Magnetic Resonance Imaging Pulse Sequence Selection for Optimal Time and Image Quality Enhancement

Abstract

This study is a comparison of three commonly used magnetic resonance imaging (MRI) pulse sequences to examine the image quality of the pulse sequences at a short acquisition time. Two tissue-equivalent gels were created. While one gel is constructed of polysaccharide and agarose, the other is made of ferrous benzoic xylenol orange (FBX). FBX gel is exposed to a 25 Grey dosage of 6MV photons from a linear accelerator. Repetition time (TR) was used to conduct experimental modifications in imaging parameters. The quantitative analysis comprises the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). Fast Spin Echo (FSE) and Fast Fluid Attenuated Inversion Recovery (FLAIR) are most comparable in SNR at 1.5 Tesla for various TR values. Conventional Spin Echo (CSE) has a CNR that is 143% and 93% higher than FSE and FLAIR, respectively. The time difference between CSE and FSE is 6 minutes and 34 seconds, whereas CSE and FLAIR is 6 minutes and 43 seconds. FSE and FLAIR provide superior image quality with quicker acquisition, suitable for patients sensitive to longer scan durations. Meanwhile, CSE stands out, delivering significantly enhanced contrast and SNR in T2-weighted images compared to other MRI pulses.

1. Introduction

1.1 Magnetic resonance imaging

An extremely versatile, noninvasive medical imaging method that produces high-quality images is magnetic resonance imaging (MRI). It opposed to ionising imaging techniques like X-ray CT. For image development, MRI does not use ionising radiation. It is a sophisticated diagnostic tool that offers precise anatomical information, good spatial resolution and strong cellular comparison. The sensitivity of MR signals to a variety of tissue factors allows for the acquisition of detailed information. MRI is the best imaging method for evaluating the cerebrum/skull, and cardiovascularactivities, in addition to the digestive organs, blood vessels, skeletal system, and abdomen because of its outstanding soft-tissue contrast [1, 2].

1.2 Image quality of MRI

Image quality is a parameter used to assess the diagnostic efficacy and interpretation of an image. The selection of the pulse sequence in MRI impacts the weighting and quality of the image, as well as its ability to respond to disorders. It is critical to understand these parameters and their interplay in order to achieve the best image quality [3]. Signal-to-noise ratio (SNR), the amount of pixels used to create an image in digital form, and difference between living organism tissue are the primary MRI picture quality and diagnostic factors for human tissues. All of these elements are interconnected and subject to the fundamental principles of NMR physics, it is difficult for them to improve at the same time. New developments in the field of MRI technology have resulted in increases in contrast and SNR through the modification of imaging parameters required. Many factors influence the MRI image quality. It is crucial to understand these aspects, how they interact, and how to get the best possible image quality. The image quality is primarily affected by four elements, which are

• Signal to noise ratio

• Contrast to noise ratio

• Image Homogeneity

• Scan time

1.3 Signal to noise ratio

The difference between the received signal’s amplitude and the noise’s average amplitude is known as the signal to noise ratio. The recipient coil generates a voltage that produces a signal as the NMV (nuclear magnetic vector) moves in a circular motion within the transverse plane.

Frequencies that randomly arrange themselves in space and time give rise to noises. Background electrical noise from the system and the patient’s presence in the magnet combined to create unnecessary noise in an MR environment. The presence of the patient in the MRI, the area being investigated, and the system’s inherent noise all contribute to this noise, which is constant for every patient. SNR increases and better images are produced when the signal gets comparatively shorter than the noise.

The SNR plays a key role in determining the clinical MRI quality; hence having the maximum SNR is necessary to avoid having a poor image quality.

1.4 Contrast to noise ratio

The comparative difference in signal intensity between two adjacent portions of a picture is also defined as contrast of the picture.

CNR quantifies the distinguishability or contrast between these regions, taking into account the noise level in the image. Contrast refers to the basic difference or distinction in the luminescence of each pixel within an image in the framework of MRI, which is determined by the intensity of the signal received from each voxel during the NMR experiment.

Variations in the spin relaxation velocities are the primary cause of the differences in signal intensity. Additionally, the hydrogen density of the tissues varies and is crucial for contrast discrimination.

MRI is a highly valuable technique for assessing oncological conditions because it offers exceptional contrast in soft tissues. By utilising a variety of pulse sequences that produce different contrasts, detailed evaluations of the disease’s size and extent can be conducted.

