Diagnostic accuracy of magnetic resonance imaging for partial tears of the long head of the biceps tendon in patients with rotator cuff tears.

Background: Magnetic resonance imaging (MRI) is useful for diagnosing shoulder diseases preoperatively. However, detection of partial tears of the long head of the biceps tendon (LHBT) using current clinical tests and imaging modalities is difficult. We aimed to evaluate the accuracy of radial-slice MRI for diagnosing partial tears of the LHBT. We hypothesized that radial-slice MRI may be a valuable diagnostic tool for assessing diagnosing tears of the LHBT. Methods: We retrospectively investigated 118 patients who underwent shoulder arthroscopy for rotator cuff tears. Intraoperative LHBT findings were compared with the identification of partial tears of the LHBT on conventional-slice MRI and radial-slice MRI, using a 3.0-T system. We calculated sensitivity, specificity, accuracy, and positive and negative predictive values for the detection of LHBT tears. Inter- and intraobserver reliability for radial-slice MRI was calculated using kappa statistics. Results: We diagnosed 69 patients (58%) without any LHBT tears and 49 with partial tears (42%), arthroscopically. Sensitivity, specificity, accuracy, and positive and negative predictive values of conventional-slice MRI for detection of partial tears of the LHBT were 52%, 94%, 78%, 92%, and 58%, respectively. Radial-slice MRI had 84% sensitivity, 90% specificity, 86% accuracy, and 92% positive and 80% negative predictive values for partial tears of the LHBT. Inter- and intraobserver reliability for radial-slice MRI was 0.69 and 0.74, respectively, corresponding to high reproducibility and defined as good. Conclusion: Radial-slice MRI demonstrated significantly higher sensitivity than conventional-slice MRI. These results indicate that radial-slice MRI is useful for diagnosing LHBT partial tears.