometer (waist being next to torso) is taken into account since the other accelerometers (located at the thigh, ankle and arm) will not be relevant to the SMM recognition task during transfer learning. We consider the waist location to be equivalent to the torso location so that the CNN pre-trained on the source learning dataset can further be transferred to the target learning task (SMM recognition). Accordingly, input instances will have 3 channels instead of 12.
\nPre-processing. The pre-processing phase is the same as in Section 3.2.2.
\nExperimental setup. In experiments below, the architecture of the CNN model in time domain and frequency domain as well as training parameters are similar to the ones in Section 3.2.2. In addition, the target learning task consists of SMM recognition of a target subject \n
In order to perform SMM recognition on a target subject using transfer knowledge from SMMs of other subjects, the following steps are performed in time and frequency domains for each study \n
Step 1: we train a randomly initialized CNN in both time and frequency domains, for 5–15 epochs, using: (i) SMM instances of all 6 atypical subjects within study j except subject \n
Step 2: we reuse all layers of this CNN except the last layer (which is a fully connected layer) which is removed and replaced by the SVM classifier. The SVM of the transfer learning framework is trained using a small subset of subject \n
“TL SVM similar domains”. As depicted in Table 3, this framework (combining part of the pre-trained CNN with an SVM) is able to identify SMMs at a mean F1-score of 74.50 and 91.81% for time and frequency domains respectively. As opposed to the technique of directly applying the pre-trained CNN for classification which fails to recognize SMMs, “TL SVM similar domains” framework is able to capture relevant features for the recognition of SMMs across subjects. Thus, we can infer that low and mid-level SMM features share the same information from one subject to another and that “TL SVM similar domains” can be used as a global framework to identify SMMs of any new atypical subject. Furthermore, low- and mid-level features captured from a source learning task can be employed as low- and mid-level features of a target learning task close to the source task.
\nApproaches | \nStudy 1 | \nStudy 2 | \nMean | \n|||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
S1 | \nS2 | \nS3 | \nS4 | \nS5 | \nS6 | \nS1 | \nS2 | \nS3 | \nS4 | \nS5 | \n||
Time domain | \n||||||||||||
CNN few data | \n72.02 | \n62.31 | \n52.98 | \n88.47 | \n60.61 | \n88.86 | \n80.90 | \n53.28 | \n16.00 | \n82.66 | \n75.82 | \n66.72 | \n
TL full fine-tuning | \n75.73 | \n71.31 | \n59.04 | \n91.67 | \n59.47 | \n91.89 | \n85.66 | \n63.84 | \n38.57 | \n92.24 | \n82.31 | \n73.79 | \n
TL limited fine-tuning | \n75.44 | \n56.50 | \n50.86 | \n91.74 | \n63.86 | \n93.11 | \n85.88 | \n62.62 | \n27.14 | \n93.63 | \n81.16 | \n71.09 | \n
TL SVM similar domains | \n75.37 | \n76.44 | \n56.53 | \n91.74 | \n63.37 | \n92.76 | \n84.86 | \n62.97 | \n41.60 | \n93.32 | \n80.55 | \n74.50 | \n
TL SVM across domains | \n71.66 | \n74.40 | \n66.80 | \n90.69 | \n61.87 | \n92.19 | \n81.35 | \n58.13 | \n35.66 | \n88.44 | \n73.98 | \n72.29 | \n
Frequency domain | \n||||||||||||
CNN few data | \n76.64 | \n96.55 | \n63.44 | \n93.13 | \n82.58 | \n94.61 | \n84.94 | \n76.42 | \n29.51 | \n93.66 | \n83.41 | \n79.54 | \n
TL full fine-tuning | \n88.51 | \n97.22 | \n88.15 | \n97.53 | \n91.29 | \n98.26 | \n92.17 | \n88.19 | \n52.17 | \n96.59 | \n91.98 | \n89.28 | \n
TL limited fine-tuning | \n87.98 | \n94.59 | \n62.70 | \n97.57 | \n87.94 | \n98.36 | \n91.62 | \n87.08 | \n40.00 | \n97.82 | \n90.82 | \n85.14 | \n
TL SVM similar domain | \n90.54 | \n97.22 | \n83.86 | \n95.24 | \n86.19 | \n98.45 | \n92.71 | \n90.49 | \n84.99 | \n97.99 | \n92.22 | \n91.81 | \n
TL SVM across domains | \n74.50 | \n91.56 | \n43.77 | \n93.11 | \n76.03 | \n94.2 | \n85.16 | \n74.67 | \n66.98 | \n93.66 | \n83.99 | \n79.78 | \n
Results of CNN approaches used in this experiment per domain (time or frequency) per subject, per study. Highest rates are in bold.
“TL SVM across domains”. Training this framework produces satisfying results with a mean score of 72.29 and 79.78% in time and frequency domains respectively (Table 3). So, fixing low and mid-level features to features of basic movements and adjusting only the high-level features by an SVM seems to give satisfying classification results, which confirms that our framework has engaged feature detectors for finding stereotypical movements in signals. These results, especially the frequency-domains results, indicate that: (i) connecting low- and mid-level features of basic movements to an SVM classifier then feeding in 2000 instances for training the SVM generates a global framework which holds relevant and general representation that adapts to SMMs of any new atypical subject \n
Moreover, both our techniques are compared against the following methods:
The “CNN with few data” technique which consists of training a CNN in time and frequency domains with randomly initialized weights using the same target training data as the ones of “TL SVM similar domains” framework (i.e., 2000 SMM instances of the target subject i). The difference between this CNN and the CNN of Section 3.2.2 is that less data is used for training (2000 versus 10,000–30,000 training instances), only torso sensor measurements are applied in the former (compared to torso, right and left wrist sensor measurements in the latter), and only rock and flap-rock SMM instances are considered in the former (compared to rocking, flap-rock and flap SMM instances in the latter). We refer to this technique as “CNN few data”.
The “transfer learning with full fine-tuning” technique (referred to as “TL full fine-tuning”) consists of identifying SMMs of subject \n
The “transfer learning with limited fine-tuning” technique (denoted as “TL limited fine-tuning”) is the same as “transfer learning with full fine-tuning” except that the fine-tuning process is effective only on weights of the last CNN layer \n
Results and properties of the three techniques are depicted in Tables 3 and 4 respectively. From these results and properties, the following observations can be made:
“TL SVM similar domains” framework performs higher than the three frameworks “CNN few data”, “TL full fine-tuning” and “TL limited fine-tuning” in both time- and frequency-domain. This can be explained by the nature of the training process of the three frameworks, which relies on updating parameters using backpropagation. And, knowing that backpropagation requires abundant data for proper training, a lack of training data (2000 SMM instances) pushes the three frameworks to overfit and be less efficient.
“TL SVM similar domains” and “TL SVM across domains” architectures perform better than “CNN few data” by 7.78 and 5.57% respectively in time-domain and by 12.27 and 0.24% respectively in frequency-domain. Therefore, both architectures engage in capturing more general features than “CNN few data”. In terms of resources, one advantage of the two architectures over “CNN few data” is that the former converge much faster than the latter. Indeed, the former require 5–15 epochs (in both time and frequency domains) for full convergence while the latter needs 20–35 epochs and 55–85 epochs in time- and frequency-domain respectively for full convergence, as shown in Table 4. Another advantage resides in the number of parameters that have to be learned, which is 500 for the former (in both time and frequency domains) versus 1.2e + 06 and 7.1e + 05 for the latter in time and frequency domain respectively (Table 4). Hence, as opposed to “TL full fine-tuning” and “TL limited fine-tuning” frameworks, the “TL SVM” can be regarded as a global, fast and light-weight framework for SMM recognition across subjects.