1.5 Image homogeneity

Image uniformity reflects how evenly the signal intensity is distributed across the image. A high level of image homogeneity is desirable as it ensures that the MR system is accurately capturing the underlying properties of the imaged object. Homogeneity is particularly important in clinical MRI to provide reliable and consistent image quality, aiding in accurate diagnosis and interpretation of the images. Factors such as magnetic field uniformity, scanner calibration, and appropriate. A common artefact in MR imaging, known by various names such as intensity non-uniformity, bias, inhomogeneity, or shading artefact, can impact the constancy of MR signal intensity.

To maximise clinical outcomes, it is important to understand the impact of changing the imaging parameters of an MR pulse sequence. The quality of an image, or MR intensity non-uniformity, is greatly influenced by the repetition time (TR) and number of echoes.

1.6 MRI scan time

MRI is a widely utilised technique for obtaining highly detailed images of various objects, such as the human body. The total scan time refers to the duration required to gather all the necessary data for generating the desired images or to complete the K-space filling process.

Longer scan times increase the patient’s likelihood of moving during the acquisition, which is critical for maintaining image quality. Any patient movement during the scan may cause the images to be compromised. However, extended acquisition times have the disadvantage of lowering image quality due to a number of artefacts, including respiratory artefacts. Reducing scan time in Magnetic Resonance Imaging (MRI) continues to be a crucial concern, particularly in clinical settings where diagnostic images need to be obtained. Acquiring images of excellent quality in a short period of time is crucial for optimising the diagnostic technique.

1.7 Acquisition time issues of MRI

Although MRI is a more accurate and non-invasive medical tool for clinical diagnosis, its lengthy acquisition time reduces its patient comfort and significance [4]. The duration of the MRI scan is crucial for maintaining image quality. Any patient movement during the scan could potentially result in blurry images [5]. However, one consequence of a long acquisition period [6] is that image quality reduction due to a variety of artefacts, including respiratory artefacts [7, 8].

Every advanced pulse sequence due to fast acquisition time possesses some negative aspect. Parallel imaging loses SNR and could result in technique-dependent artefacts, however it is a successful method for reducing scan times [9]. There is an inverse correlation between the speed of image acquisition and total image quality [10].

The available options for pulse sequences and acquisition factors are incredibly extensive and patient compliance is an issue with the selection of acquisition parameters [11]. Reducing the duration of MRI scans remains a significant challenge, especially when considering the acquisition of diagnostic images within a clinical setting [12].

Maintaining the image quality, i.e., contrast to noise ratio, signal to noise ratio and image uniformity, is crucial when selecting quick acquisition pulse sequences. Each pulse sequence performs differently as a result of its unique properties and attributes [13].

In MRI, increasing the value of TR results in longer acquisition durations for T2-weighted images. It is critical to gather high-quality images quickly in order to ensure an optimal diagnostic strategy [14]. The scan duration should always be as short as possible to minimise the possibility of patient movement.

A variety of approaches can be used to increase the quality of clinical information gained from an MR image. To analyse and choose the best technique for a certain organ, comparisons between various pulse sequences are always done. In order to evaluate the best imaging procedure for an MRI of a very short T2, in 2016, Ali Caglar Ozen and colleagues compared MRI of an ancient mummified human hand using an ultra-short echo time sequence [15]. To examine the effectiveness of widely used soft tissue suppression techniques on a quantitative level, Chang Li in 2012 compared optimised soft tissue suppression schemes for ultrashort echo time MRI [16]. Michael P. Recht increases the value of MRI by reengineering the MRI workflow to reduce MRI acquisition time [17].

1.8 Purpose of the comparison of MRI pulse sequences

The aim of this study is to evaluate and compare the commonly employed pulse sequences used at the clinical level. The purpose is to determine their performance and select the most appropriate pulse sequence based on factors such as SNR, CNR, and collection time, specifically for T2-weighted images. The optimum image quality at a quick acquisition time would be achieved with a continuous range of TR. Additionally; the effectiveness of the traditional spin echo sequence in the presence of various quick pulse sequences was examined in the present investigation. Fast Fluid Attenuated Inversion Recovery (FLAIR), Conventional Spin Echo (CSE), and Fast Spin Echo (FSE) were the three pulse sequences that were used the most frequently in the present study.

