Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Oct 14;17(10):e94561.
doi: 10.7759/cureus.94561. eCollection 2025 Oct.

Examining the Effect of Deep Learning-Based Image Reconstruction on Accelerating Shoulder Magnetic Resonance Imaging (MRI) and Its Impact on Image Quality

Jordan Zheng Ting Sim  1 , Alexander Jiawei Yap  2 , Yong Han Ting  3 , Shu Wen Goh  1 , Alvin Yong Quan Soon  1 , Minju Cho  4 , Sohyun Kim  4 , Glen Chern Yue Ong  1
Affiliations

Examining the Effect of Deep Learning-Based Image Reconstruction on Accelerating Shoulder Magnetic Resonance Imaging (MRI) and Its Impact on Image Quality

Jordan Zheng Ting Sim et al. Cureus. .

Abstract

Background Prolonged scan time remains the main obstacle to increasing magnetic resonance imaging (MRI) throughput. The advent of artificial intelligence brings forth opportunities to accelerate MRI examinations. Purpose This study compares the image quality of standard MRI versus accelerated MRI with deep learning-based image reconstruction (DLR) for shoulder MRI studies. Materials and methods Forty-nine subjects were prospectively enrolled and underwent both standard and accelerated axial proton density fat-saturated (PD FS) shoulder MRIs using a 1.5T scanner (Philips Ingenia 1.5T). Two blinded musculoskeletal radiologists independently evaluated paired datasets to assess the anatomic conspicuity of specific structures (labrum, rotator cuff footprint, cartilage, long head of the biceps tendon/rotator interval), artifacts, and overall image quality. A 5-point scale was employed, where 1 indicated the standard MRI was markedly superior and 5 indicated the accelerated MRI was markedly superior. The reduction in scan time was recorded; inter-reader variability was also analyzed. Results The DLR protocol reduced scan duration by 20.2% on average, shortening acquisition time from 184 seconds to 148 seconds. Mean scores for anatomic conspicuity ranged from 3.0 to 3.2, and mean scores for artifacts and overall image quality were 3.0 and 3.2, respectively. The Wilcoxon signed-rank test revealed statistically significant differences (p<0.001) for most categories, except for "Artifacts" as assessed by one reader. Inter-reader agreement was poor, with Cohen's kappa ranging from 0.086 to 0.183 and prevalence-adjusted bias-adjusted kappa (PABAK) scores ranging from 0.063 to 0.404. Conclusion DLR-based acceleration significantly reduces scan time while maintaining diagnostic image quality, presenting a clinically feasible and efficient solution for routine shoulder MRI.

Keywords: artificial intelligence; deep learning; image quality; image reconstruction; shoulder mri.

PubMed Disclaimer

Conflict of interest statement

Human subjects: Informed consent for treatment and open access publication was obtained or waived by all participants in this study. National Healthcare Group Domain Specific Review Board (DSRB) issued approval 2023/00390-AMD0001. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue. Conflicts of interest: In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.

Figures

Figure 1
Figure 1. DLR-processed sequence (B) was scored as notably superior to the conventional sequence (A) by both readers. DLR-processed sequence (B) shows enhanced sharpness of the cartilage and labral outlines (straight arrows) while maintaining conspicuity of mild posterior labral fraying and subscapularis tendinosis (curved arrows)
DLR: deep learning reconstruction
Figure 2
Figure 2. DLR-processed sequence (B) was scored as mildly superior to the conventional sequence (A) by both readers. Both sequences adequately depict a non-displaced anteroinferior labral tear (straight arrows) as well as glenohumeral cartilage wear (curved arrows), but the DLR-processed sequence exhibited improved sharpness and reduced noise
DLR: deep learning reconstruction
Figure 3
Figure 3. Conventional sequence (A) and DLR-processed sequence (B). Both readers indicated no perceivable difference. Findings of joint synovitis, subacromial subdeltoid bursitis, and severe subscapularis tendinosis are equally well-visualized on both sequences
DLR: deep learning reconstruction
Figure 4
Figure 4. Score distributions for each image quality criterion and reader. Scores range from 1 to 5, indicating varying levels of image quality perception from the original image being superior to the DLR image being superior. Both readers predominantly rated scores of 3-4, with overall image quality and artifact reduction most frequently favoring the DLR images
DLR: deep learning reconstruction
Figure 5
Figure 5. Wilcoxon test (reference score=3) comparing image quality scores from two radiologists (A.Y.J. and Y.H.). The chart presents mean scores across specific criteria, with asterisks (*) indicating significant differences from the reference (p<0.05). Scores above 3 reflect superior quality of DLR images. All mean scores exceeded 3, with all but one reaching statistical significance, most notably for the labrum and overall image quality
DLR: deep learning reconstruction
See this image and copyright information in PMC

References

    1. Artificial intelligence in radiology. Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts HJ. Nat Rev Cancer. 2018;18:500–510. - PMC - PubMed
    1. Diagnostic performance of a deep learning model deployed at a national COVID-19 screening facility for detection of pneumonia on frontal chest radiographs. Sim JZ, Ting YH, Tang Y, et al. Healthcare (Basel) 2022;10:175. - PMC - PubMed
    1. The impact of large language models on radiology: a guide for radiologists on the latest innovations in AI. Nakaura T, Ito R, Ueda D, et al. Jpn J Radiol. 2024;42:685–696. - PMC - PubMed
    1. Clinical impact of deep learning reconstruction in MRI. Kiryu S, Akai H, Yasaka K, et al. Radiographics. 2023;43:0. - PubMed
    1. All-in-one deep learning framework for MR image reconstruction. Jeong G, Kim H, Yang J, Jang K, Kim J. https://arxiv.org/abs/2405.03684 arXiv. 2024

LinkOut - more resources