We utilized a Retrospective chart review to conduct our research. After obtaining our IRB approval we gathered all shoulder and upper arm MRIs from the radiology department and available in PACS (picture archiving and communication system) at a tertiary center between June 2015 until February of 2021, that amounted to 2779 shoulder MRI images. An exclusion criterion was applied to the sample which included: patients younger than 24 (as acromion of patients younger than 24 years is physiologically unfused), non-Saudis, repeated examination of the same shoulder and if the examination showed inadequate imaging. After applying the criteria, we excluded 356 images, and we performed systematic random sampling to reach our target sample size of 751 shoulder MRI studies.
All MRI studies were inspected and evaluated bymusculoskeletal radiologists, holding a professor degree.We looked for OA and categorized it according to the side (right, left), the type (pre-acromion, meso-acromion and meta-acromion) and the size (length, width, and thickness). Patients with OA were further evaluated for the status of the rotator cuff and the type of pathology (Tendinopathic changes, partial tear,full-thicknesstear).
In the included sample, patients were scanned in supine position. The scanning protocol included axial, sagittal and coronal imaging. All patients were scanned on Magnetom Skyra fit 3T (Siemens Healthineers, Erlangen, Germany) with axial Proton Density (PD) weighted fat-saturated imaging. At 3T, images were obtained with the following technical parameters: fast spin echo, repetition time (TR) of 1800 to 2200 milliseconds, echo time (TE) of 34 milliseconds, fat saturation, 3.5 mm thickness, approximately 2-minute scan time, 0.7-mm skip between slices, the field of view (FOV) of 160 mm, 384 x 256 matrix, number of excitations (NEX) of 2, and echo train length (ETL) of 18 to 21.
Data collection was performed through SyngoVia reading software (Siemens Healthineers, Erlangen, Germany). Patient age, gender, underlying diagnosis, and pertinent previous surgical treatments were all recorded separately from the measurements.
For analysis, patients were categorized according to age into 4 quartiles; (25–44; 45–56; 57–67; 68–81). Chi-square test was conducted to test for significant differences between male and female, left and right and the difference between age groupsin regard to the presence of OA. Analysis was completed using the Statistical Package for Social Science (IBM Corp. Released 2015. IBM SPSS Statistics for Mac, Version 23.0. Armonk, NY: IBM Corp).