2. Phantom preparation

2.1 First gel

A polysaccharide gel along with agarose is used in an experiment at Ninewells Hospital and Medical School in Dundee, UK, to create a substance that is similar to tissue for MRI. For imaging applications, this substance contains gadolinium chloride attached to ethylene diamine tetraacetic acid (EDTA). The T1 and T2 values of this substance may differ, independently as a result of varying the proportions of each ingredient.

When creating a substance that is an equivalent to biological tissue, the gadolinium ions are chelated to the macromolecule EDTA, which has three distinct benefits. First, the chelation takes away any chance that the ions will engage in any further chemical reactions with the gel matrix. Second, chelation may prevent the hydroxyl ions from being released by the gadolinium ions. The chelation procedure moderately affects the Gd-EDTA solution’s empirical relaxing characteristics. This effect, however, becomes more pronounced at higher frequencies, notably above 30 MHz [18].

This work employs seven 12 mm diameter phantoms. For those phantoms, the T1/T2 intervals of relaxation are 608/134, 759/155, 917/135, 986/220, 1050/164, 1180/221, and 1296/200 (msec). The 1.5 T unit (Siemens MAGNETOM Avanto, UK) is used for MR imaging (Figure 1). A 1.5 mm region of interest (ROI) was cantered in the gel for estimating signal intensity and replicating the ROI to measure the signal intensity of the background noise. SNRs have been determined with the following equation.

Figure 1.

Polysaccharide, comprise the range of relaxation value for biological tissues at Siemens MAGNETOM Avanto 1.5 T.

$$\text{SNR}=\text{S}\text{I}/\text{N}$$E1

Where N is the surrounding standard error of variance and SI is the mean magnitude of the signal of the ROI situated in the middle of the gel. SNR was examined using software called Image J. Phantom scanning uses the CP Head Coil of the MRI. Certain imaging parameters of CSE, FSE, and FLAIR were constant throughout the investigation, comprising the following parameters: the entire amount of acquisitions (1), the proportion of sampling (100), the field of view (100 mm x100 mm), the pixel per mm resolution (1.280), and the segment width (4 mm). Furthermore, the FSE pulse sequence had a consistent inversion time of 860 ms and an echo train length of 5, whereas the FLAIR pulse sequence had a constant inversion time of 860 ms and an echo train length of 5.

The outcomes of our evaluation were assessed using the practical information provided by the MHRA (Medicine and Healthcare Products Regulatory Agency) Evaluation 04133 for the Siemens Magetom Avanto 1.5 T system [19]. The percentage error indicates the discrepancy between the value that was observed and the actual value. Using MATLAB 7.7 (R2008b), the curve fitting approach is used to approximate the optimised value.

Gelatine (from bovine skin, Type B), sulphuric acid (Sigma-Aldrich), Xylenol Orange Tetrasodium salt (Sigma-Aldrich), and Benzoic Acid (Sigma-Aldrich) were used to make the second gel, Ferrous Benzoic Xylenol Orange (FBX), which was developed in 1998 by Kelly RG [20] alongside other researchers [21, 22, 23].

2.2 Second gel

A one-litre capacity container was used to combine 5 ml of benzoic acid, 1 ml of Xylenol Orange, and 25 ml of sulphuric acid to create the stock solution, which was then left to sit at room temperature. The first step in creating gel is to mix 40 g of gelatine with 700 ml of distilled water, 25 ml of sulphuric acid, and a hot plate that has been prepared to 40°C. After stirring continuously for 30 minutes, the gel’s gelatine melted. 0.1 mm of ferrous sulphate was dissolved in 100 ml of Xylenol vibrant orange base solution with benzoic acid. The resulting solution was then added to the liquid gelatin. By adding 25 ml of the solution, a gel with a final volume of 1 litre was created. The preliminary oxygen concentrations in the solution affect how the Fricke gel dosimeter responds. During preparation, the gel is exposed to the air, and six test containers with a 10 ml capacity are used to pour the gel into for the irradiation of different doses. At 5°C, all gel phantoms were kept [24].

2.3 Gel irradiated and MRI scanning

Radiation was applied to the gel using a 6MV photon beam generated by a Varian Clinic 600C Linear accelerator. The dose was 25GY delivered at a Source to Surface Distance (SSD) of 95.5 cm with a field area of 55 cm². Siemens MAGNETOM Avanto, a 1.5 T machine, is used to perform MR imaging. Phantom scanning uses the CP body Coil of the MRI. The CSE, FSE, and FLAIR pulse sequences were used to image the phantom (Figure 2).

Figure 2.

