Application of deep learning–based image reconstruction in MR imaging of the shoulder joint to improve image quality and reduce scan time

Abstract

Objectives To compare the image quality and diagnostic performance of conventional motion-corrected periodically rotated overlapping parallel line with enhanced reconstruction (PROPELLER) MRI sequences with post-processed PROPELLER MRI sequences using deep learning-based (DL) reconstructions. Methods In this prospective study of 30 patients, conventional (19 min 18 s) and accelerated MRI sequences (7 min 16 s) using the PROPELLER technique were acquired. Accelerated sequences were post-processed using DL. The image quality and diagnostic confidence were qualitatively assessed by 2 readers using a 5-point Likert scale. Analysis of the pathological findings of cartilage, rotator cuff tendons and muscles, glenoid labrum and subacromial bursa was performed. Inter-reader agreement was calculated using Cohen’s kappa statistic. Quantitative evaluation of image quality was measured using the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). Results Mean image quality and diagnostic confidence in evaluation of all shoulder structures were higher in DL sequences (p value = 0.01). Inter-reader agreement ranged between kappa values of 0.155 (assessment of the bursa) and 0.947 (assessment of the rotator cuff muscles). In 17 cases, thickening of the subacromial bursa of more than 2 mm was only visible in DL sequences. The pathologies of the other structures could be properly evaluated by conventional and DL sequences. Mean SNR (p value = 0.01) and CNR (p value = 0.02) were significantly higher for DL sequences. Conclusions The accelerated PROPELLER sequences with DL post-processing showed superior image quality and higher diagnostic confidence compared to the conventional PROPELLER sequences. Subacromial bursa can be thoroughly assessed in DL sequences, while the other structures of the shoulder joint can be assessed in conventional and DL sequences with a good agreement between sequences. Key Points • MRI of the shoulder requires long scan times and can be hampered by motion artifacts. • Deep learning–based convolutional neural networks are used to reduce image noise and scan time while maintaining optimal image quality. The radial k-space acquisition technique (PROPELLER) can reduce the scan time and has potential to reduce motion artifacts. • DL sequences show a higher diagnostic confidence than conventional sequences and therefore are preferred for assessment of the subacromial bursa, while conventional and DL sequences show comparable performance in the evaluation of the shoulder joint.

Comparison of three dimensional gradient-echo sequences with conventional two dimensional sequences to assess the image quality of shoulder joint in magnetic resonance imaging

“…High-eld imaging at 3 tesla (3 T) has been shown to outperform 1.5 T for articular joint imaging. In particular, TRUEFISP at 3 T offered excellent cartilage-joint uid contrast in previous investigations and that cartilage was also better visualized with TRUEFISP than SPACE sequences [9,28,29].…”

“…Meanwhile, 3D-GRE sequences showed moderate reproducibility for glenohumeral joint space measurement, while 2D sequences displaying varying reproducibility results. The reasonable answer to slight agreement for T 2 w FSE could be the motion artifacts due to the heavy breathing of volunteers or the pulsation of the adjacent vessels lead to lower image quality of this technique [9]. Furthermore, the later echoes in longer echo train length might also be associated with a decrease in the overall SNR, CNR, and more spatial blurring [7,31].…”

“…Currently 3D-FSE and 3D gradient echo (3D-GRE) sequences are used for MSK imaging. 3D-FSE sequences (e.g., 3D-FLAIR, 3D-T2, 3D-SPACE) deliver submillimeter isotropic resolution while retaining the attributes of FSE sequences [9], but their drawback lies in lengthy scan times, typically around 10 minutes depending on parameters [2]. This impedes their routine clinical use.…”

Application of Artificial Intelligence in Shoulder Pathology

“…The image quality satisfaction survey among 400 patients received high scores in DL-MRI from all radiologists. Kaniewska et al [43] also declared that DLR could improve diagnostic accuracy and image quality with thorough assessment of the subacromial bursa and good agreement for other shoulder structures.…”

Deep learning-based reconstruction enhances image quality and improves diagnosis in magnetic resonance imaging of the shoulder joint

“…In a previous study, DL-MRI sequences were found to be more effective than traditional MRI sequences in detecting SbA bursal thickening, although the extent of this thickening was not extensively explored ( 19 ). Assessing the degree of thickness in the intra-articular bursa can provide valuable insights into the severity of the patient’s condition, as it often correlates with the level of shoulder pain experienced.…”