Perspective Chapter: Optoelectronics for Neural Interfaces

Abstract

Optoelectronics for neural interfaces is a growing field developing light-based methods for recording and stimulating neural activity. It has the potential to revolutionize the treatment of neurological disorders. The chapter will delve into optoelectronics’ basic principles, its applications, and various devices such as implantable optical fibers, microelectrode arrays, and integration with flexible materials. The chapter will highlight the challenges and opportunities facing the field, such as developing small, flexible, and biocompatible devices, controlling light delivery, understanding optogenetic stimulation effects and their scalable integration to achieve high spatiotemporal precision and low invasiveness. Despite challenges, optoelectronics for neural interfaces is a promising approach that could open up new avenues to restore vision to the blind, control prosthetic limbs, and treat diseases like epilepsy.

1. Introduction

Human nervous system is made up of billions of neurons that communicate with each other through multiple synapses and therefore create vast web of neural networks, making the brain an exceptionally powerful system. Understanding correlation between animal behavior and these neuronal pathways is one of the primary aim of neuroscience. Another research domain called Brain Computer Interface (BCI), aims to form a direct hardware-based communication link between the brain and the external world. As input, these systems convert the electrical activity from neuron and translate it to control signals, which are used to connect with an external assistive device. BCI systems range from implantable electrocardiography (ECoG) to neuroprosthetics that can rehabilitate vision and hearing. These systems can be utilized to improve our understanding of organization and operation of nervous system and may lead to advancing the current state-of- the-art technologies for treating severe neurological disorders such as paralysis, Alzheimer’s disease etc. To achieve this, development of multimodal systems, with capability to record and regulate activities of target neurons is required.

Optoelectronic devices have become ubiquitous in our daily lives, seamlessly integrated into consumer electronics, communication systems, automobiles, and medical technologies. Their remarkable versatility and ability to bridge the gap between light and electricity have opened a world of possibilities, revolutionizing various aspects of modern life. Optogenetics combines genetic and optical techniques to control neural activities with light. It is a promising tool for selectively targeting nerve cells that allow extensive studies of cell activity which can eventually be extended to neural network connectivities. In more detail, light sensitive ion channels called opsins are injected into target tissue. These opsins are extracted from algae or bacteria and light activation can be of excitatory (e.g. channelrhodopsin i.e. ChR2) or inhibitory (e.g. halorhodopsin i.e. NpHR) nature [1, 2, 3]. Therefore, in contrast to traditional electrical stimulation where the surrounding electrode tissue is also affected by the stimulation, optogenetics permits regulation of neural activity with both cell-type specificity and high temporal resolution.

High spatiotemporal resolution is essential to track neuronal activity in order to understand how information is regulated in neural networks. MEAs can be used to detect extracellular action potentials (APs) as well as transient electrical signals generated from summation of excitatory and inhibitory potentials from a cluster of neurons called as local field potentials (LFPs). Synchronous recording and modulation of specific neuronal population in target tissue can be achieved with integration of electrical recording and optogenetic capabilities. This is important to establish their role in information regulation process and therefore gain further insights into animal behavior [4, 5].

Research in the last decade has indicate that integrating electrophysiology and optogenetics capabilities has created a powerful tool to enhance our understanding of neural activity, behavioral outputs as well as sources of defects in nervous system that lead to neurological disorders [6, 7, 8]. However, substantial challenges exist in current combined optogenetics & electrophysiology techniques which need to be addressed. This chapter delves into advances in implantable neural probes along with their material and design requirements. It further provides an overview of recent developments in multimodal neural probes for simultaneous optoelectronic modulation and recording via two distinct approaches i.e. using active and passive optical probes. Finally, challenges faced and potential opportunities of further developing these multimodal neural interfaces to potentially use for clinical applications is discussed.

2. Neural Interface technology

2.1 Advancement of microelectrode arrays (MEAs)

Neural prosthetics are BCI devices intended to revive the lost functionality of motor or sensory organs in patients suffering from neural injuries and disorders. Advancement of fabrication of neural prosthetics from single channel devices to flexible Microelectrode arrays (MEAs) consisting of few hundred to thousands of channels (thus improving spatial resolution of input electrical signal as well as the signal-to-noise ratio) has made the clinical applications of these devices more viable. These arrays can be utilized for recording neuronal activity to aid in gaining further insights into functioning of specific areas of the brain, that may eventually lead to diagnosis of neurodegenerative diseases. Another application of these arrays involves electrical stimulation of the brain by injecting electrical impulses with an aim to restore degenerating or impaired motor systems by activating damaged neurons and creating an alternate neural pathway. Additionally, these arrays can also be used for detection of neurotransmitters, which has been an effective approach to gain a better understanding of neurological disorders such as Alzheimer’s disease, Parkinson’s disease as these are characteristically associated with imbalance. Another interesting area of research using these arrays is the field of drug delivery which has been plausible by addition of microfluidic channels [9].