FXG phantom (a) after irradiation; deliver dose is 25 Grey with linear accelerator, 600 MV X-ray energy (b) MR image of FGX phantom in CSE.

A 1.5 mm square ROI was set in the gel’s middle to analyse signal intensities for quantitative image processing. The same ROI area was used to measure noise within the background. This process is performed throughout each pulse sequence. The following formula is used to determine contrast to noise ratios (CNRs):

$${\text{CNR=SNR}}_{\text{A}}{\text{–SNR}}_{\text{B}}$$E2

The phantom’s contrast to noise ratio of exposed to radiation and non-exposed to radiation regions is denoted by SNR_A and SNR_B, respectively.

For the T2-weighted investigation, imaging parameters for CSE, FSE, and FLAIR are kept constant (100 x 100 mm field of view; 4 mm slice thickness). The number of acquisitions was 1; for the T2-weighted study, the FLAIR inversion time was 2500 ms; the length of the echo train was 21; and for the FSE, the echo train length was 21.

2.4 T1/T2 calculation of FXG gel with variation of deliver dose

The same methods used by Afzal et al. [25, 26], Bartusek et al. [20, 27], and MATLAB version (R2008b) are utilised to calculate T1/T2 for FGX Gel (Table 1).

Dose	T1 (msec)	T2 (msec)
O Grey (No dose)	812	166
25 Grey	628	58

Table 1.

Calculated values of FGX at 0 gray and 25 gray.

3. Analysis of CSE, FSE, and FLAIR responses for SNR and CNR for various phantoms and parameters

The T2-weighted images and prolonged acquisition time made CSE less noteworthy compared to other pulse sequences. Both CSE and FLAIR generate the necessary SNR, but FLAIR has an advantage because to its quick acquisition time, as seen in Table 2, T1/T2 time of the phantom is 608/134 (msec). For accurate SNR of the image, FLAIR takes 56% less time to acquire than CSE (Figure 3a).

Sr. no	Pulse sequences	T1/T2 of phantom (msec)	TR (msec)	SNR	Percentage error (%)	Acquisition time (min: sec)
1	CSE	608/134	1800	130.23	−11.40	5.9
			2000	140.72	−4.27	6.29
			2200	146.83		7.08
			2400	149.44		8.59
2	FSE	608/134	4000	127.52	−13.24	0.48
			4200	133.44	−9.22	1.29
			4400	139.02	−5.42	1.5
			4600	160.33	2.12	2.18
3	FLAIR	608/134	4000	128.19	−12.79	1.28
			4500	133.19	−9.39	1.42
			5000	134.43	−8.55	1.5
			6000	154.59		2.58

Table 2.

Time optimization with variation of TR for SNR. T1/T2 608/134 (msec).

Figure 3.

Comparison between the pulse sequences for the appropriate SNR at short acquisition time (a) T1/T2 of the phantom is 608/134 msec. (b) T1/T2 of the phantom is 759/155 msec. (c) T1/T2 of the phantom is 917/135 msec. (d) T1/T2 of the phantom is 986/220 msec.

The minimum TR value in Table 3, the phantom’s T1/T2 time is 759/155 (msec), CSE is the one that best MRI technique that satisfies the SNR requirement. FSE or FLAIR, on the other hand, can be considered better pulse sequences due to their fast acquisition times. For achieving an appropriate SNR of the image, the FSE and FLAIR pulse sequences were found to take 79% and 80% less time, respectively, compared to the CSE sequence (Figure 3b).

Sr. no	Pulse sequences	T1/T2 of phantom (msec)	TR (msec)	SNR	Percentage error (%)	Acquisition time (min: sec)
1	CSE	759/155	1800	153.43		5.9
			2000	188.21	19.88	6.29
			2200	196.39	25.08	7.08
			2400	201.07	28.07	8.59
2	FSE	759/155	4000	141.78	−3.54	0.48
			4200	148.05		1.29
			4400	154.30		1.5
			4600	155.76		2.18
3	FLAIR	759/155	4000	143.99	−2.04	1.28
			4500	149.73		1.42
			5000	151.26		1.5
			6000	153.61		2.58

Table 3.

Time optimization with variation of TR for SNR. T1/T2 759/155 (msec).

As demonstrated in Table 4, only CSE produces SNR for images with a minimal TR. T1/T2 is 917/135 (msec) for the phantom. The significance of other pulse sequences is diminished for this particular phantom due to inadequate SNR (Figure 3c).

