Quantitative MRI of visually intact rotator cuff muscles by multiecho Dixon-based fat quantification and diffusion tensor imaging.

BACKGROUND: Quantitative MRI allows assessment of shoulder rotator cuff (RC) muscles by Dixon MR sequences with calculation of fractional fat content (FF%) maps and diffusion tensor imaging (DTI) including tractography. PURPOSE: To compare FF% and DTI derived parameters among visually intact RC muscles, to compare 2D with 3D DTI measurements and to establish normative values. STUDY TYPE: Prospective. SUBJECTS: Forty patients aged >18 years undergoing shoulder MR arthrography were included. FIELD STRENGTH/SEQUENCE: MR arthrography of the shoulder including 3D multiecho Dixon and 3D echo-planar DTI sequences (15 gradient encoding directions, b-value 600 s/mm2 ) was performed at 3.0T. ASSESSMENT: Muscles affected by RC tears or fatty infiltration of Goutallier grade ≥1 were excluded. Two independent radiologists measured FF%, apparent diffusion coefficient (ADC), and fractional anisotropy (FA) by region-of-interest (ROI) placements at the Y-position of the scapula and 3D tractography of each muscle with qualitative evaluation was performed. STATISTICAL TESTS: Intraclass correlation coefficients (ICCs) and Cohen's kappa were used for interreader agreement and Pearson correlation coefficient to correlate quantitative measures with each other and age, independent-samples t-test, one-way analysis of variance (ANOVA), and Kruskal-Wallis test were performed to investigate differences between genders and muscles. RESULTS: Qualitative and quantitative measurements showed moderate (κ = 0.41-0.56) to almost perfect (ICC = 0.75-0.99) agreement. There were weak but significant positive correlations of FF% with age (r = 0.273, P < 0.05) and FA-2D (r = 0.319-0.383, P < 0.05). Significant differences were found among RC muscles for ADC, radial diffusivity (RD), and tract homogeneity (all P < 0.05) but not between genders (all P ≥ 0.05). High correlations of 2D with 3D measurements for ADC (r = 0.639, P < 0.001) and FA (r = 0.628, P < 0.001) were seen. DATA CONCLUSION: Quantitative MRI with estimation of FF% and DTI parameters shows significant age-associated changes and differences among visually intact RC muscles. High reproducibility and correlations of 2D with 3D DTI measurements can be expected. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;49:109-117.

Shoulder muscle volume and fat content in healthy adult volunteers: quantification with DIXON MRI to determine the influence of demographics and handedness.

OBJECTIVE: We aimed to provide mean values for fat-fraction and volume for full-length bilateral rotator cuff and deltoid muscles in asymptomatic adults selected on the basis of their good musculoskeletal and systemic health, and to understand the influence of gender, age, and arm dominance. MATERIALS AND METHODS: Seventy-six volunteers aged 20 to 60 years who were screened for normal BMI and high general health were included in the study. MRI was performed at 3 Tesla using three-point DIXON sequences. Volume and fat-signal fraction of the rotator cuff muscles and the deltoid muscle were determined with semi-automated segmentation of entire muscle lengths. Differences according to age, gender, and handedness per muscle were evaluated. RESULTS: Fat-signal fractions were comparable between genders (mean ± 2 SD, 95% CI, women 7.0 ± 3.0; 6.8-7.2%, men 6.8 ± 2.7; 6.7-7.0%) but did not show convincing changes with age. Higher shoulder muscle volume and lower fat-signal fraction in the dominant arm were shown for teres minor and deltoid (p < 0.01) with similar trends shown for the other rotator cuff muscles. CONCLUSIONS: Bilateral fat-signal fractions and volumes based on entire length shoulder muscles in asymptomatic 20-60 year old adults may provide reference for clinicians. Differences shown according to arm dominance should be considered and may rationalize the need for bilateral imaging in determining appropriate management.