Microelectrode arrays (MEAs) have been employed for neural recording and stimulation application with different degrees of precision and invasiveness [10]. For example, Electroencephalography (EEG) electrodes which is non-invasive as the electrodes are implanted on the surface of the brain and thus offer low-resolution data of smoothed field potentials representative of neural activity of the whole cortical surface. Electrocorticography (ECoG) electrodes which is semi-invasive as the electrodes may be positioned outside the dura mater (epidural) or under the dura mater (subdural) and it requires craniotomy for implantation of the electrodes and multiple layers of cortical laminae can be targeted with different designs (e.g. shank type) of electrodes. These provide higher temporal and spatial resolution and are routinely used to identify seizure loci in epilepsy patients [11].

Accurate mapping of neural activity is of significant clinical importance extending beyond the cortex to encompass deep brain regions (such as the subthalamic nucleus in individuals with Parkinson’s disease), the spinal cord, and peripheral nerves (particularly in cases of trauma or chronic pain). Furthermore, many nervous system related disorders are associated with atypical activity in specific neuronal classes, justifying the critical need to achieve single-neuron resolution for development of effective therapeutic strategies.

2.2 Material and design requirements for MEAs

Essential requirements for a prospective material for neural probes are as follows:

The material must be biocompatible in order to promote seamless integration with neurons and therefore minimize foreign-body response by causing minimal inflammation and neuronal cell loss.

• The material must be exhibit adequate mechanical strength to withstand compression and tension forces during insertion and retraction stages of implantation procedure. Additionally, it must decrease the mechanical mismatch between the sensing electrode material and the surrounding tissue for suppressing chronic tissue encapsulation.

• The material must be able to inject sufficient charge in the surrounding tissue to cause electrical brain stimulation at clinically relevant charge densities without causing any significant tissue damage due to corrosion or delamination of insulating or substrate layers of the implant.

• During stimulation operation, irreversible reaction should not occur in the brain i.e. the material should have the ability to provide safe levels of therapeutic stimulation.

• The material should be able to withstand implantation as well as exhibit stable electrochemical characteristics for long-term.

Materials for neural probes can be categorized based on their mechanism of charge injection during stimulation i.e. capacitive or faradaic. Capacitive charge injection involves charging and discharging of the electrode-electrolyte (in-vitro testing) or electrode-tissue (in-vivo testing) double layer while the faradic charge injection involves transfer of electrons resulting in oxidation or reduction of surface bound species. Examples of some capacitive charge-injection materials employed for the purpose of neural stimulation are Titanium nitride, Tantalum/Tantalum oxide while some of the Faradaic materials include noble metals like Platinum and Platinum-iridium alloys, Iridium-oxide, PEDOT. Some of the common materials used for recording electrodes are stainless-steel, tungsten, PEDOT, platinum, platinum-iridium alloys [9].

3. Integration of optoelectronics

As the field of optogenetics has progressed, it has become possible to selectively excite or inhibit specific neurons with millisecond precision [2, 12]. This is realized by genetically targeting light-sensitive proteins called ‘opsins’ (extracted from algae or bacteria) to demonstrate neuronal response (sensitivity) to various wavelength in visible spectrum of light. Opsins can play excitatory (e.g. channelrhodopsin i.e. ChR2) or inhibitory role (e.g. halorhodopsin i.e. NpHR) based on application of interest. Optogenetics offers a distinct advantage over the electrical stimulation by enabling the precise activation of specific target cell population without impacting the surrounding tissue [13].

A major challenge for these light-based techniques is the highly scattering and absorptive behavior of mammalian tissue in the visible light range, especially in deep neural tissue. Therefore, inclusion of optical waveguides or light emitting device integrated at site of intervention is essential for efficient light delivery and collection within deep tissue. Materials, design and biocompatibility considerations must be addressed for these waveguides similar to considerations for neural recording and stimulation electrodes.

Factors such as proposed implantation site and therefore volume and density of neuron population to be optically stimulated, dictates the required size and illumination area for optical part of probe. Based on the mechanism of light source, probes can be categorized as ‘Passive optical neural probes’ which route light into tissue from light sources located outside the body while ‘Active optical neural probes’ where light source directly assembled on source. Using passive probes minimizes the risk of thermal damage to the tissue as the light source is located outside the body while active implants generate light directly inside tissue by converting electricity to light.

