The conventional magnetic resonance method may overestimate off-track Hill-Sachs lesions

Three-dimensional (3D)-reconstructed computed tomography (CT) images have been used to evaluate on-track/off-track Hill-Sachs lesions. In the magnetic resonance imaging (MRI) method, there is no need to reconstruct images to determine the measurement position, and measurements can be easily performed using the standard axial and sagittal imaging sequences. For these reasons, recent studies have increasingly adopted MRI for measurements. The purpose of this study was to investigate whether or not the measurements using magnetic resonance (MR) images were as accurate as those using 3D-CT images.

This study is a retrospective case-control study. Thirty-two cases with recurrent anterior dislocation of the shoulder that underwent both MR and CT examinations preoperatively were enrolled. A multihelical CT scanner was used for CT and a 1.5T scanner was used for MRI. In the measurement using 3D-CT images (CT method), the Hill-Sachs interval was measured in the plane passing the center of the humeral head and perpendicular to the Hill-Sachs lesion. In the MR measurement (conventional MR method), the Hill-Sachs interval was measured at the widest point using axial images. The Hill-Sachs interval was measured in the same way as CT method on MR images (modified MR method).

Five of 32 cases were judged as an off-track lesion in the CT method, whereas 8 and 6 cases in the conventional and modified MR methods, respectively. There were 3 mismatched cases between the CT and conventional MR methods. The Hill-Sachs lesions in these 3 cases were judged as on-track in the CT method but as off-track in the conventional MR method. In these 3 cases, the Hill-Sachs interval measured by the MR method was significantly greater than that by the CT method (P = .0028). There was only 1 mismatched case between the CT and modified MR methods.

Our results demonstrated that the conventional MR method overestimated an off-track lesion compared to the CT method. The difference in the measurement plane between CT method and MR method was likely to cause a difference between the 2 methods. On the other hand, our modified MR method used the same plane as the CT method, which enabled us to evaluate off-track lesions more accurately than the conventional MR method. The conventional MR method may overestimate off-track lesions. In contrast, the modified MR method is expected to achieve accuracy closer to that of the CT method, making it a potentially useful measurement technique.