Effect of Tear Classification on Subscapularis Muscle Volume: A Deep Learning-based Semi-automatic Analysis of Pre- and Postoperative Changes in 246 Rotator Cuff Repair Patients With and Without First Facet Subscapularis Tears.

Background: Partial-thickness subscapularis (SSC) tendon tears, although less studied than full-thickness tears, are prevalent and clinically significant. Recent classifications have proposed subdivisions for these partial-thickness tears. However, their clinical implications and correlation with SSC muscle volume remain poorly understood. Purpose: To evaluate SSC muscle volume in relation to tendon tear classification and assess the changes in muscle volume both pre- and postoperatively using deep learning-based magnetic resonance imaging (MRI) segmentation in patients undergoing arthroscopic rotator cuff repair. Study Design: Cohort study; Level of evidence, 3. Methods: This study included 246 patients who underwent arthroscopic rotator cuff repair between January 2018 and December 2023. SSC tendon tears were classified using the Yoo and Rhee system. SSC muscle volumes were measured on preoperative MRIs and 6 months postoperatively using a validated deep learning segmentation tool. These volumes were normalized to the scapular volume. Results: Preoperative normalized SSC volume (nSSC) varied significantly across tear types (F = 7.21; Conclusion: Even in partial-thickness tears affecting only the first facet, nSSC muscle volume decreases significantly with increasing tear severity. Furthermore, SSC muscle volume significantly decreases 6 months postoperatively in patients with SSC tendon tears, although the extent of this reduction does not vary among different treatment modalities.