“TL full fine-tuning” has a slightly higher performance than “TL limited fine-tuning” by 2.71 and 4.14% in time- and frequency-domain respectively, suggesting that fine-tuning weights of layers \n
“TL SVM across domains” framework yields a lower performance than “TL SVM similar domains” by 2.21% and 12.02% in time- and frequency-domain respectively. This implies the superiority in the SMM recognition task of low and mid-level features learned from SMMs over the ones learned from basic human movements. However, the latter features are more global. In time-domain, the small rate difference (2.21%) between “TL SVM across domains” and “TL SVM similar domains” suggests that the low- and mid-level feature space generated by human activities shares common details with the one generated by movements of specific atypical subjects. This is not the case for frequency-domain series, which can be explained by the difference in the frequency range between human activities and SMMs. Indeed, the FFT amplitude of human activities is contained below 10 Hz, pre-training the CNN on human activity frequency-signals from 0 to 3 Hz and not from 0 to 10 Hz results in imperfect human activity features which, combined with the SVM, do not seem to yield good classification results on the recognition of SMMs. If we were to have a new target learning task whose data signals are within the same frequency range as data signals of the source learning task, then “TL SVM across domains” would have achieved the same performance as “TL SVM similar domains”.
One advantage of “TL SVM similar domains” and “TL SVM across domains” is that they can be implemented in Android portable devices, as shown in Table 4. Indeed, an expert could receive continuous acceleration signals from the torso accelerometer of a subject, and label them on the fly (as SMM/non-SMM) as the subject performs his activities/movements. This results in annotated time series which are then preprocessed and fed into either “TL SVM similar domains” or “TL SVM across domains” for training. A one-minute recording of these signals is sufficient to train one of the two frameworks. Afterwards, this framework is ready to use for recognizing further SMMs on that same subject.
Approaches | \n# parameters updated for one pass (1 batch) | \n# batches per iteration | \n# iterations (epochs) | \nImplementation on Android device | \n
---|---|---|---|---|
CNN few data | \n1.2e + 06 (time-domain) 7.1e + 05 (frequency-domain) | \n14 (2000/150) | \n20–35 (time-domain) 55–85 (frequency-domain) | \nNo (too much memory consumption) | \n
TL full fine-tuning | \n1.2e + 06 (time-domain) 7.1e + 05 (frequency-domain) | \n14 (2000/150) | \n5–15 | \nNo (too much memory consumption) | \n
TL limited fine-tuning | \n1000 (500*2) | \n14 (2000/150) | \n5–15 | \nNo (hard to run back-propagation on mobile devices) | \n
TL SVM (similar domains and across domains) | \n500 | \n1 | \n1 | \nYes (easy to train SVM on mobile devices) | \n
Properties and resources used for the different techniques implemented in the experiment.
Time series pose important challenges to existing approaches which perform predictive modeling for classification tasks. In this paper we present a review on our previous works. Our contributions are aggregated into two categories: data-level and algorithm-level approaches. Our data-level approach consists of encoding time series using STin order to produce noise-free input signals which offer a more efficient CNN training. At the algorithm level, one approach is the adaptive convolutional layer filter approach which consists of determining the size of the filter based on an analysis of the input time series signals and fluctuations present within them. Indeed, choosing the proper 1st layer filter generates features maps which are more informative about the input signals and which capture the whole peaks within input signals. Furthermore, “TL SVM similar domains” and “TL SVM across domains” are algorithm-level approaches dealing with tasks with limited annotated data, which are regarded as two global, fast and light-weight techniques for these kinds of tasks. These two CNN approaches generate features general and global enough to recognize time series of the target learning task, given time series of a source learning task that is similar or different but related to the target learning task. All these approaches were implemented on the recognition of human activities, including normal activities performed by typical subjects and disorder-based activities performed by atypical subjects (such as SMMs of autistic subjects). Experimental results have showed the superiority of our techniques and their ability to extract relevant features from time series inputs. As a perspective, knowing that time series datasets often contain outliers either due to noisy time series or mislabeled time series (e.g. incorrect labels), we aim at studying a robust CNN that is insensitive to outliers. As opposed to our data-level CNN technique (mentioned in this paper) whose goal is to eliminate noise from time series, this robust CNN is an algorithm-level technique with acts at the level of loss functions by controlling high error values caused by outliers.
\nThe authors declare that they have no conflicts of interest.
\nYou have been successfully unsubscribed.
",metaTitle:"Unsubscribe Successful",metaDescription:"You have been successfully unsubscribed.",metaKeywords:null,canonicalURL:"/page/unsubscribe-successful",contentRaw:'[{"type":"htmlEditorComponent","content":""}]'},components:[{type:"htmlEditorComponent",content:""}]},successStories:{items:[]},authorsAndEditors:{filterParams:{sort:"featured,name"},profiles:[{id:"6700",title:"Dr.",name:"Abbass A.",middleName:null,surname:"Hashim",slug:"abbass-a.-hashim",fullName:"Abbass A. Hashim",position:null,profilePictureURL:"https://mts.intechopen.com/storage/users/6700/images/1864_n.jpg",biography:"Currently I am carrying out research in several areas of interest, mainly covering work on chemical and bio-sensors, semiconductor thin film device fabrication and characterisation.\nAt the moment I have very strong interest in radiation environmental pollution and bacteriology treatment. The teams of researchers are working very hard to bring novel results in this field. I am also a member of the team in charge for the supervision of Ph.D. students in the fields of development of silicon based planar waveguide sensor devices, study of inelastic electron tunnelling in planar tunnelling nanostructures for sensing applications and development of organotellurium(IV) compounds for semiconductor applications. I am a specialist in data analysis techniques and nanosurface structure. I have served as the editor for many books, been a member of the editorial board in science journals, have published many papers and hold many patents.",institutionString:null,institution:{name:"Sheffield Hallam University",country:{name:"United Kingdom"}}},{id:"54525",title:"Prof.",name:"Abdul Latif",middleName:null,surname:"Ahmad",slug:"abdul-latif-ahmad",fullName:"Abdul Latif Ahmad",position:null,profilePictureURL:"//cdnintech.com/web/frontend/www/assets/author.svg",biography:null,institutionString:null,institution:null},{id:"20567",title:"Prof.",name:"Ado",middleName:null,surname:"Jorio",slug:"ado-jorio",fullName:"Ado Jorio",position:null,profilePictureURL:"//cdnintech.com/web/frontend/www/assets/author.svg",biography:null,institutionString:null,institution:{name:"Universidade Federal de Minas