Utility of Thin-slice Single-shot T2-weighted MR Imaging with Deep Learning Reconstruction as a Protocol for Evaluating Pancreatic Cystic Lesions

Q: What is the purpose of this study?

To determine the effect of industry-developed DLR-SR on single-shot turbo spin-echo T2-weighted images when using 2 mm thickness compared with 5 mm thickness in pancreatic cystic lesion patients.

Q: How were images acquired and evaluated?

Thirty consecutive patients underwent abdominal MRI between June and July 2024. Image quality was assessed both qualitatively and quantitatively for SshTSE2mm and SshTSE5mm with and without DLR-SR.

Q: What were the key imaging quality findings?

SNRs and CNRs did not decrease across thickness and reconstruction settings. SshTSE2mm with DLR-SR achieved the best performance for edge sharpness, coarseness perception, pancreatic duct clarity, noise and artifacts, overall image quality, and diagnostic confidence (P < 0.0001).

Q: Are thin-slice SshTSE images with DLR-SR considered clinically useful?

Yes. The study concludes that SshTSE2mm with DLR-SR provides better image quality than other configurations and is feasible and clinically useful for evaluating pancreatic cystic lesions.

Sep 29, 2026

Uploaded by Elsa

Objective: assess how industry-developed deep learning reconstruction with super resolution (DLR-SR) affects single-shot turbo spin-echo (SshTSE) images acquired with 2 mm thickness (DLR-SshTSE2mm) versus 5 mm thickness (DLR-SshTSE5mm) in patients with pancreatic cystic lesions. Methods: evaluate 30 consecutive observed patients undergoing abdominal MRI between June and July 2024, comparing image qualities qualitatively and quantitatively with and without DLR-SR. Results: pancreas-related SNR and CNR for pancreas-to-lesion and fat remain stable across thickness and reconstruction; SshTSE2mm with DLR-SR yields the highest sharpness, duct clarity, lower coarseness, fewer artifacts, overall quality, and diagnostic confidence (P < 0.0001). Conclusion: thin-slice SshTSE with DLR-SR is feasible and clinically useful without degrading SNR or CNR.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages15

