Abstract
This study was designed to evaluate the value of MR imaging compared with arthro-bursography for incomplete tears of the rotator cuff and to classify the post-operative MRI findings of an incomplete tear. Pre-operative MRI were taken on 42 shoulders and we applied arthro-bursography to 99 shoulders. The pre-operateve MRI findings were classified into complete and incomplete tears which were classified into 3 patterns; high signal of the bursal side of the tendon, intratendinous high signal and high signal of the articular side of the tendon. The post-operative MRI findings were classified into 4 types; Typel-homogeneous pattern, Type2a-partial intratendinous high signal pattern, Type2b-intratendinous high signal pattern in almost all slices of the sutured area, Type3-high signal at the suture site of the bone groove. Sensitivity and specificity of the MRI in incomplete tear cases were 78.6% and 89.7%, respectively. The diagnosis rate of the MRI, especially in the intratendinous tear, was higher than arthrobursography. There was no correlation between post-operative MRI findings and the clinical results. The intratendinous high signal patterns of the MRI at the suture site, which were seen at 6 months postop., tend to become lower in patients with a 1 year or longer follow-up.