Gerais",country:{name:"Brazil"}}},{id:"47940",title:"Dr.",name:"Alberto",middleName:null,surname:"Mantovani",slug:"alberto-mantovani",fullName:"Alberto Mantovani",position:null,profilePictureURL:"//cdnintech.com/web/frontend/www/assets/author.svg",biography:null,institutionString:null,institution:null},{id:"12392",title:"Mr.",name:"Alex",middleName:null,surname:"Lazinica",slug:"alex-lazinica",fullName:"Alex Lazinica",position:null,profilePictureURL:"https://mts.intechopen.com/storage/users/12392/images/7282_n.png",biography:"Alex Lazinica is the founder and CEO of IntechOpen. After obtaining a Master's degree in Mechanical Engineering, he continued his PhD studies in Robotics at the Vienna University of Technology. Here he worked as a robotic researcher with the university's Intelligent Manufacturing Systems Group as well as a guest researcher at various European universities, including the Swiss Federal Institute of Technology Lausanne (EPFL). During this time he published more than 20 scientific papers, gave presentations, served as a reviewer for major robotic journals and conferences and most importantly he co-founded and built the International Journal of Advanced Robotic Systems- world's first Open Access journal in the field of robotics. Starting this journal was a pivotal point in his career, since it was a pathway to founding IntechOpen - Open Access publisher focused on addressing academic researchers needs. Alex is a personification of IntechOpen key values being trusted, open and entrepreneurial. Today his focus is on defining the growth and development strategy for the company.",institutionString:null,institution:{name:"TU Wien",country:{name:"Austria"}}},{id:"19816",title:"Prof.",name:"Alexander",middleName:null,surname:"Kokorin",slug:"alexander-kokorin",fullName:"Alexander Kokorin",position:null,profilePictureURL:"https://mts.intechopen.com/storage/users/19816/images/1607_n.jpg",biography:"Alexander I. Kokorin: born: 1947, Moscow; DSc., PhD; Principal Research Fellow (Research Professor) of Department of Kinetics and Catalysis, N. Semenov Institute of Chemical Physics, Russian Academy of Sciences, Moscow.\r\nArea of research interests: physical chemistry of complex-organized molecular and nanosized systems, including polymer-metal complexes; the surface of doped oxide semiconductors. He is an expert in structural, absorptive, catalytic and photocatalytic properties, in structural organization and dynamic features of ionic liquids, in magnetic interactions between paramagnetic centers. The author or co-author of 3 books, over 200 articles and reviews in scientific journals and books. He is an actual member of the International EPR/ESR Society, European Society on Quantum Solar Energy Conversion, Moscow House of Scientists, of the Board of Moscow Physical Society.",institutionString:null,institution:{name:"Semenov Institute of Chemical Physics",country:{name:"Russia"}}},{id:"62389",title:"PhD.",name:"Ali Demir",middleName:null,surname:"Sezer",slug:"ali-demir-sezer",fullName:"Ali Demir Sezer",position:null,profilePictureURL:"https://mts.intechopen.com/storage/users/62389/images/3413_n.jpg",biography:"Dr. Ali Demir Sezer has a Ph.D. from Pharmaceutical Biotechnology at the Faculty of Pharmacy, University of Marmara (Turkey). He is the member of many Pharmaceutical Associations and acts as a reviewer of scientific journals and European projects under different research areas such as: drug delivery systems, nanotechnology and pharmaceutical biotechnology. Dr. Sezer is the author of many scientific publications in peer-reviewed journals and poster communications. Focus of his research activity is drug delivery, physico-chemical characterization and biological evaluation of biopolymers micro and nanoparticles as modified drug delivery system, and colloidal drug carriers (liposomes, nanoparticles etc.).",institutionString:null,institution:{name:"Marmara University",country:{name:"Turkey"}}},{id:"61051",title:"Prof.",name:"Andrea",middleName:null,surname:"Natale",slug:"andrea-natale",fullName:"Andrea Natale",position:null,profilePictureURL:"//cdnintech.com/web/frontend/www/assets/author.svg",biography:null,institutionString:null,institution:null},{id:"100762",title:"Prof.",name:"Andrea",middleName:null,surname:"Natale",slug:"andrea-natale",fullName:"Andrea Natale",position:null,profilePictureURL:"//cdnintech.com/web/frontend/www/assets/author.svg",biography:null,institutionString:null,institution:{name:"St David's Medical Center",country:{name:"United States of America"}}},{id:"107416",title:"Dr.",name:"Andrea",middleName:null,surname:"Natale",slug:"andrea-natale",fullName:"Andrea Natale",position:null,profilePictureURL:"//cdnintech.com/web/frontend/www/assets/author.svg",biography:null,institutionString:null,institution:{name:"Texas Cardiac Arrhythmia",country:{name:"United States of America"}}},{id:"64434",title:"Dr.",name:"Angkoon",middleName:null,surname:"Phinyomark",slug:"angkoon-phinyomark",fullName:"Angkoon Phinyomark",position:null,profilePictureURL:"https://mts.intechopen.com/storage/users/64434/images/2619_n.jpg",biography:"My name is Angkoon Phinyomark. I received a B.Eng. degree in Computer Engineering with First Class Honors in 2008 from Prince of Songkla University, Songkhla, Thailand, where I received a Ph.D. degree in Electrical Engineering. My research interests are primarily in the area of biomedical signal processing and classification notably EMG (electromyography signal), EOG (electrooculography signal), and EEG (electroencephalography signal), image analysis notably breast cancer analysis and optical coherence tomography, and rehabilitation engineering. I became a student member of IEEE in 2008. During October 2011-March 2012, I had worked at School of Computer Science and Electronic Engineering, University of Essex, Colchester, Essex, United Kingdom. In addition, during a B.Eng. I had been a visiting research student at Faculty of Computer Science, University of Murcia, Murcia, Spain for three months.\n\nI have published over 40 papers during 5 years in refereed journals, books, and conference proceedings in the areas of electro-physiological signals processing and classification, notably EMG and EOG signals, fractal analysis, wavelet analysis, texture analysis, feature extraction and machine learning algorithms, and assistive and rehabilitative devices. I have several computer programming language certificates, i.e. Sun Certified Programmer for the Java 2 Platform 1.4 (SCJP), Microsoft Certified Professional Developer, Web Developer (MCPD), Microsoft Certified Technology Specialist, .NET Framework 2.0 Web (MCTS). I am a Reviewer for several refereed journals and international conferences, such as IEEE Transactions on Biomedical Engineering, IEEE Transactions on Industrial Electronics, Optic Letters, Measurement Science Review, and also a member of the International Advisory Committee for 2012 IEEE Business Engineering and Industrial Applications and 2012 IEEE Symposium on Business, Engineering and Industrial Applications.",institutionString:null,institution:{name:"Joseph Fourier University",country:{name:"France"}}},{id:"55578",title:"Dr.",name:"Antonio",middleName:null,surname:"Jurado-Navas",slug:"antonio-jurado-navas",fullName:"Antonio Jurado-Navas",position:null,profilePictureURL:"https://mts.intechopen.com/storage/users/55578/images/4574_n.png",biography:"Antonio Jurado-Navas received the M.S. degree (2002) and the Ph.D. degree (2009) in Telecommunication Engineering, both from