Assessment of the supraspinatus muscle fiber architecture with diffusion tensor imaging in healthy volunteers

Objectives: This study presents a framework for the calculation of supraspinatus (SSP) muscle pennation angles (PAs) from diffusion tensor imaging (DTI).

Materials and methods: Ten healthy individuals (five females and five males; age 32.0 ± 4.7 years) underwent three sessions of 3-T MRI, including a stimulated echo acquisition mode DTI sequence. The imaging plane of the DTI sequence was angled along the intramuscular part of the SSP tendon. A custom-built software was developed and implemented to compute DTI-based PAs of the anterior and posterior SSP in relation to the orientation of the tendon. Subsequently, three readers measured PAs from the post-processed images. Test-retest reliability, inter-reader agreement, and intra-reader agreement of PA measurements were evaluated with intraclass correlation coefficients (ICCs).

Results: The mean PA in the anterior SSP was 15.6 ± 2.1° and 10.7 ± 0.9° in the posterior SSP. MRI-derived PAs showed good to excellent test-retest reliability (ICC: 0.856-0.945), inter-reader agreement (ICC: 0.863-0.955), and intra-reader agreement (ICC: 0.804-0.955).

Conclusion: PAs derived from DTI demonstrated good to excellent test-retest reliability, inter-reader agreement, and intra-reader agreement. We successfully implemented a highly standardized technique for evaluating PAs of the SSP muscle.

Critical relevance statement: This proposed low-complex method might facilitate the increased use of the PA as a biomarker for pathological conditions of the rotator cuff.

Key points: A low-complex method for measuring PAs of the SSP might help identify pathology early. The mean PA was 15.6 ± 2.1° and 10.7 ± 0.9° in the anterior and posterior SSP, respectively. ICCs were ≥ 0.856 for test-retest reliability, ≥ 0.863 for inter-reader agreement, and ≥ 0.804 for intra-reader agreement.