3.1 Passive optical neural probes

Essential requirement of optical waveguide for passive optical neural probes is a high refractive index core engulfed by a lower refractive index cladding layer to support optimal transmission of light to the desired neuronal population in target area. Materials exhibiting this behavior in addition to flexibility and biocompatibility, as described in Section 2.2, are needed to form a flexible optical waveguide to integrate into neural implants. As silicon micromachining has the longest history of producing successful neural probes and allows easy integration of electronic circuity, it was a go-to choice as a substrate material in many designs [14, 15]. Other examples include use of silicon dioxide, silicon nitride and zinc oxide as waveguide materials [16].

Alternatively, polymer materials like SU8, Parylene and ORMOCER [17, 18, 19, 20, 21, 22] have been used but transparency in most cases is not as high as silicon-based materials. Using silicon as a substrate is a well-established practice, but it’s known for its rigidity, which could potentially result in elevated foreign body responses when compared to more pliable materials. As an alternative substrate, polymers have been employed, featuring integrated waveguides and the possibility of fluidic channels for precise vector injection during optical stimulation and electrical recording at the desired location.

Polymer probes are preferred due to their flexibility and reduction in mechanical mismatch to the tissue i.e. polymers are less stiff than silicon but their flexibility i.e. Young’s modulus is still some orders of magnitude larger than nervous tissue. Additionally, larger attenuation of light is observed in polymers as compared to waveguides based on dielectrics. Parylene photonic waveguides have been implemented by [20, 23] where Parylene N is used as core and Parylene C is used as cladding material. A high contrast of refractive index between Parylene-C (used as core here) and PDMS (used as cladding) is favorable and results in highly restricted optical modes [24, 25].

While the overall design of these passive flexible neural probes is important, careful consideration muse be given to the input ports for each optical waveguide. These ports need to be specifically designed to enable efficient light coupling from external backend sources, ensuring optimal performance.

3.2 Active optical neural probes

To avoid connecting an external light source, Active optical neural probes introduce a light source to the neural probe, either in combination with a waveguide or directly on an electrode array. This ensures delivery of light deep into the target tissue with high spatial resolution. These light sources include light emitting diodes (LEDs) or lasers that emit light when powered by electricity [26]. There have been recent developments of implants with LDs and LEDs assembled or monolithically integrated on silicon based as well as polymer-based substrates [18, 27, 28, 29, 30].

For example, Gallium Nitride (GaN) LEDs have been utilized in the creating active photonic neural implants that produce blue light within the wavelength spectrum that aligns with the absorption range of Channelrhodopsin (ChR2), a frequently employed opsin in optogenetics, as well as fluorescent markers used for both structural and functional brain imaging. Pre-made LEDs have been packaged with neural probes such that metal traces and bondpads on neural probes are lithographically defined on a polymer substrate and LEDs are subsequently flip-chip bonded onto the polymer via bondpads. This method provides lot of flexibility such that LEDs emitting light at different wavelengths can be integrated on the same array to stimulate different opsins in any arbitrary arrangement, e.g. in cochlear implants [31].

Powering these light sources can result in substantial temperature increment which must be restricted in order to prevent damage to surrounding tissue. To prevent this – pulsing of light pulses can be incorporated rather than continuous use, to keep below the advised temperature increment of ~2 K for medical devices per ISO 14708-1. Additionally, design can be adjusted to incorporate wide spatial distribution between LEDs that stimulate simultaneously by distributing to multiple shanks or levels in the electrode array. Using waveguide internally, light source and illumination volume can be segregated to prevent overheating the tissue [19].

The materials employed for encapsulation, packaging, and waveguides must demonstrate biostability in terms of both insulation and optical transmission to guarantee the extended lifespan of devices during chronic implantation. Recently, a new monolithic fabrication process has been developed where flexible micro-LEDs are directly fabricated on silicon substrate with epitaxially grown GaN layers such that silicon layer is fully etched to release flexible devices. This approach provides flexibility in terms of size and arrangement of micro-LEDs that can be integrated in conjunction with recording electrodes [32]. Most of the approaches have only been tested for weeks and successful demonstration of reliable stimulation and recording performance is yet to be performed. Apart from shank type arrays, some studies on modification of planar electrodes for electrocardiograms [24, 25, 33, 34] and cuff-electrodes for peripheral nerves [35, 36] with optoelectronic stimulation capabilities are going on. Research efforts in this domain are gaining momentum and progressing along the right trajectory to enhance material capabilities and yield promising results.

4. Challenges and opportunities for optoelectronics in neural interfaces

Neuroscience has significantly progressed with advancement in neural modulation and recording techniques. Specific neuronal populations in target brain regions can be simultaneously recorded and stimulated using multimodal neural interfaces that combine optogenetics with electrophysiology and thus provide a powerful technique for understanding neural circuit functionality. Research and development on the optoelectronic front have emerged relatively recently with promising results for both passive and active optical neural implants. However, many challenges remain to be addressed in the development of multimodal tools for combined optogenetics and electrophysiology. Given the diverse strengths and limitations of each technology platform and their impact on suitability for specific applications, this section compares the current state of passive and active flexible optoelectronic technologies, focusing on the present challenges and opportunities for achieving a high longevity combined optogenetics and electrophysiology platform.