Sr. No	Pulse sequences	T1/T2 of phantom (msec)	TR (msec)	SNR	Percentage error (%)	Acquisition time (min:sec)
1	CSE	917/135	1800	169.85	8.18	5.9
			2000	186.43	18.74	6.29
			2200	190.23	21.16	7.08
			2400	187.61	19.49	8.59
2	FSE	917/135	4000	89.48	−39.12	0.48
			4200	92.97	−36.75	1.29
			4400	96.79	−34.15	1.5
			4600	105.05	−28.53	2.18
3	FLAIR	917/135	4000	97.37	−33.76	1.28
			4500	102.25	−30.43	1.42
			5000	103.54	−29.55	1.5
			6000	115.05	−21.73	2.58

Table 4.

Time optimization with variation of TR for SNR. T1/T2 917/135 (msec).

CSE and FSE both are present in the significant domain of SNR, as seen in Table 5; T1/T2 time for that phantom is 986/220 (msec) while the faster acquisition time of T2-weighted images makes FSE more prevalent. For images with a good SNR, the FSE acquires data 79% quicker compared to the CSE (Figure 3d).

Sr. No	Pulse sequences	T1/T2 of phantom (msec)	TR (msec)	SNR	Percentage error (%)	Acquisition time (msec)
1	CSE	986/220	1800	124.39	−15.64	5.9
			2000	139.46	−5.124	6.29
			2200	145.72		7.08
			2400	149.55		8.59
2	FSE	986/220	4000	135.31	−7.951	0.48
			4200	141.76	−3.561	1.29
			4400	148.07		1.5
			4600	155.30		2.18
3	FLAIR	986/220	4000	130.37	−11.311	1.28
			4500	134.51	−8.495	1.42
			5000	135.06	−8.115	1.5
			6000	141.75	−3.565	2.58

Table 5.

Time optimization with variation of TR for SNR. T1/T2 986/220 (msec).

As shown in Table 6, the T/T2 time of this phantom is 1050/164 (msec), FSE and FLAIR created the image’s SNR in the shortest possible period of time, FLAIR and CSE also provided a good image with the least amount of error for given values. FLAIR can be a good option for selecting various TR values because it has a quicker acquisition time than CSE. FSE and FLAIR have precise SNR of the image and acquisition times that are 63% and 56% faster than CSE, respectively (Figure 4a).

Sr. no	Pulse sequences	T1/T2 of phantom (msec)	TR (msec)	SNR	Percentage error (%)	Acquisition time (msec)
1	CSE	1050/164	1800	158.06		5.9
			2000	162.18	3.30	6.29
			2200	165.16	5.19	7.08
			2400	166.33	5.94	8.59
2	FSE	1050/164	4000	120.02	−18.35	0.48
			4200	124.97	−14.98	1.29
			4400	127.54	−13.23	1.5
			4600	146.27		2.18
3	FLAIR	1050/164	4000	133.31	−9.31	1.28
			4500	139.82	−4.88	1.42
			5000	141.91	−3.45	1.5
			6000	159.56		2.58

Table 6.

Time optimization with variation of TR for SNR. T1/T2 1050/164 (msec).

Figure 4.

Comparison between the pulse sequences for appropriate SNR at short acquisition time. (a) T1/T2 of the phantom is 1050/164 msec. (b) T1/T2 of the phantom is 1180/220 msec. (c) T1/T2 of the phantom is 1296/200 msec. (d) Comparison among pulse sequences for good CNR at short acquisition time,T1/T2 of the phantoms are 628/58 and 812/166 (msec).

While FSE and FLAIR may be preferable in terms of the time of acquisition, as demonstrated in Table 7, the phantom’s T1/T2 time is 1180/221 (msec); CSE, FSE, and FLAIR equally generate accurate SNR. To achieve an accurate signal-to-noise ratio (SNR) of the image, both the FSE and FLAIR pulse sequences require 76% and 77% less time, respectively, compared to the CSE sequence (Figure 4b).

Sr. no	Pulse sequences	T1/T2 of phantom (msec)	TR (msec)	SNR	Percentage error (%)	Acquisition time (msec)
1	CSE	1180/221	1800	140.68	−4.76	5.9
			2000	150.88		6.29
			2200	156.62		7.08
			2400	159.58	1.64	8.59
2	FSE	1180/221	4000	139.54	−5.06	0.48
			4200	145.79		1.29
			4400	152.29	−0.82	1.5
			4600	155.36		2.18
3	FLAIR	1180/221	4000	140.58	−4.36	1.28
			4500	146.24		1.42
			5000	147.43		1.5
			6000	152.34		2.58

Table 7.

