MR patterns of involvement of humeral head osteonecrosis.

PURPOSE: To characterize osteonecrosis of the humeral head on coronal and axial MR images. METHODS: We retrospectively reviewed MR examinations in patients with humeral head osteonecrosis. The angle of the entire affected articular surface at three levels was measured and an angle subtended by the margins of osteonecrosis was also measured. RESULTS: The appearance of osteonecrosis is identical to that in the femoral head with signal abnormality (13/13), double-line sign (7/13), and subchondral fractures (3/13). The superior aspect of the humeral head was the most common location of osteonecrosis. The MR staging resulted as follows: stage II (6/13), stage IIIA (1/13), stage IIIB (1/13), and stage IV (5/13). The maximum percentage of involvement of the articular surface demonstrated the following distribution: 0% to 25%, 0/11; 25% to 50%, 1/13; 50% to 75%, 5/13; and 75% to 100%, 7/13. CONCLUSIONS: Quantitative analysis of the percentage of involvement of the humeral head was performed and indicates that in most cases, 75% to 100% of the articular surface at the levels measured was involved.

Magnetic resonance characteristics of medial plica of the knee: correlation with arthroscopic resection.

OBJECTIVE: The purpose of this study was to determine if there are magnetic resonance (MR) characteristics of the medial plica that correlate with the likelihood of plica resection. METHODS: Sixty-six knee MR examinations from patients who underwent subsequent knee arthroscopy were evaluated retrospectively for the presence of a medial plica. The plicae were then characterized by relative width, thickness, relation to the trochlear cartilage, associated cartilage changes, and effusion. The arthroscopy reports from each knee were reviewed for the presence, description, and resection of medial plicae. RESULTS: A medial plica was demonstrated by MR in 46 of 66 (69.7%) cases. A medial plica was mentioned in 16 of 66 (24.2%) arthroscopic reports. No MR characteristics were significantly predictive of resection. Location of a plica adjacent to cartilage had a P value of 0.0786, plica width had a P value of 0.0858, and plica thickness had a P value of 0.1685. CONCLUSION: No MR characteristics of medial plicae were found to be predictive of subsequent resection at arthroscopy.