Multiplexed sensitivity-encoding diffusion-weighted imaging (MUSE) in diffusion-weighted imaging for rectal MRI: a quantitative and qualitative analysis at multiple b-values

Abstract

Purpose:

To compare four diffusion-weighted imaging (DWI) sequences for image quality, rectal contour, and lesion conspicuity, and to assess the difference in their signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), and apparent diffusion coefficient (ADC).

Methods:

In this retrospective study of 36 consecutive patients who underwent 3.0 T rectal MRI from January–June 2020, DWI was performed with single-shot echo planar imaging (ss-EPI) (b800 s/mm²), multiplexed sensitivity encoding (MUSE) (b800 s/mm²), MUSE (b1500 s/mm²), and field-of-view optimized and constrained undistorted single-shot (FOCUS) (b1500 s/mm²). Two radiologists independently scored image quality using a 5-point Likert scale. Inter-reader agreement was assessed using the weighted Cohen’s κ. SNR, CNR, and ADC measurements were compared using the paired t-test.

Results:

For both readers, MUSE b800 scored significantly higher for image quality, rectal contour, and lesion conspicuity compared to ss-EPI; MUSE b800 also scored significantly higher for image quality and rectal contour compared to all other sequences. Lesion conspicuity was equally superior for MUSE b800 and MUSE b1500 compared to the other two sequences. There was good to excellent inter-reader agreement for all qualitative features (κ=0.72–0.88). MUSE b800 had the highest SNR; MUSE b1500 had the highest CNR. A significant difference in ADC was observed between ss-EPI compared to the other sequences (p<0.001) and between MUSE b800 and FOCUS. No significant difference in ADC was found between MUSE b1500 and FOCUS b1500.

Conclusion:

MUSE b800 improved image quality over ss-EPI and both MUSE b800 and b1500 showed better tumor conspicuity compared to conventional ss-EPI.

INTRODUCTION

The most common imaging sequence used in diffusion-weighted magnetic resonance imaging (DW-MRI) is single-shot echo planar imaging (ss-EPI) [1]. High-quality DW-MRI images are critical for imaging of rectal cancer, and radiologists often look to improve DWI quality and explore advanced MRI techniques to improve patient care [2–4].

With an ss-EPI sequence, the entirety of the k-space is acquired with a single excitation, with typical acquisition times ranging on the order of tens of milliseconds [5, 6]. This fast acquisition essentially “freezes” the effects of patient motion, creating images with negligible motion artifacts. However, ss-EPI is very susceptible to main field inhomogeneities, local gradients such as air-tissue and bone-tissue boundaries, and chemical shift artifacts, all of which can lead to severe image degradation.

Interleaving EPI trajectories along the phase-encoding direction, commonly referred to as multi shot-EPI (ms-EPI) [7] or segmental EPI, has been used to divide k-space into multiple segments. In ms-EPI, the echo train is divided into several parts in an interleaved manner to reduce the echo train of a single image along the phase-encoding dimension and to increase bandwidth per pixel in the phase-encoding direction. The advantages of ms-EPI are that spatial resolution can be increased; thus, the distortion artifacts of EPI acquisitions are reduced. The major drawbacks of ms-EPI include reduced robustness against motion artifacts as compared to ss-EPI, and the increased acquisition time.

Multiplexed sensitivity encoding (MUSE) was recently developed for the reconstruction of ms-EPI to correct motion-induced phase errors. For each segment, a phase navigator is inherently calculated and, along with parallel imaging calibration, used to solve the motion-induced phase errors [8]. Recent studies suggest that DW images acquired with MUSE have superior performance in several areas, including high-resolution DWI of the brain [9], breast [10], and prostate [11], and in the assessment of small bowel disease [12].

Another promising approach to diffusion-weighted imaging — which has been developed and studied in the spine [13], for tumors of the neck [14], and in pancreatic [15], breast [16], prostate [17], and rectal [18] cancers — is field-of-view (FOV) optimized and constrained undistorted single-shot (FOCUS; GE Healthcare, Milwaukee, Wisconsin) DWI. FOCUS uses a spatially 2D-selective echo-planar radiofrequency excitation pulse and a 180° refocusing pulse to reduce the FOV and allow for fast acquisition of data from a limited volume using a 2D spatially selective [13, 19] pulse. Images can be generated with a reduced FOV in the phase-encode direction without the need for a longer readout, and contiguous multi-slice images can be acquired with fat signal suppression.