Time optimization with variation of TR for SNR. T1/T2 1180/221(msec).

Table 8 demonstrates that FLAIR is the only technique that can provide an image’s SNR with the smallest errors and the shortest acquisition time (Figure 4c).

Sr. No	Pulse sequences	T1/T2 of phantom (msec)	TR (msec)	SNR	Percentage error (%)	Acquisition time (min: sec)
1	CSE	1296/200	1800	169.75	8.12	5.9
			2000	195.49	24.51	6.29
			2200	200.60	27.77	7.08
			2400	203.09	29.35	8.59
2	FSE	1296/200	4000	123.82	−15.76	0.48
			4200	128.08	−12.86	1.29
			4400	133.62	−9.09	1.5
			4600	137.85	−6.22	2.18
3	FLAIR	1296/200	4000	133.38	−9.26	1.28
			4500	140.73	−4.26	1.42
			5000	142.76	−2.87	1.5
			6000	147.13		2.58

Table 8.

Time optimization with variation of TR for SNR. T1/T2 1296/200 (msec).

According to Table 9, T1/T2 628/58 (msec) & 812/166 (msec), the acquisition time of CSE in the T2-weighted study is longer than that of FSE and FLAIR. When compared to the FSE and FLAIR pulse sequences, the CSE sequence has a 144% and 94% superior contrast-to-noise ratio (CNR). CSE must be used for a long time in T2-weighted images to create a contrast between tissues (Figure 4d).

Sr. No	Pulse sequences	T1/T2 of phantom (msec)	TR (msec)	CNR	Percentage increase in CNR (%)	Acquisition time (min: sec)
1	CSE	628/58	1800	41.864	13%	5.51
		812/166	2000	47.292	7%	6.49
			2200	50.687	7%	7.07
			2400	54.267		7.48
2	FSE	628/58	3800	18.028	17%	0.45
		812/166	4000	21.147	2%	0.47
			4200	21.567	3%	0.59
			4400	22.283		1.14
3	FLAIR	628/58	3500	17.075	19%	0.43
		812/166	4000	20.331	19%	0.46
			4500	24.149	16%	0.54
			5000	28.002		1.05

Table 9.

Time optimization with variation of TR on CNR. Deliver dose 25 Grey & 0 Grey. T1/T2 628/58(msec) & 812/166 (msec).

4. Importance of acquisition time and comparison of pulse sequences

In clinical practice, the duration of acquisition is crucial in T2-weighted MRI imaging. The three most frequently used pulse sequences that are regularly employed at the clinical level for diagnostic purposes have been chosen. The idea of selecting the right sequence and optimising the experimental parameters to achieve image excellence at a quick acquisition time was shown by comparing these pulse sequences. Regarding FBX and polysaccharide gels, we assessed the CNR and SNR, respectively. We evaluated the performance of different pulse sequences by comparing SNR and CNR at the appropriate acquisition time.

Many options for TR in each pulse sequence were tested and researched in order to identify the pulse sequence with the most favourable combination for improved image quality with a respectably fast acquisition time. Each pulse sequence is usually exceptionally effective at selecting the right parameters.

SNR is a measurement of image quality. With an increase in TR, the image’s SNR rises. The TR response is determined by the transverse time to the relaxation of a phantom object or biological cells [28] because T2 decay is caused by energy transfer between spins. To make T2 the dominant factor in the signal decay, a very long TR will be required. TR is a component that lengthens the process of acquiring the image. For the image’s SNR and acquisition time to be maintained, an appropriate TR is essential. Image noise and contrast may become a limiting factor when TR is decreased to shorten the time required for image acquisition [29].

According to Tables 2 and 6, by considering the phantoms 608/134 (msec) and 1050/164 (msec), SNR of the pulse sequences CSE and FLAIR in the T2-weighted examinations are comparable. But CSE and FLAIR have quite different acquisition times. When compared to CSE, FLAIR’s acquisition time is 63% and 56% faster. While other pulse sequences failed to sustain SNR at particular TR values, for the identical phantom, FLAIR covers a range of −33% to −21% while FSE has a percentage error between −39% and − 28%., Table 4 shows that CSE is noticeably good for 917/155 (msec). For large T1/T2 weighting phantoms with an acquisition time of 1180/221 (msec), Table 7 compares CSE to FLAIR and FSE, despite the fact that CSE has an acquisition time that is 372% and 398% as high as FLAIR and FSE, correspondingly.

