Efficacy of biceps-radial-slice magnetic resonance images for the diagnosis of biceps and pulley lesions.

Abstract

Injuries to the biceps reflection pulley lead to instability of the long head of the biceps tendon. However, conventional magnetic resonance imaging has low diagnostic accuracy for long head of the biceps tendon and pulley lesions. Here, we investigated the usefulness of novel biceps-radial magnetic resonance imaging for evaluating long head of the biceps tendon and pulley lesions. Biceps-radial magnetic resonance images of 84 patients (84 shoulders) were prospectively analyzed. The biceps-radial magnetic resonance imaging protocol includes sequences acquired in radial planes perpendicular to the long head of the biceps tendon in the shoulder joint. All patients underwent shoulder arthroscopy, and the intraoperative long head of the biceps tendon and pulley lesion findings were compared to the preoperative evaluations. The diagnostic accuracies of the biceps-radial magnetic resonance images and conventional magnetic resonance images were determined. A normal long head of the biceps tendon was observed in 30 (31.6%) patients, partial tears in 43 (52.6%), and complete tears in 11 (15.8%). Normal long head of the biceps tendon stability was present in 54 (61.4%) patients, subluxation in 24 (31.6%), and dislocation in 6 (7.0%). The biceps-radial magnetic resonance (kappa coefficient: 0.94) and conventional magnetic resonance (kappa coefficient: 0.68) images accurately identified long head of the biceps tendon tears. The biceps-radial magnetic resonance images excellently agreed with the arthroscopic findings (kappa coefficient: 0.91), whereas the conventional magnetic resonance images poorly agreed (kappa coefficient: 0.17) regarding long head of the biceps tendon instability. A normal medial wall of the pulley was observed in 26 (31.0%) patients, partial tears in 30 (35.7%), and complete tears in 28 (33.3%). A normal lateral wall of the pulley was observed in 30 (35.7%) patients, partial tears in 21 (25.0%), and complete tears in 33 (39.3%). The biceps-radial magnetic resonance images excellently agreed with the arthroscopic findings (kappa coefficient: 0.89), whereas the conventional magnetic resonance images moderately agreed (kappa coefficient: 0.50) regarding medial pulley lesions. The biceps-radial magnetic resonance images excellently agreed with the arthroscopic findings (kappa coefficient: 0.95), and the conventional magnetic resonance images moderately agreed (kappa coefficient: 0.56) regarding lateral pulley lesions. Biceps-radial magnetic resonance imaging allows for tracking of the long head of the biceps tendon and pulley from the supraglenoid tuberosity to the bicipital groove in the glenohumeral joint and accurate evaluations of long head of the biceps tendon and pulley lesions.