the University of Málaga (Spain). He first worked as a consultant at Vodafone-Spain. From 2004 to 2011, he was a Research Assistant with the Communications Engineering Department at the University of Málaga. In 2011, he became an Assistant Professor in the same department. From 2012 to 2015, he was with Ericsson Spain, where he was working on geo-location\ntools for third generation mobile networks. Since 2015, he is a Marie-Curie fellow at the Denmark Technical University. His current research interests include the areas of mobile communication systems and channel modeling in addition to atmospheric optical communications, adaptive optics and statistics",institutionString:null,institution:{name:"University of Malaga",country:{name:"Spain"}}}],filtersByRegion:[{group:"region",caption:"North America",value:1,count:5681},{group:"region",caption:"Middle and South America",value:2,count:5161},{group:"region",caption:"Africa",value:3,count:1683},{group:"region",caption:"Asia",value:4,count:10200},{group:"region",caption:"Australia and Oceania",value:5,count:886},{group:"region",caption:"Europe",value:6,count:15610}],offset:12,limit:12,total:117095},chapterEmbeded:{data:{}},editorApplication:{success:null,errors:{}},ofsBooks:{filterParams:{topicId:"14"},books:[{type:"book",id:"10216",title:"Paraffin - Thermal Energy Storage Applications",subtitle:null,isOpenForSubmission:!0,hash:"456090b63f5ba2290e24e655abd119bf",slug:null,bookSignature:"Dr. Elsayed Zaki and Dr. Abdelghaffar S. Dhmees",coverURL:"https://cdn.intechopen.com/books/images_new/10216.jpg",editedByType:null,editors:[{id:"220156",title:"Dr.",name:"Elsayed",surname:"Zaki",slug:"elsayed-zaki",fullName:"Elsayed Zaki"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"10411",title:"Nanocrystals",subtitle:null,isOpenForSubmission:!0,hash:"be29908600b7067c583ac21da1544a2d",slug:null,bookSignature:"Dr. Awadesh Mallik",coverURL:"https://cdn.intechopen.com/books/images_new/10411.jpg",editedByType:null,editors:[{id:"178218",title:"Dr.",name:"Awadesh",surname:"Mallik",slug:"awadesh-mallik",fullName:"Awadesh Mallik"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"10412",title:"Transition Metals",subtitle:null,isOpenForSubmission:!0,hash:"bd7287b801dc0ac77e01f66842dc1d99",slug:null,bookSignature:"Dr. Sajjad Haider and Dr. Adnan Haider",coverURL:"https://cdn.intechopen.com/books/images_new/10412.jpg",editedByType:null,editors:[{id:"110708",title:"Dr.",name:"Sajjad",surname:"Haider",slug:"sajjad-haider",fullName:"Sajjad Haider"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"10609",title:"Zeolites",subtitle:null,isOpenForSubmission:!0,hash:"90681a8fef45a03f68f4b9276acba2d3",slug:null,bookSignature:"Dr. Pavel Krivenko",coverURL:"https://cdn.intechopen.com/books/images_new/10609.jpg",editedByType:null,editors:[{id:"180922",title:"Dr.",name:"Pavel",surname:"Krivenko",slug:"pavel-krivenko",fullName:"Pavel Krivenko"}],productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"10847",title:"Aluminium Alloys",subtitle:null,isOpenForSubmission:!0,hash:"22f9ea5292994565616af16fac347ba0",slug:null,bookSignature:"",coverURL:"https://cdn.intechopen.com/books/images_new/10847.jpg",editedByType:null,editors:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"10848",title:"Tribology",subtitle:null,isOpenForSubmission:!0,hash:"316ca42e0526e3c775e16b929cc702f9",slug:null,bookSignature:"",coverURL:"https://cdn.intechopen.com/books/images_new/10848.jpg",editedByType:null,editors:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}}],filtersByTopic:[{group:"topic",caption:"Agricultural and Biological Sciences",value:5,count:9},{group:"topic",caption:"Biochemistry, Genetics and Molecular Biology",value:6,count:17},{group:"topic",caption:"Business, Management and Economics",value:7,count:2},{group:"topic",caption:"Chemistry",value:8,count:7},{group:"topic",caption:"Computer and Information Science",value:9,count:10},{group:"topic",caption:"Earth and Planetary Sciences",value:10,count:5},{group:"topic",caption:"Engineering",value:11,count:14},{group:"topic",caption:"Environmental Sciences",value:12,count:2},{group:"topic",caption:"Immunology and Microbiology",value:13,count:5},{group:"topic",caption:"Materials Science",value:14,count:4},{group:"topic",caption:"Mathematics",value:15,count:1},{group:"topic",caption:"Medicine",value:16,count:60},{group:"topic",caption:"Nanotechnology and Nanomaterials",value:17,count:1},{group:"topic",caption:"Neuroscience",value:18,count:1},{group:"topic",caption:"Pharmacology, Toxicology and Pharmaceutical Science",value:19,count:6},{group:"topic",caption:"Physics",value:20,count:2},{group:"topic",caption:"Psychology",value:21,count:3},{group:"topic",caption:"Robotics",value:22,count:1},{group:"topic",caption:"Social Sciences",value:23,count:3},{group:"topic",caption:"Technology",value:24,count:1},{group:"topic",caption:"Veterinary Medicine and Science",value:25,count:2}],offset:12,limit:12,total:6},popularBooks:{featuredBooks:[{type:"book",id:"9208",title:"Welding",subtitle:"Modern Topics",isOpenForSubmission:!1,hash:"7d6be076ccf3a3f8bd2ca52d86d4506b",slug:"welding-modern-topics",bookSignature:"Sadek Crisóstomo Absi Alfaro, Wojciech Borek and Błażej Tomiczek",coverURL:"https://cdn.intechopen.com/books/images_new/9208.jpg",editors:[{id:"65292",title:"Prof.",name:"Sadek Crisostomo Absi",middleName:"C. Absi",surname:"Alfaro",slug:"sadek-crisostomo-absi-alfaro",fullName:"Sadek Crisostomo Absi Alfaro"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9139",title:"Topics in Primary Care Medicine",subtitle:null,isOpenForSubmission:!1,hash:"ea774a4d4c1179da92a782e0ae9cde92",slug:"topics-in-primary-care-medicine",bookSignature:"Thomas F. Heston",coverURL:"https://cdn.intechopen.com/books/images_new/9139.jpg",editors:[{id:"217926",title:"Dr.",name:"Thomas F.",middleName:null,surname:"Heston",slug:"thomas-f.-heston",fullName:"Thomas F. Heston"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"8697",title:"Virtual Reality and Its Application in Education",subtitle:null,isOpenForSubmission:!1,hash:"ee01b5e387ba0062c6b0d1e9227bda05",slug:"virtual-reality-and-its-application-in-education",bookSignature:"Dragan Cvetković",coverURL:"https://cdn.intechopen.com/books/images_new/8697.jpg",editors:[{id:"101330",title:"Dr.",name:"Dragan",middleName:"Mladen",surname:"Cvetković",slug:"dragan-cvetkovic",fullName:"Dragan Cvetković"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9785",title:"Endometriosis",subtitle:null,isOpenForSubmission:!1,hash:"f457ca61f29cf7e8bc191732c50bb0ce",slug:"endometriosis",bookSignature:"Courtney Marsh",coverURL:"https://cdn.intechopen.com/books/images_new/9785.jpg",editors:[{id:"255491",title:"Dr.",name:"Courtney",middleName:null,surname:"Marsh",slug:"courtney-marsh",fullName:"Courtney Marsh"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9343",title:"Trace Metals in the Environment",subtitle:"New Approaches and Recent Advances",isOpenForSubmission:!1,hash:"ae07e345bc2ce1ebbda9f70c5cd12141",slug:"trace-metals-in-the-environment-new-approaches-and-recent-advances",bookSignature:"Mario Alfonso Murillo-Tovar, Hugo Saldarriaga-Noreña and Agnieszka Saeid",coverURL:"https://cdn.intechopen.com/books/images_new/9343.jpg",editors:[{id:"255959",title:"Dr.",name:"Mario Alfonso",middleName:null,surname:"Murillo-Tovar",slug:"mario-alfonso-murillo-tovar",fullName:"Mario Alfonso Murillo-Tovar"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"7831",title:"Sustainability in Urban Planning and