3D quantitative CT study to assess rotator cuff muscle fatty infiltration

Keywords:

Fatty infiltration, rotator cuff, 3D CT scan, Goutallier


Published online: Oct 21 2024

https://doi.org/10.52628/90.2.12827

L. DERAEDT1, C. DIEPENDAELE1, D. CARDON1, A. JALALIJAM1, L. DE WILDE1,2, A. VAN TONGEL1,2

1 Ghent University, De Pintelaan 185, 9000 Gent, Belgium
2 Department of Orthopaedic Surgery and Traumatology, Ghent University Hospital, Corneel Heymanslaan 10, 9000 Gent, Belgium

Abstract

Fatty infiltration is a degenerative condition of the tendon-muscle unit of rotator cuff muscles, characterized by fatty accumulation within and around the muscles. This study compares a 3D method to the Goutallier classification for evaluating fatty infiltration of the rotator cuff muscles. Employing a retrospective study design, four researchers performed 3D segmentation of the rotator cuff muscles on 65 CT scans of patients with rotator cuff arthropathy. The Goutallier classification was graded on screenshots of the Y view and two axial slices. Measurements, including HU, volume, cross-sectional area, and Goutallier grade, were performed on the 3D segmentations and the respective 2D slices. The inter- and intra-rater variability for 2D and 3D methods were calculated using the Intraclass Correlation Coefficient (ICC). The ICC for 2D and 3D methods was excellent (ICC: 0.90-0.95 and 0.81-0.99, respectively) and good for the Goutallier classification (ICC: 0.62-0.81). Overall, the relative 3D fatty infiltration was 0% for Goutallier grade 0 muscles, 7% for grade 1, 19% for grade 2, 33% for grade 3, and 37% for grade 4. The relative 2D fatty infiltration was 0%, 8%, 25%, 37%, and 43%, respectively. We conclude that 3D segmentation on CT scans is better reproducible and evaluates the muscle entirely. However, the time-intensive nature of the 3D method currently limits its clinical practicality. Quantitive 2D evaluation is excellently reproducible but may overestimate the actual fatty infiltration percentage of the whole muscle.