Reading Time38 min

AccessFree to download

Introduction

Epidemiology and clinical need for pancreatic cystic lesion surveillance

MRI and T2-weighted imaging role

SshTSE challenges with slice thickness

Deep learning reconstruction rationale

Magn Reson Med Sci 2025 ; 24: 2025-0042 [doi:10.2463/mrms.mp.2025-0042 Published Online: June 14](doi:10.2463/mrms.mp.2025-0042 Published Online: June 14), 2025 MAJOR PAPER Utility of Thin-slice Single-shot T2-weighted MR Imaging with Deep Learning Reconstruction as a Protocol for Evaluating Pancreatic Cystic Lesions Kumi Ozaki1* , Hanae Hasegawa1, Jihun Kwon2, Yasutomo Katsumata2, Masami Yoneyama2, Shota Ishida3, Takafumi Iyoda1, Masataka Sakamoto4, Shuuhei Aramaki5,6, Yukichi Tanahashi1, and Satoshi Goshima1 Purpose: To assess the effects of industry-developed deep learning reconstruction with super resolution (DLRSR) on single-shot turbo spin-echo (SshTSE) images with thickness of 2 mm with DLR (SshTSE2mm)relative to those of images with a thickness of 5 mm with DLR (SSshTSE5mm) in the patients with pancreatic cystic lesions. Methods: Thirty consecutive patients who underwent abdominal MRI examinations because of pancreatic cystic lesions under observation between June 2024 and July 2024 were enrolled. We qualitatively and quantitatively evaluated the image qualities of SshTSE2mm and SshTSE5mm with and without DLR-SR. Results: The SNRs ofthe pancreas, spleen, paraspinal muscle, peripancreatic fat, and pancreatic cystic lesions of SshTSE2mm with and without DLR-SR did not decrease in compared to that ofSshTSE5mm with and without DLRSR. There were no significant differences in contrast-to-noise ratios(CNRs)ofthe pancreas-to-cystic lesions and fat between 4 types of images. SshTSE2mm with DLR-SR had the highest image quality related to pancreas edge sharpness, perceived coarseness pancreatic duct clarity, noise, artifacts, overall image quality, and diagnostic confidence of cystic lesions, followed by SshTSE2mm without DLR-SR and SshTSE5mm with and without DLRSR (P <0.0001). Conclusions: SshTSE2mm with DLR-SR images had better quality than the other images and did not have decreased SNRs and CNRs. The thin-slice SshTSE with DLR-SR maybe feasible and clinically useful for the evaluation of patients with pancreatic cystic lesions. Keywords: deep learning reconstruction, magnetic resonance imaging, pancreatic cystic lesion, single-shot turbo spin echo, thin-slice 1Department of Radiology, Hamamatsu University School of Medicine, Hamamatsu, Shizuoka, Japan 2Philips Japan, Tokyo, Japan 3Department of Radiological Technology, Faculty of Medical Science, Kyoto College of Medical Science, Nantan, Kyoto, Japan 4Department of Radiology, Hamamatsu University Hospital, Hamamatsu, Shizuoka, Japan 5Radiation Oncology, Hamamatsu University School of Medicine, Hamamatsu, Shizuoka, Japan 6Startup Support and URA Office, Institute of Photonics Medicine, Hamamatsu University School of Medicine, Hamamatsu, Shizuoka, Japan *Corresponding author: Department of Radiology, Hamamatsu University School of Medicine, 1-20-1, Handayama, Chuo-ku, Hamamatsu, Shizuoka 431-3192, Japan. Phone: +81-53-435-2242, Fax: +81-53-435-2241, E-mail: [ozakik-rad@umin.org](ozakik-rad@umin.org) This work is licensed under a Creative Commons Attribution 4 .0 International License. ©2025 Japanese Society for Magnetic Resonance in Medicine Received: March 19, 2025 | Accepted: April 11, 2025 Introduction Pancreatic cystic lesions are found in approximately 2.5% of asymptomatic individuals, with their prevalence increasing to 10% in those aged 70 years or older, as demonstrated by computed tomography (CT) and MRI examinations.1,2 Their detection rate has increased significantly due to increased use of abdominal imaging, aging populations, and improved imaging sensitivity.2–4 The majority of neoplastic pancreatic cystic lesions consist of 3 main types: intraductal papillary mucinous neoplasms (IPMNs), mucinous cystic neoplasms (MCN), and serous cystic neoplasms (SCN), with IPMNs comprising over half of imaging-detected cases.5,6 Given their varying malignant potential, accurate diagnosis and radiological monitoring are essential for ea

FAQs

What is the purpose of this study?

To determine the effect of industry-developed DLR-SR on single-shot turbo spin-echo T2-weighted images when using 2 mm thickness compared with 5 mm thickness in pancreatic cystic lesion patients.

How were images acquired and evaluated?

Thirty consecutive patients underwent abdominal MRI between June and July 2024. Image quality was assessed both qualitatively and quantitatively for SshTSE2mm and SshTSE5mm with and without DLR-SR.

What were the key imaging quality findings?

SNRs and CNRs did not decrease across thickness and reconstruction settings. SshTSE2mm with DLR-SR achieved the best performance for edge sharpness, coarseness perception, pancreatic duct clarity, noise and artifacts, overall image quality, and diagnostic confidence (P < 0.0001).

Are thin-slice SshTSE images with DLR-SR considered clinically useful?

Yes. The study concludes that SshTSE2mm with DLR-SR provides better image quality than other configurations and is feasible and clinically useful for evaluating pancreatic cystic lesions.

Sep 29, 2026

Uploaded by Elsa