Design",subtitle:null,isOpenForSubmission:!1,hash:"c924420492c8c2c9751e178d025f4066",slug:"sustainability-in-urban-planning-and-design",bookSignature:"Amjad Almusaed, Asaad Almssad and Linh Truong - Hong",coverURL:"https://cdn.intechopen.com/books/images_new/7831.jpg",editors:[{id:"110471",title:"Dr.",name:"Amjad",middleName:"Zaki",surname:"Almusaed",slug:"amjad-almusaed",fullName:"Amjad Almusaed"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"8468",title:"Sheep Farming",subtitle:"An Approach to Feed, Growth and Sanity",isOpenForSubmission:!1,hash:"838f08594850bc04aa14ec873ed1b96f",slug:"sheep-farming-an-approach-to-feed-growth-and-sanity",bookSignature:"António Monteiro",coverURL:"https://cdn.intechopen.com/books/images_new/8468.jpg",editors:[{id:"190314",title:"Prof.",name:"António",middleName:"Cardoso",surname:"Monteiro",slug:"antonio-monteiro",fullName:"António Monteiro"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"8816",title:"Financial Crises",subtitle:"A Selection of Readings",isOpenForSubmission:!1,hash:"6f2f49fb903656e4e54280c79fabd10c",slug:"financial-crises-a-selection-of-readings",bookSignature:"Stelios Markoulis",coverURL:"https://cdn.intechopen.com/books/images_new/8816.jpg",editors:[{id:"237863",title:"Dr.",name:"Stelios",middleName:null,surname:"Markoulis",slug:"stelios-markoulis",fullName:"Stelios Markoulis"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9376",title:"Contemporary Developments and Perspectives in International Health Security",subtitle:"Volume 1",isOpenForSubmission:!1,hash:"b9a00b84cd04aae458fb1d6c65795601",slug:"contemporary-developments-and-perspectives-in-international-health-security-volume-1",bookSignature:"Stanislaw P. Stawicki, Michael S. Firstenberg, Sagar C. Galwankar, Ricardo Izurieta and Thomas Papadimos",coverURL:"https://cdn.intechopen.com/books/images_new/9376.jpg",editors:[{id:"181694",title:"Dr.",name:"Stanislaw P.",middleName:null,surname:"Stawicki",slug:"stanislaw-p.-stawicki",fullName:"Stanislaw P. Stawicki"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"7769",title:"Medical Isotopes",subtitle:null,isOpenForSubmission:!1,hash:"f8d3c5a6c9a42398e56b4e82264753f7",slug:"medical-isotopes",bookSignature:"Syed Ali Raza Naqvi and Muhammad Babar Imrani",coverURL:"https://cdn.intechopen.com/books/images_new/7769.jpg",editors:[{id:"259190",title:"Dr.",name:"Syed Ali Raza",middleName:null,surname:"Naqvi",slug:"syed-ali-raza-naqvi",fullName:"Syed Ali Raza Naqvi"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9279",title:"Concepts, Applications and Emerging Opportunities in Industrial Engineering",subtitle:null,isOpenForSubmission:!1,hash:"9bfa87f9b627a5468b7c1e30b0eea07a",slug:"concepts-applications-and-emerging-opportunities-in-industrial-engineering",bookSignature:"Gary Moynihan",coverURL:"https://cdn.intechopen.com/books/images_new/9279.jpg",editors:[{id:"16974",title:"Dr.",name:"Gary",middleName:null,surname:"Moynihan",slug:"gary-moynihan",fullName:"Gary Moynihan"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"7807",title:"A Closer Look at Organizational Culture in Action",subtitle:null,isOpenForSubmission:!1,hash:"05c608b9271cc2bc711f4b28748b247b",slug:"a-closer-look-at-organizational-culture-in-action",bookSignature:"Süleyman Davut Göker",coverURL:"https://cdn.intechopen.com/books/images_new/7807.jpg",editors:[{id:"190035",title:"Associate Prof.",name:"Süleyman Davut",middleName:null,surname:"Göker",slug:"suleyman-davut-goker",fullName:"Süleyman Davut Göker"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}}],offset:12,limit:12,total:5126},hotBookTopics:{hotBooks:[],offset:0,limit:12,total:null},publish:{},publishingProposal:{success:null,errors:{}},books:{featuredBooks:[{type:"book",id:"9208",title:"Welding",subtitle:"Modern Topics",isOpenForSubmission:!1,hash:"7d6be076ccf3a3f8bd2ca52d86d4506b",slug:"welding-modern-topics",bookSignature:"Sadek Crisóstomo Absi Alfaro, Wojciech Borek and Błażej Tomiczek",coverURL:"https://cdn.intechopen.com/books/images_new/9208.jpg",editors:[{id:"65292",title:"Prof.",name:"Sadek Crisostomo Absi",middleName:"C. Absi",surname:"Alfaro",slug:"sadek-crisostomo-absi-alfaro",fullName:"Sadek Crisostomo Absi Alfaro"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9139",title:"Topics in Primary Care Medicine",subtitle:null,isOpenForSubmission:!1,hash:"ea774a4d4c1179da92a782e0ae9cde92",slug:"topics-in-primary-care-medicine",bookSignature:"Thomas F. Heston",coverURL:"https://cdn.intechopen.com/books/images_new/9139.jpg",editors:[{id:"217926",title:"Dr.",name:"Thomas F.",middleName:null,surname:"Heston",slug:"thomas-f.-heston",fullName:"Thomas F. Heston"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"8697",title:"Virtual Reality and Its Application in Education",subtitle:null,isOpenForSubmission:!1,hash:"ee01b5e387ba0062c6b0d1e9227bda05",slug:"virtual-reality-and-its-application-in-education",bookSignature:"Dragan Cvetković",coverURL:"https://cdn.intechopen.com/books/images_new/8697.jpg",editors:[{id:"101330",title:"Dr.",name:"Dragan",middleName:"Mladen",surname:"Cvetković",slug:"dragan-cvetkovic",fullName:"Dragan Cvetković"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9785",title:"Endometriosis",subtitle:null,isOpenForSubmission:!1,hash:"f457ca61f29cf7e8bc191732c50bb0ce",slug:"endometriosis",bookSignature:"Courtney Marsh",coverURL:"https://cdn.intechopen.com/books/images_new/9785.jpg",editors:[{id:"255491",title:"Dr.",name:"Courtney",middleName:null,surname:"Marsh",slug:"courtney-marsh",fullName:"Courtney Marsh"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9343",title:"Trace Metals in the Environment",subtitle:"New Approaches and Recent Advances",isOpenForSubmission:!1,hash:"ae07e345bc2ce1ebbda9f70c5cd12141",slug:"trace-metals-in-the-environment-new-approaches-and-recent-advances",bookSignature:"Mario Alfonso Murillo-Tovar, Hugo Saldarriaga-Noreña and Agnieszka Saeid",coverURL:"https://cdn.intechopen.com/books/images_new/9343.jpg",editors:[{id:"255959",title:"Dr.",name:"Mario Alfonso",middleName:null,surname:"Murillo-Tovar",slug:"mario-alfonso-murillo-tovar",fullName:"Mario Alfonso Murillo-Tovar"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"8468",title:"Sheep Farming",subtitle:"An Approach to Feed, Growth and Sanity",isOpenForSubmission:!1,hash:"838f08594850bc04aa14ec873ed1b96f",slug:"sheep-farming-an-approach-to-feed-growth-and-sanity",bookSignature:"António Monteiro",coverURL:"https://cdn.intechopen.com/books/images_new/8468.jpg",editors:[{id:"190314",title:"Prof.",name:"António",middleName:"Cardoso",surname:"Monteiro",slug:"antonio-monteiro",fullName:"António Monteiro"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"8816",title:"Financial Crises",subtitle:"A Selection of Readings",isOpenForSubmission:!1,hash:"6f2f49fb903656e4e54280c79fabd10c",slug:"financial-crises-a-selection-of-readings",bookSignature:"Stelios Markoulis",coverURL:"https://cdn.intechopen.com/books/images_new/8816.jpg",editors:[{id:"237863",title:"Dr.",name:"Stelios",middleName:null,surname:"Markoulis",slug:"stelios-markoulis",fullName:"Stelios Markoulis"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"7831",title:"Sustainability in Urban Planning and Design",subtitle:null,isOpenForSubmission:!1,hash:"c924420492c8c2c9751e178d025f4066",slug:"sustainability-in-urban-planning-and-design",bookSignature:"Amjad Almusaed, Asaad Almssad and Linh Truong - Hong",coverURL:"https://cdn.intechopen.com/books/images_new/7831.jpg",editors:[{id:"110471",title:"Dr.",name:"Amjad",middleName:"Zaki",surname:"Almusaed",slug:"amjad-almusaed",fullName:"Amjad Almusaed"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"9376",title:"Contemporary Developments and Perspectives in International Health Security",subtitle:"Volume 1",isOpenForSubmission:!1,hash:"b9a00b84cd04aae458fb1d6c65795601",slug:"contemporary-developments-and-perspectives-in-international-health-security-volume-1",bookSignature:"Stanislaw P. Stawicki, Michael S. Firstenberg, Sagar C. Galwankar, Ricardo Izurieta and Thomas Papadimos",coverURL:"https://cdn.intechopen.com/books/images_new/9376.jpg",editors:[{id:"181694",title:"Dr.",name:"Stanislaw P.",middleName:null,surname:"Stawicki",slug:"stanislaw-p.