The active optoelectronic implants with micro-LEDs as functional element potentially allow a very high density of light sources. Miniature electrical traces can be densely routed (using advanced lithography techniques) to power these active light sources. On the other hand, passive optoelectronic implants require entire routing from backend to probe shank and cannot be defined over trajectories with sudden sharp bends. However, heat generation and dissipation in the probe shank due to lower efficiency of micro-LEDs is concerning. Temperature change of less than 1°C is prescribed as safe operating condition [37] but it depends upon the duration of exposure as well [38]. Two potential approaches can be adopted to alleviate this – (1) adding a heat-sink layer [39] (2) pulsing active light source to avoid heat accumulation [29]. Different geometric profiles of the micro-LEDs such as integrated planar [40] and hemispherical mirrors [41] have been explored to enhance efficiency of optical stimulation. Additional research is required to improve efficiency of these light sources and control heat dissipation into tissue. Conversely, passive optical waveguides deliver light from external sources to brain regions while minimizing heat transfer to the surrounding tissue in case of passive optoelectronic implants.

Majority of passive flexible optoelectronic interfaces developed so far consist of few optical channels. Miniaturization of waveguide core dimensions and therefore reduction of pitch between adjacent waveguides for high density routing (and therefore high number of channels) is limited by core waveguide materials as well as manufacturing methods. Additional overhead is added from assembly methods for integration of optical elements. Innovative approaches to ease this manufacturing overhead are required. One design-based approach to increase the density of optical channels, can be exploring multilevel metal traces as well as multilevel photonic guides while maintaining minimal crosstalk between adjacent channels.

Alternatively, a simple thermal drawing process has been explored by Polina et. Al [42], which is inspired by optical fiber production as a method of fabrication for multi-functional neural probes. During this process, using low-end manufacturing processing, a macroscale preform is fabricated which can be drawn into fiber such that lateral dimension can be scaled as much as 10,000-fold using multiple drawing processes & length can be stretched by a factor of ~100. This creates an extended fibrous device where cross-section structures on nanoscale can be created without the need of high-resolution fabrication technology. Another approach to achieve multi-site illumination without modifying the design or fabrication process is wavelength-based multiplexing. This technique involves routing multiple wavelengths of light from a single input waveguide to different output waveguides located at various spatial locations along the probe shank. Each output waveguide then transmits its assigned wavelength to a specific target site, enabling simultaneous stimulation at multiple locations. This unique assignment of different wavelengths neither requires additional power to operate nor generates additional heat [36].

Development of hermetic packaging for both active and passive optoelectronic approaches is another area of focus, where one approach being explored is combining inorganic layers with polymers to protect integrated micro-electronic circuits and conductors [43]. Longevity is a critical aspect to enable transfer of all the exciting approaches to chronic studies. Proof of concept experiments with multimodal process have demonstrated general feasibility such that existing active flexible optoelectronic probes are best suited for acute studies, where high-density light illumination is expected while passive optoelectronic probes have the potential for chronic long-term studies. However, current passive optoelectronic systems are too bulky with fragility concerns regarding the assembly of optoelectronic components to neural probes and therefore challenging to use for the chronic studies in freely behaving animals. Only a few groups have explored passive optoelectronic systems for small animal models [44, 45]. Development of fully implantable systems with wireless power supply and bidirectional data transmission would be the ideal direction to enable translational studies aimed at exploring therapeutic potential of optogenetics and employ miniaturized optoelectronic probes to potentially revolutionize bioelectronic medicine.

5. Conclusion

Progress in the field of neuroscience have been driven by the evolution of techniques for neural recording and stimulation. Multimodal neural probes, seamlessly integrating optoelectronics and electrophysiology, offer the potential for groundbreaking advancements in the field. These probes enable simultaneous recording and stimulation of specific neuronal populations within targeted brain tissue, making them powerful tools for fundamental neuroscientific research. Notably, in a groundbreaking study, Canales et al. (2015) presented a novel multifunctional fiber capable of simultaneously performing optical stimulation, neural recording, and drug delivery [46].

Further research efforts can be directed to development of a high longevity multifaceted probe combining additional capabilities like microfluidic drug delivery to potentially alter gene expression, deliver peptide ligands, and eventually manipulate animal behavior. Implementing these modalities with optoelectronics and electrophysiology, especially wirelessly [5, 47, 48, 49] is extremely challenging but if executed, it holds massive potential to enhance our understanding of mysteries of the brain and potentially develop promising solutions for diagnosis and treatment of neurological disorders.