Study population
Our institutional review board approved this retrospective study, and the requirement for informed consent was waived for the patients group. All patients who underwent shoulder MR examination suspicious for RCT were retrospectively reviewed, and inclusion criteria were any positive findings on MR images and physical examinations. In addition, healthy volunteers without any shoulder symptoms or positive findings of shoulder disorders were recruited prospectively as control subjects, and written informed consent was obtained. A musculoskeletal radiologist and an orthopaedic surgeon confirmed the eligibility.
Consequently, 13 patients (4 males and 9 females; age, 47–82 years; mean age, 65.3 years) and 20 controls (10 males and 10 females; 21–24 years; 22.0 years) satisfied eligibility criteria. Among the patients, 4 patients had a complete tear, and 9 patients had a partial tear.
MR examination
MR examination was performed using a 3.0-T MR system with a shoulder coil. The symptomatic shoulders of patients and right shoulders of the controls were examined. Both routine clinical and MDME sequences were acquired. The MDME sequence is a quantification sequence based on spin-echo acquisition, and T1 and T2 values were measured independently using four saturation delay times and two echo times [14]. The acquisition time of the MDME sequence was 4 min 59 s. The MR imaging parameters are summarized in Table 1.
MR image analysis
Proton density-weighted images (PDWI) and T2-weighted short-tau inversion-recovery (T2-STIR) images were generated for T1 and T2 measurements using the SyMRI software (SyntheticMR AB). The quantitative values were directly measured on synthetic PDWI, which contained all quantitative information of T1, T2, and PD values. Two radiological technologists measured T1 and T2 values of the supraspinatus tendon, supraspinatus muscle, and glenohumeral cartilage. Slices with the largest anatomical structures were selected, and the region of interest (ROI) was manually drawn on the border of the structure to avoid synovial fluid or other tissues, referring to synthetic T2-STIR (Fig. 1a). For patients with complete tears, the ROI was placed in the intact region to measure the supraspinatus tendon (Fig. 1b). When it was difficult to measure a specific structure due to muscle atrophy or cartilage defects, the structure was excluded from the analysis. The measurements were conducted independently by the two raters, and one of them measured twice at least two-week intervals. Thereafter, the quantitative values of the three measurements were averaged.
A musculoskeletal radiologist measured the maximum ratio of the torn tendon thickness (tear size ratio), which correlates with tear progression [15, 16] on conventional coronal T2-weighted images with fat suppression (FS-T2WI). It is noted that the tear size ratio of the complete tear was set to 1.0.
Statistical analysis
The Mann-Whitney U test was performed to compare the T1 and T2 values between the two groups. The intra- and inter-rater reliabilities of quantitative measurements were assessed using the intra-class correlation coefficient (ICC), and ICCs greater than 0.80 were considered almost perfect reliability [17]. For the patient group, Spearman’s correlation coefficients (r) were calculated to analyze the association between the tear size ratio and quantitative values in each structure. Correlation coefficients greater than 0.70 were considered strong correlation [18]. Statistical significance was set at p < 0.05.