-stawicki",fullName:"Stanislaw P. Stawicki"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}},{type:"book",id:"7769",title:"Medical Isotopes",subtitle:null,isOpenForSubmission:!1,hash:"f8d3c5a6c9a42398e56b4e82264753f7",slug:"medical-isotopes",bookSignature:"Syed Ali Raza Naqvi and Muhammad Babar Imrani",coverURL:"https://cdn.intechopen.com/books/images_new/7769.jpg",editors:[{id:"259190",title:"Dr.",name:"Syed Ali Raza",middleName:null,surname:"Naqvi",slug:"syed-ali-raza-naqvi",fullName:"Syed Ali Raza Naqvi"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}}],latestBooks:[{type:"book",id:"8468",title:"Sheep Farming",subtitle:"An Approach to Feed, Growth and Sanity",isOpenForSubmission:!1,hash:"838f08594850bc04aa14ec873ed1b96f",slug:"sheep-farming-an-approach-to-feed-growth-and-sanity",bookSignature:"António Monteiro",coverURL:"https://cdn.intechopen.com/books/images_new/8468.jpg",editedByType:"Edited by",editors:[{id:"190314",title:"Prof.",name:"António",middleName:"Cardoso",surname:"Monteiro",slug:"antonio-monteiro",fullName:"António Monteiro"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9523",title:"Oral and Maxillofacial Surgery",subtitle:null,isOpenForSubmission:!1,hash:"5eb6ec2db961a6c8965d11180a58d5c1",slug:"oral-and-maxillofacial-surgery",bookSignature:"Gokul Sridharan",coverURL:"https://cdn.intechopen.com/books/images_new/9523.jpg",editedByType:"Edited by",editors:[{id:"82453",title:"Dr.",name:"Gokul",middleName:null,surname:"Sridharan",slug:"gokul-sridharan",fullName:"Gokul Sridharan"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9785",title:"Endometriosis",subtitle:null,isOpenForSubmission:!1,hash:"f457ca61f29cf7e8bc191732c50bb0ce",slug:"endometriosis",bookSignature:"Courtney Marsh",coverURL:"https://cdn.intechopen.com/books/images_new/9785.jpg",editedByType:"Edited by",editors:[{id:"255491",title:"Dr.",name:"Courtney",middleName:null,surname:"Marsh",slug:"courtney-marsh",fullName:"Courtney Marsh"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9018",title:"Some RNA Viruses",subtitle:null,isOpenForSubmission:!1,hash:"a5cae846dbe3692495fc4add2f60fd84",slug:"some-rna-viruses",bookSignature:"Yogendra Shah and Eltayb Abuelzein",coverURL:"https://cdn.intechopen.com/books/images_new/9018.jpg",editedByType:"Edited by",editors:[{id:"278914",title:"Ph.D.",name:"Yogendra",middleName:null,surname:"Shah",slug:"yogendra-shah",fullName:"Yogendra Shah"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"8816",title:"Financial Crises",subtitle:"A Selection of Readings",isOpenForSubmission:!1,hash:"6f2f49fb903656e4e54280c79fabd10c",slug:"financial-crises-a-selection-of-readings",bookSignature:"Stelios Markoulis",coverURL:"https://cdn.intechopen.com/books/images_new/8816.jpg",editedByType:"Edited by",editors:[{id:"237863",title:"Dr.",name:"Stelios",middleName:null,surname:"Markoulis",slug:"stelios-markoulis",fullName:"Stelios Markoulis"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9585",title:"Advances in Complex Valvular Disease",subtitle:null,isOpenForSubmission:!1,hash:"ef64f11e211621ecfe69c46e60e7ca3d",slug:"advances-in-complex-valvular-disease",bookSignature:"Michael S. Firstenberg and Imran Khan",coverURL:"https://cdn.intechopen.com/books/images_new/9585.jpg",editedByType:"Edited by",editors:[{id:"64343",title:null,name:"Michael S.",middleName:"S",surname:"Firstenberg",slug:"michael-s.-firstenberg",fullName:"Michael S. Firstenberg"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"10150",title:"Smart Manufacturing",subtitle:"When Artificial Intelligence Meets the Internet of Things",isOpenForSubmission:!1,hash:"87004a19de13702d042f8ff96d454698",slug:"smart-manufacturing-when-artificial-intelligence-meets-the-internet-of-things",bookSignature:"Tan Yen Kheng",coverURL:"https://cdn.intechopen.com/books/images_new/10150.jpg",editedByType:"Edited by",editors:[{id:"78857",title:"Dr.",name:"Tan Yen",middleName:null,surname:"Kheng",slug:"tan-yen-kheng",fullName:"Tan Yen Kheng"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9386",title:"Direct Numerical Simulations",subtitle:"An Introduction and Applications",isOpenForSubmission:!1,hash:"158a3a0fdba295d21ff23326f5a072d5",slug:"direct-numerical-simulations-an-introduction-and-applications",bookSignature:"Srinivasa Rao",coverURL:"https://cdn.intechopen.com/books/images_new/9386.jpg",editedByType:"Edited by",editors:[{id:"6897",title:"Dr.",name:"Srinivasa",middleName:"P",surname:"Rao",slug:"srinivasa-rao",fullName:"Srinivasa Rao"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9139",title:"Topics in Primary Care Medicine",subtitle:null,isOpenForSubmission:!1,hash:"ea774a4d4c1179da92a782e0ae9cde92",slug:"topics-in-primary-care-medicine",bookSignature:"Thomas F. Heston",coverURL:"https://cdn.intechopen.com/books/images_new/9139.jpg",editedByType:"Edited by",editors:[{id:"217926",title:"Dr.",name:"Thomas F.",middleName:null,surname:"Heston",slug:"thomas-f.-heston",fullName:"Thomas F. Heston"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9208",title:"Welding",subtitle:"Modern Topics",isOpenForSubmission:!1,hash:"7d6be076ccf3a3f8bd2ca52d86d4506b",slug:"welding-modern-topics",bookSignature:"Sadek Crisóstomo Absi Alfaro, Wojciech Borek and Błażej Tomiczek",coverURL:"https://cdn.intechopen.com/books/images_new/9208.jpg",editedByType:"Edited by",editors:[{id:"65292",title:"Prof.",name:"Sadek Crisostomo Absi",middleName:"C. Absi",surname:"Alfaro",slug:"sadek-crisostomo-absi-alfaro",fullName:"Sadek Crisostomo Absi Alfaro"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}}]},subject:{topic:{id:"158",title:"Metals and Nonmetals",slug:"metals-and-nonmetals",parent:{title:"Materials Science",slug:"materials-science"},numberOfBooks:112,numberOfAuthorsAndEditors:2695,numberOfWosCitations:2995,numberOfCrossrefCitations:1927,numberOfDimensionsCitations:4373,videoUrl:null,fallbackUrl:null,description:null},booksByTopicFilter:{topicSlug:"metals-and-nonmetals",sort:"-publishedDate",limit:12,offset:0},booksByTopicCollection:[{type:"book",id:"9343",title:"Trace Metals in the Environment",subtitle:"New Approaches and Recent Advances",isOpenForSubmission:!1,hash:"ae07e345bc2ce1ebbda9f70c5cd12141",slug:"trace-metals-in-the-environment-new-approaches-and-recent-advances",bookSignature:"Mario Alfonso Murillo-Tovar, Hugo Saldarriaga-Noreña and Agnieszka Saeid",coverURL:"https://cdn.intechopen.com/books/images_new/9343.jpg",editedByType:"Edited by",editors:[{id:"255959",title:"Dr.",name:"Mario Alfonso",middleName:null,surname:"Murillo-Tovar",slug:"mario-alfonso-murillo-tovar",fullName:"Mario Alfonso Murillo-Tovar"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"8787",title:"Bismuth",subtitle:"Fundamentals and Optoelectronic Applications",isOpenForSubmission:!1,hash:"7751170d0b538f61d14a27a56e6567a5",slug:"bismuth-fundamentals-and-optoelectronic-applications",bookSignature:"Yanhua Luo, Jianxiang Wen and Jianzhong