The aim of the study was to compare ss-EPI b800, FOCUS b1500, MUSE b800, and b1500 in rectal MRI for image quality, rectal contour, and lesion conspicuity, and to assess the difference in their signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), and apparent diffusion coefficient (ADC).

MATERIAL AND METHODS

Inclusion of Patients

Our institutional review board waived the requirement for informed consent for this retrospective study, which was compliant with the Health Insurance Portability and Accountability Act. During a 5-month period between January and June 2020, we evaluated 64 consecutive patients with the following inclusion criteria: patients with rectal cancer who underwent a 3.0 T MRI exam of the rectum either for baseline staging or re-staging. All studies must include DW-MRI with ss-EPI b800, FOCUS b1500, MUSE b800, and MUSE b1500 performed in one exam.

The exclusion criteria for qualitative analysis include: (1) patients who did not get all the DWI sequences being studied; (2) DW images deemed to be of non-diagnostic quality (i.e., non-diagnostic secondary to motion artifact or susceptibility artifact from prosthesis or bowel peristalsis); and (3) DWI images that showed no lesions. In addition, for quantitative analysis, patients were excluded if they had undergone chemoradiotherapy (Fig. 1).

Fig. 1.

Patient inclusion flowchart. Abbreviations: CRT, chemoradiotherapy; MRI, magnetic resonance imaging; MUSE, multiplexed sensitivity-encoding diffusion-weighted imaging; FOCUS, field-of-view optimized and constrained undistorted single-shot; ss-EPI, single-shot echo planar imaging

MRI Technique

All imaging was performed on a single 3 T MRI scanner (Signa Premier, GE Healthcare, Milwaukee, WI, USA) using a body coil for excitation and a flexible surface phased-array coil (30 channel AIR Anterior Coil and Poterior Coil, GE Healthcare) for reception. MRI sequences included axial and sagittal large FOV T2-weighted imaging (T2WI), and axial and coronal oblique T2WI through the tumor/tumor bed, followed by axial DWI using ss-EPI with b-values of 0 and b800 s/mm², FOCUS with b-values of 0 and b1500 s/mm², MUSE with b-values of 0 and b800 s/mm², and MUSE with b-values of 0 and b1500 s/mm². The technical parameters of all MRI sequences are listed in Table 1.

Table 1:

Sequence Parameters

	ss-EPI (b=800)	MUSE (b=800)	MUSE (b=1500)	FOCUS (b=1500)	Sagittal T2W	Axial T2W	Oblique axial T2W
b-value, s/mm2	0, 800	0, 800	0, 1500	0, 1500	NA	NA	NA
TR, ms	2905–5181	3642–5081	3000–5138	3079–5415	4000–6000	4000–6000	4000–6000
Minimum TE, ms	53.7	74.5	74.5	55.3	102	102	102
Slice thickness (mm)	5	5	4	4	4	5	3
Number of slices	46	40	32	32	28	49	21
Matrix	128×128	120×120	320×192	140×70	320×320	320×320	320×320
FOV, mm2	240×240	240×240	240×240	240×120	180×180	200×200	180×180
In-plane resolution, mm	1.9×1.9	2.0×2.0	0.75×1.25	1.7×1.7	0.56×0.56	0.625×0.625	0.6×0.6
Pixel bandwidth, Hz/pixel	1953	1953	1953	1953	32	50	32
Averages	14	14	16	16	2	1	4
Acquisition time, min	6–8	6–8	6–8	6–8	4–5	1–2	5–6

Abbreviations: FOV, field of view; MUSE, multiplexed sensitivity-encoding diffusion-weighted imaging; FOCUS, field-of-view optimized and constrained undistorted single-shot; ss-EPI, single-shot echo planar imaging; T2W, T2-weighted; TE, echo time; TR, repetition time

Qualitative Image Analysis

All images were analyzed separately by two radiologists (DB with 5 years of experience reading rectal MRI and ME with 2 years of experience). Both readers were blinded to the clinical findings and DWI sequence and were asked to interpret the quality of the DW images. Images were reviewed on a picture archiving and communication system workstation (Centricity PACS, GE Healthcare, Milwaukee, WI). The readers scored the images for image quality (based on noise and geometric distortion), sharpness of the rectum, and rectal lesion conspicuity. Qualitative scoring was performed using a 5-point Likert scale ranging from 1–5 (Table 2).

Table 2.

