Try a new search

Format these results:

Searched for:

person:shuklm01

in-biosketch:true

Total Results:

2


Deep Learning Diagnosis and Classification of Rotator Cuff Tears on Shoulder MRI

Lin, Dana J; Schwier, Michael; Geiger, Bernhard; Raithel, Esther; von Busch, Heinrich; Fritz, Jan; Kline, Mitchell; Brooks, Michael; Dunham, Kevin; Shukla, Mehool; Alaia, Erin F; Samim, Mohammad; Joshi, Vivek; Walter, William R; Ellermann, Jutta M; Ilaslan, Hakan; Rubin, David; Winalski, Carl S; Recht, Michael P
BACKGROUND:Detection of rotator cuff tears, a common cause of shoulder disability, can be time-consuming and subject to reader variability. Deep learning (DL) has the potential to increase radiologist accuracy and consistency. PURPOSE:The aim of this study was to develop a prototype DL model for detection and classification of rotator cuff tears on shoulder magnetic resonance imaging into no tear, partial-thickness tear, or full-thickness tear. MATERIALS AND METHODS:This Health Insurance Portability and Accountability Act-compliant, institutional review board-approved study included a total of 11,925 noncontrast shoulder magnetic resonance imaging scans from 2 institutions, with 11,405 for development and 520 dedicated for final testing. A DL ensemble algorithm was developed that used 4 series as input from each examination: fluid-sensitive sequences in 3 planes and a sagittal oblique T1-weighted sequence. Radiology reports served as ground truth for training with categories of no tear, partial tear, or full-thickness tear. A multireader study was conducted for the test set ground truth, which was determined by the majority vote of 3 readers per case. The ensemble comprised 4 parallel 3D ResNet50 convolutional neural network architectures trained via transfer learning and then adapted to the targeted domain. The final tear-type prediction was determined as the class with the highest probability, after averaging the class probabilities of the 4 individual models. RESULTS:The AUC overall for supraspinatus, infraspinatus, and subscapularis tendon tears was 0.93, 0.89, and 0.90, respectively. The model performed best for full-thickness supraspinatus, infraspinatus, and subscapularis tears with AUCs of 0.98, 0.99, and 0.95, respectively. Multisequence input demonstrated higher AUCs than single-sequence input for infraspinatus and subscapularis tendon tears, whereas coronal oblique fluid-sensitive and multisequence input showed similar AUCs for supraspinatus tendon tears. Model accuracy for tear types and overall accuracy were similar to that of the clinical readers. CONCLUSIONS:Deep learning diagnosis of rotator cuff tears is feasible with excellent diagnostic performance, particularly for full-thickness tears, with model accuracy similar to subspecialty-trained musculoskeletal radiologists.
PMID: 36728041
ISSN: 1536-0210
CID: 5502202

Ultrasound evaluation of the ulnar collateral ligament of the elbow: Which method is most reproducible?

Shukla, Mehool; Keller, Robert; Marshall, Nathan; Ahmed, Hafeez; Scher, Courtney; Moutzouros, Vasilios Bill; van Holsbeeck, Marnix
INTRODUCTION/BACKGROUND:The ulnar collateral ligament (UCL) is an important medial stabilizer of the elbow, particularly in overhead-throwing athletes. However, there is no universally accepted method for evaluating UCL thickness with ultrasound (US). OBJECTIVE:To assess reproducibility of previously published methods, as well as a modified technique, for evaluating the UCL via US. We hypothesize that a modified technique would show greater reproducibility. MATERIAL AND METHODS/METHODS:Using US, the thickness of the UCL in 50 volunteers was measured by two musculoskeletal trained radiologists using two different measurement techniques. The techniques utilized were as described by Nazarian and Jacobson/Ward (JW). Technique measurements were evaluated using interclass correlation coefficients (ICC) to determine the reproducibility of each method. Twenty-eight of the subjects also underwent measurement via a modified JW technique, measured perpendicular to the ligament rather than the frame of imaging. This technique was also evaluated with ICC values. RESULTS:The ICC value for the Nazarian technique was 0.82 (very good) and 0.51 (moderate) for the JW technique. When using the modified JW technique, we found an ICC value of 0.84 (very good). Mean ligament thickness was greatest with the Nazarian technique, 6.41 mm, with the JW technique measuring 1.86 mm and the modified technique measuring 1.38 mm. CONCLUSION/CONCLUSIONS:US assessment of UCL thickness by all three measurement techniques are reproducible. The JW technique had less interobserver agreement when compared to the Nazarian method, whereas the modified JW technique had greater reproducibility compared to the JW technique and similar to the Nazarian technique.
PMID: 28424849
ISSN: 1432-2161
CID: 3826332