Zhang",coverURL:"https://cdn.intechopen.com/books/images_new/8787.jpg",editedByType:"Edited by",editors:[{id:"226148",title:"Dr.",name:"Yanhua",middleName:null,surname:"Luo",slug:"yanhua-luo",fullName:"Yanhua Luo"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9949",title:"Lead Chemistry",subtitle:null,isOpenForSubmission:!1,hash:"b2f999b9583c748f957f612227976570",slug:"lead-chemistry",bookSignature:"Pipat Chooto",coverURL:"https://cdn.intechopen.com/books/images_new/9949.jpg",editedByType:"Edited by",editors:[{id:"197984",title:"Ph.D.",name:"Pipat",middleName:null,surname:"Chooto",slug:"pipat-chooto",fullName:"Pipat Chooto"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"9190",title:"Stability and Applications of Coordination Compounds",subtitle:null,isOpenForSubmission:!1,hash:"3f07c532e478beb8fcd2fe53b8c9bcfd",slug:"stability-and-applications-of-coordination-compounds",bookSignature:"Abhay Nanda Srivastva",coverURL:"https://cdn.intechopen.com/books/images_new/9190.jpg",editedByType:"Edited by",editors:[{id:"293623",title:"Dr.",name:"Abhay Nanda",middleName:"Nanda",surname:"Srivastva",slug:"abhay-nanda-srivastva",fullName:"Abhay Nanda Srivastva"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"7787",title:"Rare Earth Elements and Their Minerals",subtitle:null,isOpenForSubmission:!1,hash:"7ba4060b0830f7a68f00557da8ed8a39",slug:"rare-earth-elements-and-their-minerals",bookSignature:"Michael Aide and Takahito Nakajima",coverURL:"https://cdn.intechopen.com/books/images_new/7787.jpg",editedByType:"Edited by",editors:[{id:"185895",title:"Dr.",name:"Michael",middleName:"Thomas",surname:"Aide",slug:"michael-aide",fullName:"Michael Aide"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"7722",title:"Recent Advancements in the Metallurgical Engineering and Electrodeposition",subtitle:null,isOpenForSubmission:!1,hash:"0d7ff67bd6f4c13830658bc6f9a75851",slug:"recent-advancements-in-the-metallurgical-engineering-and-electrodeposition",bookSignature:"Uday Basheer Al-Naib, Dhanasekaran Vikraman and K. Karuppasamy",coverURL:"https://cdn.intechopen.com/books/images_new/7722.jpg",editedByType:"Edited by",editors:[{id:"182041",title:null,name:"Uday",middleName:"M.",surname:"Basheer Al-Naib",slug:"uday-basheer-al-naib",fullName:"Uday Basheer Al-Naib"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"7775",title:"Metallic Glasses",subtitle:null,isOpenForSubmission:!1,hash:"665fb007e1e410d119fc09d709c41cc3",slug:"metallic-glasses",bookSignature:"Dragica Minić and Milica Vasić",coverURL:"https://cdn.intechopen.com/books/images_new/7775.jpg",editedByType:"Edited by",editors:[{id:"30470",title:"Prof.",name:"Dragica",middleName:"M",surname:"Minić",slug:"dragica-minic",fullName:"Dragica Minić"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"8653",title:"Electromagnetic Materials and Devices",subtitle:null,isOpenForSubmission:!1,hash:"0cc0489a203ae888b1105719a4e70ecd",slug:"electromagnetic-materials-and-devices",bookSignature:"Man-Gui Han",coverURL:"https://cdn.intechopen.com/books/images_new/8653.jpg",editedByType:"Edited by",editors:[{id:"250649",title:"Prof.",name:"Man-Gui",middleName:null,surname:"Han",slug:"man-gui-han",fullName:"Man-Gui Han"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"8886",title:"Cobalt Compounds and Applications",subtitle:null,isOpenForSubmission:!1,hash:"0241f740fc6e17cd9dc69362ef388d04",slug:"cobalt-compounds-and-applications",bookSignature:"Yasemin Yıldız and Aynur Manzak",coverURL:"https://cdn.intechopen.com/books/images_new/8886.jpg",editedByType:"Edited by",editors:[{id:"208129",title:"Dr.",name:"Yasemin",middleName:null,surname:"Yıldız",slug:"yasemin-yildiz",fullName:"Yasemin Yıldız"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"8416",title:"Non-Equilibrium Particle Dynamics",subtitle:null,isOpenForSubmission:!1,hash:"2c3add7639dcd1cb442cb4313ea64e3a",slug:"non-equilibrium-particle-dynamics",bookSignature:"Albert S. Kim",coverURL:"https://cdn.intechopen.com/books/images_new/8416.jpg",editedByType:"Edited by",editors:[{id:"21045",title:"Prof.",name:"Albert S.",middleName:null,surname:"Kim",slug:"albert-s.-kim",fullName:"Albert S. Kim"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"8408",title:"Titanium Alloys",subtitle:"Novel Aspects of Their Manufacturing and Processing",isOpenForSubmission:!1,hash:"e5533136b732dc4ada818553023d4d55",slug:"titanium-alloys-novel-aspects-of-their-manufacturing-and-processing",bookSignature:"Maciej Motyka, Waldemar Ziaja and Jan Sieniawsk",coverURL:"https://cdn.intechopen.com/books/images_new/8408.jpg",editedByType:"Edited by",editors:[{id:"101690",title:"Associate Prof.",name:"Maciej",middleName:null,surname:"Motyka",slug:"maciej-motyka",fullName:"Maciej Motyka"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}},{type:"book",id:"6850",title:"Crystal Growth",subtitle:null,isOpenForSubmission:!1,hash:"6dfc582440368fdfd9e72511c1a23f94",slug:"crystal-growth",bookSignature:"Vadim Glebovsky",coverURL:"https://cdn.intechopen.com/books/images_new/6850.jpg",editedByType:"Edited by",editors:[{id:"101945",title:"Dr.",name:"Vadim",middleName:null,surname:"Glebovsky",slug:"vadim-glebovsky",fullName:"Vadim Glebovsky"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter",authoredCaption:"Edited by"}}],booksByTopicTotal:112,mostCitedChapters:[{id:"37067",doi:"10.5772/35482",title:"Fourier Transform Infrared Spectroscopy for Natural Fibres",slug:"fourier-transform-infrared-spectroscopy-for-natural-fibres",totalDownloads:8205,totalCrossrefCites:113,totalDimensionsCites:275,book:{slug:"fourier-transform-materials-analysis",title:"Fourier Transform",fullTitle:"Fourier Transform - Materials Analysis"},signatures:"Mizi Fan, Dasong Dai and Biao Huang",authors:[{id:"104647",title:"Prof.",name:"Mizi",middleName:null,surname:"Fan",slug:"mizi-fan",fullName:"Mizi Fan"}]},{id:"46243",doi:"10.5772/57255",title:"Corrosion Inhibitors – Principles, Mechanisms and Applications",slug:"corrosion-inhibitors-principles-mechanisms-and-applications",totalDownloads:12941,totalCrossrefCites:29,totalDimensionsCites:100,book:{slug:"developments-in-corrosion-protection",title:"Developments in Corrosion Protection",fullTitle:"Developments in Corrosion Protection"},signatures:"Camila G. Dariva and Alexandre F. Galio",authors:[{id:"169261",title:"Dr.",name:"Camila",middleName:"G.",surname:"Dariva",slug:"camila-dariva",fullName:"Camila Dariva"},{id:"170138",title:"Dr.",name:"Alexandre",middleName:"Ferreira",surname:"Galio",slug:"alexandre-galio",fullName:"Alexandre Galio"}]},{id:"60680",doi:"10.5772/intechopen.76082",title:"Environmental Contamination by Heavy Metals",slug:"environmental-contamination-by-heavy-metals",totalDownloads:11768,totalCrossrefCites:54,totalDimensionsCites:100,book:{slug:"heavy-metals",title:"Heavy Metals",fullTitle:"Heavy Metals"},signatures:"Vhahangwele Masindi and Khathutshelo L. Muedi",authors:[{id:"225304",title:"Dr.",name:"Vhahangwele",middleName:null,surname:"Masindi",slug:"vhahangwele-masindi",fullName:"Vhahangwele Masindi"},{id:"241403",title:"M.Sc.",name:"Khathutshelo",middleName:"Lilith",surname:"Muedi",slug:"khathutshelo-muedi",fullName:"Khathutshelo Muedi"}]}],mostDownloadedChaptersLast30Days:[{id:"60680",title:"Environmental Contamination by Heavy Metals",slug:"environmental-contamination-by-heavy-metals",totalDownloads:11768,totalCrossrefCites:54,totalDimensionsCites:100,book:{slug:"heavy-metals",title:"Heavy Metals",fullTitle:"Heavy Metals"},signatures:"Vhahangwele Masindi and Khathutshelo