Description of DWI qualitative assessment using a 5-point Likert scale

Score	Image Quality	Rectal Contour	Lesion Conspicuity
1	Non-diagnostic quality	Non-diagnostic quality	Barely conspicuous
2	Substantial deficits in image quality	Substantial deficits in image quality	Fairly conspicuous
3	Moderate image quality	Moderate image quality	Moderately conspicuous
4	Good image quality	Good image quality	Well seen
5	Excellent image quality	Excellent image quality	Very well and sharply defined

Figures 2–3 show representative images of rectal cancer obtained with ss-EPI (b = 800 s/mm²), FOCUS (b = 1500 s/mm²), MUSE (b = 800 s/mm2), and MUSE (b = 1500 s/mm²).

Fig. 2.

An 83-year-old male underwent MRI for rectal cancer staging. Oblique axial T2-weighted imaging (a) shows a circumferential rectal tumor (arrow). Multiplexed sensitivity-encoding diffusion-weighted imaging (MUSE) b800 (b), MUSE b1500 (c), single-shot echo planar imaging (ss-EPI) (d), and field-of-view optimized and constrained undistorted single-shot (FOCUS) (e) sequences show restricting tumor with a corresponding low apparent diffusion coefficient (ADC) map (f). Both readers had 100% agreement that MUSE b1500 delivered excellent image quality, rectal wall contour, and very sharply defined tumor (score 5). MUSE b800 and ss-EPI had excellent image quality and rectal wall contour (score 5) with well visualized tumor (score 4). FOCUS was found to have good image quality and rectal wall contour (score 4) with moderately conspicuous tumor (score 3)

Fig. 3.

A 58-year-old female underwent MRI for rectal cancer restaging post-chemoradiotherapy. Oblique axial T2-weighted imaging (a) shows scar tissue with intermediate signal intensity tissue, suspicious for residual tumor (arrow). Multiplexed sensitivity-encoding diffusion-weighted imaging (MUSE) b800 (b), MUSE b1500 (c), single-shot echo planar imaging (ss-EPI) (d), and field-of-view optimized and constrained undistorted single-shot (FOCUS) (e) sequences show restricting tumor with a corresponding low apparent diffusion coefficient (ADC) map (f). Both readers considered the tumor well-visualized (score 4) on MUSE b1500 with good rectal wall contour (score 4). Reader 2 found MUSE b800 to visualize the lesion better than did Reader 1 (moderately conspicuous)

Quantitative Image Analysis

Elliptical regions of interest (ROIs) were drawn in PACS by one of the radiologists (DB) on each of the DWI sequences to include as much of the rectal tumor as possible. The ROI for normal tissue were drawn in normal rectal muscularis layer where possible, separate from the tumor or adjacent structures (intraluminal air or mesorectal fat, for example). As background noise level cannot be reliably measured [20, 21], SD_tissue was used to measure local noise for SNR and CNR calculations.

The mean signal intensity on corresponding ADC maps and standard deviation were obtained from each ROI measurement.

A medical physicist (YM) calculated SNR and CNR for each DWI sequence. The SNR was defined as the ratio between the mean signal intensity inside the tumor (S_tumor) and the standard deviation of tissue (SD_tissue) [20, 21] as given by:

$$SNR=\frac{S_{\mathrm{t}\mathrm{u}\mathrm{m}\mathrm{o}\mathrm{r}}}{{SD}_{\mathrm{t}\mathrm{i}\mathrm{s}\mathrm{s}\mathrm{u}\mathrm{e}}}$$ [1]

The CNR was defined as the ratio of the mean signal intensity difference between tumor and normal tissue (S_tissue) divided by the sum of the squared standard deviation of the tumor and squared standard deviation of tissue using the following equation:

$$CNR=\frac{|S_{\mathrm{t}\mathrm{u}\mathrm{m}\mathrm{o}\mathrm{r}}-S_{\mathrm{t}\mathrm{i}\mathrm{s}\mathrm{s}\mathrm{u}\mathrm{e}}|}{\sqrt{{SD}_{\mathrm{t}\mathrm{u}\mathrm{m}\mathrm{o}\mathrm{r}}^2+{SD}_{\mathrm{t}\mathrm{i}\mathrm{s}\mathrm{s}\mathrm{u}\mathrm{e}}^2}}$$ [2]

Statistical Data Analysis

Statistical analysis, including power analysis for the given number of included patients, was performed using MATLAB 2016a (MathWorks, Natick, MA, USA). Inter-reader agreement was assessed by applying the procedure of weighted Cohen’s κ. Accordingly, the range of κ > 0.8 was considered excellent, κ = 0.61–0.80 good, κ = 0.61–0.80 moderate, κ = 0.21–0.40 fair, and κ = 0–0.20 poor [22]. Image scores were compared using the Wilcoxon rank-sum test with Bonferroni correction. SNR, CNR, and ADC values calculated for the ROIs were averaged to report the mean and standard deviation (SD). SNR, CNR, and ADC measurements between sequences were compared using the paired t-test. A P value < .05 was considered statistically significant. The Bonferroni correction was used to address multiple comparisons between radiological evaluations.