In T2 weighted study as indicated in Table 9, the provided dosages are 25 grey and 0 grey, and T1/T2 values are 628/48 (msec) and 812/166 (msec) respectively, the signal intensity variation throughout the tissues is very significant. Table 9 demonstrates the great disparity among tissues generated by the CSE at long TR. At the diagnostic stage, this pinnacle of quality in CNR is extremely appreciable and desired.

In MRI study, CSE’s T2-weighted acquisition time is 556% and 612% longer than FSE, FLAIR, correspondingly; nevertheless, CSE’s CNR is 144% and 94% better with TR selection. As a result, as illustrated in Tables 2 and 9, CSE must be utilised for an extended period of time to establish contrast among biological cells, in images that are T2-weighted.

As illustrated in Tables 3 and 6, FSE demonstrated very favourable outcomes for the phantoms of & 759/155 (msec), 1180/221 (msec). Similar to this, FSE for the phantom 608/134 (msec), 986/220 (msec), and 1296/220 (msec) in Tables 2,5, and 8 correspondingly show high SNR with low percentage error. As shown in Table 6, the FSE cannot be selected for the T1//T2 phantom of 917/135 (msec), the percentage inaccuracy is large, and consequently the SNR is very low for the phantom of 1050/164 (msec). The level of accuracy offered by CSE could never be matched by CNR in FSE. Because all the echoes were averaged into one k-space in the T2-weighted investigation, the adoption of a long turbo factor reduced the significance of FSE [30]. The MT (magnetization transfer), which reduces the contrast between normal and pathological tissues, also has an impact on FSE. However, by adjusting the echo factor, the contrast of the tissues can be adjusted [31, 32, 33, 34, 35]. In a T2-weighted investigation of FSE, optimal parameters are required to obtain improved CNR.

For a number of tissues, FLAIR and FSE are equivalent to one another in terms of SNR, and imagine acquisition time for tissues with T1/T2 time of 608/134 (msec) and 1296/200 (msec), respectively, as shown in Tables 8 and 9, FLAIR is an excellent substitute. The optimal TR selection improves CNR by 55% in FLAIR, which is often comparable to T2-weighted CSE. CNR for FLAIR has increased by 55%, however it is still 48% lower than CSE’s. FLAIR has a strong tendency by employing TI (the time which relates to the null point of particular tissues), making an image contrast between tissues more visible by nullifying the signal for specific cells [36]. Images with poor inversion timing lack contrast between adjacent tissues. The inversion time and repetition time have a significant impact on the signal intensity differences of diseased tissues [13]. To get a strong contrast between tissues, the best inversion time is needed. These results underline the need for precision in determining the ideal imaging parameters at the diagnostic stage, in addition to the application of T2-weighted pulse sequences for certain tissue.

5. Conclusion of the comparison of different pulse sequences

In this study, we analysed factors within a pulse sequence optimally with the goal of obtaining image quality with a quick acquisition time. The correct TR value in each individual pulse sequence has a substantial impact on the precision of T2 measurement in MRI. It is clear from the results that CSE produced the highest SNR for a variety of tissues and produced a striking contrast in the images that are T2-weighted. Studies have demonstrated this, while the greater time required for acquisition seems less appealing, it does not reduce its utility in T2-weighted MRI studies. Despite the benefits of a quick collecting time and high image quality, FSE pulse sequence may be the best option once the obstacles connected with the complex interaction between imaging parameters and echo train length have been overcome. Regarding SNR, FLAIR is equivalent to FSE for a number of tissues. In a T2-weighted MRI scan, FLAIR images can also be advantageous pulse sequences with good SNR and quick acquisition times for different tissues. A pulse sequence with exceptional image quality in a quick period of acquisition was chosen as a result of comparing pulse sequences based on acquisition time. Indeed, a pulse sequence that offers excellent image quality and a fast acquisition period would be an ideal choice for clinical MRI. By combining high-quality images with efficient acquisition times, healthcare professionals can enhance patient care and improve workflow efficiency in a clinical setting.

Acknowledgments

We are thankful to the staff of the MRI department of Ninewells Hospital and Medical School, Dundee, UK for their support in collecting data and facilitating this research work.