Evaluation of articular cartilage thickness of the humeral head and the glenoid fossa by MR arthrography: anatomic correlation in cadavers

Objective. To evaluate the accuracy of MR arthrography in determining the thickness of articular cartilage of the humeral head and glenoid fossa. Design and patients. MR arthrography of the glenohumeral joint was performed in 17 cadaveric shoulders. Articular cartilage thickness was measured on the MR arthrographic images and corresponding anatomic sections. Results. The correlation coefficients for MR arthrographic measurement versus anatomic measurement of the cartilage thickness were 0.7324 and 0.8757 for humeral head and glenoid fossa, respectively. With regard to the humeral head, there was a tendency to overestimate regions of thin cartilage and underestimate regions of thick cartilage. This tendency was not found in the assessment of glenoid cartilage. The mean of the absolute value of MR-anatomic differences was similar on the glenoid side (0.27 mm) and the humeral side (0.29 mm). The accuracy of measurement was significantly better on the glenoid side (Fisher’s r-to-Z transformation: Z=5.21, P=0.000001). Conclusion. MR arthrography causes a moderate degree of error in the naked-eye measurement of the cartilage of the glenohumeral joint. The accuracy is higher on the glenoid side than on the humeral side.