Results

Patient Sample

Thirty-six patients (median age, 57 years; age range, 25–92 years, 24 men, 12 women) were included in the qualitative assessment. Six patients had post-chemoradiotherapy rectal MRI and were therefore excluded from the quantitative analysis. Thus, for qualitative analysis, a total of 36 patients were included, and for quantitative analysis, a total of 30 patients were included, as shown in Fig. 1 (patient inclusion flowchart).

Qualitative Image Analysis

For both readers, MUSE b800 scored higher in terms of image quality, rectal contour, and conspicuity than ss-EPI (Table 3–4). The average scores between the readers for ss-EPI versus MUSE b800 were as follows: For Reader 1, image quality 3.89 ± 0.52 versus 4.58 ± 0.55 (P = 0.001), rectal contour 3.83 ± 0.64 versus 4.17 ± 0.61 (P = 0.010), and lesion conspicuity 3.17 ± 0.91 versus 3.7 ± 1.09 (P = 0.001). For Reader 2: image quality 3.00 ± 1.04 versus 4.22 ± 0.87, rectal contour 3.17 ± 0.97 versus 4.19 ± 0.75, and lesion conspicuity 2.9 ± 1.16 versus 4.03 ± 0.76 (P < 0.001 for all). The median κ scores between readers were high, between 0.722–0.833.

Table 3.

Qualitative analysis regarding image quality, rectal contour, and lesion conspicuity for the four different DWI sequences

	ss-EPI (b800)			MUSE (b800)			FOCUS (b1500)			MUSE (b1500)
	Reader 1	Reader 2	κ	Reader 1	Reader 2	κ	Reader 1	Reader 2	κ	Reader 1	Reader 2	κ
Image Quality	4 (4–4)	3 (2–4)	0.722	5 (4–5)	4 (4–5)	0.829	3 (3–3)	3 (2.75–4)	0.868	3 (3–4)	4 (3–4)	0.875
	3.89±0.52	3.00±1.04		4.58±0.55	4.22±0.87		3.11±0.40	3.03±0.81		3.36±0.76	3.64±0.99
Rectal Contour	4 (4–3)	3 (2.75–4)	0.792	4 (4–5)	4 (4–5)	0.882	3 (3–4)	3 (3–3)	0.840	4 (3–4)	4 (3–4)	0.868
	3.83±0.64	3.17±0.97		4.17±0.61	4.19±0.75		3.27±0.61	2.92±0.77		3.67±0.83	3.58±0.97
Lesion Conspicuity	3 (3–4)	3 (2–4)	0.833	4 (3–4)	4 (4–4.75)	0.850	4 (3–4)	3 (3–4)	0.833	4 (4–4.75)	4 (3.25–5)	0.850
	3.17±0.91	2.90±1.16		3.70±1.09	4.03±0.76		3.5±0.82	3.20±0.91		4.07±0.74	3.93±1.14

Values are expressed as median (IQR); mean + standard deviation

Abbreviations: MUSE, multiplexed sensitivity-encoding diffusion-weighted imaging; FOCUS, field-of-view optimized and constrained undistorted single-shot; ss-EPI, single-shot echo planar imaging

Table 4.

P values of qualitatively assessed image quality, rectal contour, and lesion conspicuity comparing four different DWI sequences

	ss-EPI (b800) vs. MUSE (b800)		ss-EPI (b800) vs. FOCUS (b1500)		ss-EPI (b800) vs. MUSE (b1500)		MUSE (b800) vs. FOCUS (b1500)		MUSE (b800) vs. MUSE (b1500)		FOCUS (b1500) vs. MUSE (b1500)
	Reader 1	Reader 2	Reader 1	Reader 2	Reader 1	Reader 2	Reader 1	Reader 2	Reader 1	Reader 2	Reader 1	Reader 2
Image Quality	<0.001	<0.001	<0.001	0.913	0.007	0.022	<0.001	<0.001	<0.001	0.016	0.074	0.021
Contour	0.010	<0.001	<0.001	0.208	0.197	0.098	<0.001	<0.001	0.010	0.008	0.017	0.008
Conspicuity	<0.001	<0.001	0.160	0.710	<0.001	0.008	0.374	<0.001	0.160	0.881	0.005	0.009

MUSE b800 had the highest score for image quality and contour, which was significant for both readers, compared to all other acquisitions. Lesion conspicuity was better on MUSE b800 than FOCUS for Reader 2. There was no significant difference between MUSE at different b-values for the assessment of lesion conspicuity (P values 0.16–0.881); however, image quality and rectal wall contour were significantly better on MUSE b800 than MUSE b1500.