L. Muedi",authors:[{id:"225304",title:"Dr.",name:"Vhahangwele",middleName:null,surname:"Masindi",slug:"vhahangwele-masindi",fullName:"Vhahangwele Masindi"},{id:"241403",title:"M.Sc.",name:"Khathutshelo",middleName:"Lilith",surname:"Muedi",slug:"khathutshelo-muedi",fullName:"Khathutshelo Muedi"}]},{id:"59857",title:"Introductory Chapter: Introducing Heavy Metals",slug:"introductory-chapter-introducing-heavy-metals",totalDownloads:4140,totalCrossrefCites:3,totalDimensionsCites:8,book:{slug:"heavy-metals",title:"Heavy Metals",fullTitle:"Heavy Metals"},signatures:"Martin Koller and Hosam M. Saleh",authors:[{id:"144691",title:"Prof.",name:"Hosam",middleName:"M.",surname:"Saleh",slug:"hosam-saleh",fullName:"Hosam Saleh"}]},{id:"60518",title:"Synthetic Methods for Titanium Dioxide Nanoparticles: A Review",slug:"synthetic-methods-for-titanium-dioxide-nanoparticles-a-review",totalDownloads:3104,totalCrossrefCites:9,totalDimensionsCites:15,book:{slug:"titanium-dioxide-material-for-a-sustainable-environment",title:"Titanium Dioxide",fullTitle:"Titanium Dioxide - Material for a Sustainable Environment"},signatures:"Pardon Nyamukamba, Omobola Okoh, Henry Mungondori,\nRaymond Taziwa and Simcelile Zinya",authors:[{id:"196100",title:"Dr.",name:"Raymond",middleName:null,surname:"Taziwa",slug:"raymond-taziwa",fullName:"Raymond Taziwa"},{id:"219920",title:"Prof.",name:"Omobola",middleName:null,surname:"Okoh",slug:"omobola-okoh",fullName:"Omobola Okoh"},{id:"226567",title:"Dr.",name:"Pardon",middleName:null,surname:"Nyamukamba",slug:"pardon-nyamukamba",fullName:"Pardon Nyamukamba"},{id:"239758",title:"Mr.",name:"Simcelile",middleName:null,surname:"Zinya",slug:"simcelile-zinya",fullName:"Simcelile Zinya"}]},{id:"59905",title:"Synthesis of Silver Nanoparticles",slug:"synthesis-of-silver-nanoparticles",totalDownloads:4855,totalCrossrefCites:1,totalDimensionsCites:5,book:{slug:"silver-nanoparticles-fabrication-characterization-and-applications",title:"Silver Nanoparticles",fullTitle:"Silver Nanoparticles - Fabrication, Characterization and Applications"},signatures:"Remziye Güzel and Gülbahar Erdal",authors:[{id:"226613",title:"Dr.",name:"Remziye",middleName:null,surname:"Güzel",slug:"remziye-guzel",fullName:"Remziye Güzel"},{id:"240772",title:"MSc.",name:"Gülbahar",middleName:null,surname:"Erdal",slug:"gulbahar-erdal",fullName:"Gülbahar Erdal"}]},{id:"58868",title:"Iron Ore Pelletizing Process: An Overview",slug:"iron-ore-pelletizing-process-an-overview",totalDownloads:3036,totalCrossrefCites:2,totalDimensionsCites:4,book:{slug:"iron-ores-and-iron-oxide-materials",title:"Iron Ores and Iron Oxide Materials",fullTitle:"Iron Ores and Iron Oxide Materials"},signatures:"Sandra Lúcia de Moraes, José Renato Baptista de Lima and Tiago\nRamos Ribeiro",authors:[{id:"216788",title:"Dr.",name:"Sandra",middleName:"Lúcia",surname:"De Moraes",slug:"sandra-de-moraes",fullName:"Sandra De Moraes"},{id:"233466",title:"Prof.",name:"José Renato Baptista",middleName:null,surname:"De Lima",slug:"jose-renato-baptista-de-lima",fullName:"José Renato Baptista De Lima"},{id:"233467",title:"MSc.",name:"Tiago Ramos",middleName:null,surname:"Ribeiro",slug:"tiago-ramos-ribeiro",fullName:"Tiago Ramos Ribeiro"}]},{id:"51497",title:"The Review of Some Commonly Used Methods and Techniques to Measure the Thermal Conductivity of Insulation Materials",slug:"the-review-of-some-commonly-used-methods-and-techniques-to-measure-the-thermal-conductivity-of-insul",totalDownloads:4057,totalCrossrefCites:13,totalDimensionsCites:27,book:{slug:"insulation-materials-in-context-of-sustainability",title:"Insulation Materials in Context of Sustainability",fullTitle:"Insulation Materials in Context of Sustainability"},signatures:"Numan Yüksel",authors:[{id:"178245",title:"Dr.",name:"Numan",middleName:null,surname:"Yüksel",slug:"numan-yuksel",fullName:"Numan Yüksel"}]},{id:"47427",title:"Corrosion and Surface Treatment of Magnesium Alloys",slug:"corrosion-and-surface-treatment-of-magnesium-alloys",totalDownloads:3366,totalCrossrefCites:10,totalDimensionsCites:23,book:{slug:"magnesium-alloys-properties-in-solid-and-liquid-states",title:"Magnesium Alloys",fullTitle:"Magnesium Alloys - Properties in Solid and Liquid States"},signatures:"Henry Hu, Xueyuan Nie and Yueyu Ma",authors:[{id:"170745",title:"Prof.",name:"Henry",middleName:null,surname:"Hu",slug:"henry-hu",fullName:"Henry Hu"}]},{id:"43029",title:"Chemical Vapor Transport Reactions–Methods, Materials, Modeling",slug:"chemical-vapor-transport-reactions-methods-materials-modeling",totalDownloads:8826,totalCrossrefCites:15,totalDimensionsCites:28,book:{slug:"advanced-topics-on-crystal-growth",title:"Advanced Topics on Crystal Growth",fullTitle:"Advanced Topics on Crystal Growth"},signatures:"Peer Schmidt, Michael Binnewies, Robert Glaum and Marcus Schmidt",authors:[{id:"162521",title:"Prof.",name:"Peer",middleName:null,surname:"Schmidt",slug:"peer-schmidt",fullName:"Peer Schmidt"}]},{id:"57525",title:"Investigation of Martensitic Transformation Induced by Cyclic Plastic Deformation in Austenitic Steels",slug:"investigation-of-martensitic-transformation-induced-by-cyclic-plastic-deformation-in-austenitic-stee",totalDownloads:740,totalCrossrefCites:0,totalDimensionsCites:0,book:{slug:"austenitic-stainless-steels-new-aspects",title:"Austenitic Stainless Steels",fullTitle:"Austenitic Stainless Steels - New Aspects"},signatures:"Jerzy Kaleta, Przemysław Wiewiórski and Wojciech Wiśniewski",authors:[{id:"19069",title:"Prof.",name:"Jerzy",middleName:null,surname:"Kaleta",slug:"jerzy-kaleta",fullName:"Jerzy Kaleta"},{id:"220028",title:"MSc.",name:"Przemysław",middleName:null,surname:"Wiewiórski",slug:"przemyslaw-wiewiorski",fullName:"Przemysław Wiewiórski"},{id:"220029",title:"MSc.",name:"Wojciech",middleName:null,surname:"Wiśniewski",slug:"wojciech-wisniewski",fullName:"Wojciech Wiśniewski"}]},{id:"70192",title:"Stability Constants of Metal Complexes in Solution",slug:"stability-constants-of-metal-complexes-in-solution",totalDownloads:1226,totalCrossrefCites:0,totalDimensionsCites:0,book:{slug:"stability-and-applications-of-coordination-compounds",title:"Stability and Applications of Coordination Compounds",fullTitle:"Stability and Applications of Coordination Compounds"},signatures:"Jagvir Singh, Abhay Nanda Srivastav, Netrapal Singh and Anuradha Singh",authors:null}],onlineFirstChaptersFilter:{topicSlug:"metals-and-nonmetals",limit:3,offset:0},onlineFirstChaptersCollection:[],onlineFirstChaptersTotal:0},preDownload:{success:null,errors:{}},aboutIntechopen:{},privacyPolicy:{},peerReviewing:{},howOpenAccessPublishingWithIntechopenWorks:{},sponsorshipBooks:{sponsorshipBooks:[{type:"book",id:"10176",title:"Microgrids and Local Energy Systems",subtitle:null,isOpenForSubmission:!0,hash:"c32b4a5351a88f263074b0d0ca813a9c",slug:null,bookSignature:"Prof. Nick Jenkins",coverURL:"https://cdn.intechopen.com/books/images_new/10176.jpg",editedByType:null,editors:[{id:"55219",title:"Prof.",name:"Nick",middleName:null,surname:"Jenkins",slug:"nick-jenkins",fullName:"Nick Jenkins"}],equalEditorOne:null,equalEditorTwo:null,equalEditorThree:null,productType:{id:"1",chapterContentType:"chapter"}}],offset:8,limit:8,total:1},route:{name:"profile.detail",path:"/profiles/67291/srinivasan-prabakar",hash:"",query:{},params:{id:"67291",slug:"srinivasan-prabakar"},fullPath:"/profiles/67291/srinivasan-prabakar",meta:{},from:{name:null,path:"/",hash:"",query:{},params:{},fullPath:"/",meta:{}}}},function(){var e;(e=document.currentScript||document.scripts[document.scripts.length-1]).parentNode.removeChild(e)}()