There was good to excellent inter-reader agreement for all qualitative features (κ = 0.72–0.88).

Quantitative Image Analysis

The average ADC of the rectal wall was significantly higher for ss-EPI (1.12 ± 0.14) than MUSE b1500 (0.98 ± 0.13) and FOCUS (0.96 ± 0.14) (P < 0.001 for both) (Tables 5 and 6). MUSE b800 had a significantly higher average ADC than FOCUS (1.11± 0.15 versus 0.96 ± 0.14) and MUSE b1500 (1.11± 0.15 versus 0.98 ± 0.13) (P < 0.001 for both).

Table 5.

Quantitative analysis regarding contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR), and average apparent diffusion coefficient (ADC) for the four different DWI sequences

	ss-EPI (b800)	MUSE (b800)	FOCUS (b1500)	MUSE (b1500)
SNR	15.80±6.07	31.00±14.63	18.16±6.71	28.62±10.82
CNR	5.94±3.49	9.27±4.64	7.29±4.80	9.99±6.60
ADC	1.12±0.14	1.11±0.15	0.96±0.14	0.98±0.13

Values are expressed as mean + standard deviation

Abbreviations: ADC, apparent diffusion coefficient; contrast-to-noise ratio (CNR), MUSE, multiplexed sensitivity-encoding diffusion-weighted imaging; FOCUS, field-of-view optimized and constrained undistorted single-shot; signal-to-noise ratio (SNR); ss-EPI, single-shot echo planar imaging

Table 6.

P values of qualitatively assessed contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR), and average apparent diffusion coefficient (ADC) comparing four different DWI sequences

	ss-EPI (b800) vs. MUSE (b800)	ss-EPI (b800) vs. FOCUS (b1500)	ss-EPI (b800) vs. MUSE (b1500)	MUSE (b800) vs. FOCUS (b1500)	MUSE (b800) vs. MUSE (b1500)	FOCUS (b1500) vs. MUSE (b1500)
SNR	<0.001	0.316	<0.001	<0.001	0.477	<0.001
CNR	0.016	0.220	0.022	0.327	0.327	0.302
ADC	0.999	<0.001	<0.001	<0.001	<0.001	0.999

Abbreviations: ADC, apparent diffusion coefficient; contrast-to-noise ratio (CNR), MUSE, multiplexed sensitivity-encoding diffusion-weighted imaging; FOCUS, field-of-view optimized and constrained undistorted single-shot; signal-to-noise ratio (SNR); ss-EPI, single-shot echo planar imaging

MUSE b800 had a higher SNR (31.0± 15.63) in comparison to ss-EPI and FOCUS (15.8 ± 6.07 and 18.16 ± 6.71, respectively) (P < 0.001). No significant difference for SNR was found for MUSE b800 and b1500 (P = 0.477).

The average CNR of MUSE b800 and b1500 were 9.27± 4.64 and 9.99 ± 6.6 in comparison to ss-EPI 5.94 ± 3.49 (P = 0.22 and 0.016). No statistical significance was found in CNR between MUSE b800 and b1500 (P = 0.327) (Figure 4).

Fig. 4.

Contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR), and average apparent diffusion coefficient (ADC) for the four diffusion sequences

DISCUSSION

In this study we investigated the qualitative and quantitative performance of MUSE b800 and MUSE b1500 compared to ss-EPI and FOCUS in rectal cancer. Our results showed that MUSE b800 had superior image quality, rectal wall contour and lesion conspicuity compared to ss-EPI, probably due to improvement in geometric distortion, with higher SNR and CNR and no significantly different average ADC compared to ss-EPI in rectal tumor. Thus, MUSE b800 improved image quality with good intrareader agreement in the qualitative analysis. It should be noted that SNR is reduced with higher b-values. Therefore, whereas comparison between sequences with the same b-value provides insight into SNR trade-offs, there is inherent bias when the comparison is performed between sequences with different b-values.

In rectal cancer, DW images are important for staging and restaging MRI. However, ss-EPI is limited in its spatial resolution [23] and is affected by patient motion and magnetic field inhomogeneity, resulting in artifacts that limit its role in fine structures where high resolution is required [24]. These problems can be improved by using MUSE, which has a higher SNR ratio and higher spatial resolution with no increased acquisition time [9]. MUSE has been applied in breast MRI [10] and was shown to be superior to ss-EPI in improving image quality by decreasing geometric distortion and better visualizing breast lesions, with no significant difference in ADC maps, similar to the results of our study. In brain MRI, MUSE was able to eliminate aliasing artifacts and achieve higher SNR [9], similar to our results, where MUSE images had better image quality and less geometric distortion compared to ss-EPI. Studies of ms-EPI in prostate MRI confirmed its superior image quality with no effect on ADC maps, as compared to ss-EPI [25]. Table 7 provides a summary of some of the applications of MUSE DWI in the literature to date.

Table 7.

Different applications of MUSE DWI in the literature to date

Organ	Aim	Results
Brain [9]	Comparison between MUSE and SENSE sequences in the normal brain	Motion-induced aliasing artifacts can be removed using MUSE with higher SNR in comparison with SENSE images
Breast [10]	Comparison between MUSE and single-shot DWI for the visualization of lesions and for the differentiation of malignant and benign lesions	MUSE showed better image quality compared to single-shot DWI and better visibility of lesions. MUSE DWI ADC values showed a significant difference between malignant and benign breast lesions.
Prostate [25]	Reproducibility of quantitative diffusion measurements between ssEPI, rFOV, and multishot EPI (msEPI) in phantoms, healthy volunteers, and patients	No significant difference in ADC was found between the 3 pulse sequences in phantoms, healthy volunteers, and patients. msEPI has high resolution with less distortion compared to ssEPI.
Small bowel [12]	Comparison between ssEPI, high resolution ssEPI (HRssEPI), and MUSE for the assessment of bowel inflammation in Crohn’s disease using enhanced MRI as reference	MUSE had significantly better image quality, less geometric distortion, and better tissue texture conspicuity compared to the other 2 sequences. It showed as well higher sensitivity and accuracy than ssEPI for detecting inflammation.

Abbreviations: ADC, apparent diffusion coefficient; MUSE, multiplexed sensitivity-encoding diffusion-weighted imaging; rFOV, reduced field of view; SENSE, sensitivity encoding; ss-EPI, single-shot echo planar imaging

Our study is the first to compare MUSE b800 and b1500 to ss-EPI and FOCUS sequences in rectal MRI for rectal cancer patients. Our results showed the superior quality, rectal wall contour visualization, and lesion conspicuity of MUSE b800 compared to ss-EPI, which can be attributed to improved geometric distortion.

Diffusion plays an important role in staging rectal cancer and an even more crucial role in the restaging of rectal cancer. The addition of DWI to T2WI in restaging rectal cancer increases sensitivity from 50.4% to 83.6% with no significant change in specificity [26]. A study by Gollub et al [27] showed that in restaging MRI, positive DWI findings are concordant with endoscopy and can show tumor regrowth, even prior to endoscopy, highlighting the importance of obtaining good DWI image quality.

A small study by Bates et al. [2] compared the sensitivity and specificity of b800 and FOCUS sequences in post-treatment MRI with four readers. The study showed no significant difference between readers, except one who had higher sensitivity with b1500. In our study, we obtained similar results, with no significant difference in lesion conspicuity between ss-EPI and FOCUS; image quality and rectal wall contour were significantly better for only one of the readers.

Our study has multiple limitations. First, the retrospective nature of our study may have resulted in a selection bias with regard to patients. Second, the number of patients included was relatively small and may not be generalizable at this time. Third, readers were not completely blinded to which sequence they’re analyzing, as FOCUS 1500 has a small field-of-view compared to the other sequences. Finally, all MRIs were performed on 3 T using the same vendor; future studies should address the reproducibility of our results at different field strengths and using different coils and scanners from different vendors.

Conclusion

MUSE b800 has better image quality, rectal wall contour, and lesion conspicuity in rectal cancer compared to the traditional ss-EPI, with no significant change in ADC values for both readers. MUSE b1500 has same lesion conspicuity as MUSE b800.

Data availability:

The datasets used and analyzed in this study are not publicly available due to patient privacy requirements but are available upon reasonable request from the corresponding author.