Computed Tomographic and Low-Field Magnetic Resonance Arthrography: A Comparison of Techniques For Observing...

To compare low-field magnetic resonance arthrography (MRA) and computed tomographic arthrography (CTA) for identification of intra-articular ligaments and joint cartilage in the normal canine shoulder with the aid of cross-sectional anatomy.

Normal adult Foxhounds (n = 3) and 4 Foxhound cadaver shoulders.

CTA of 3 dogs was performed after intra-articular administration of contrast medium in both shoulders. Transverse slices and reformatted sagittal and dorsal images were obtained. Later, an MRA study of all 6 shoulder joints was performed using a low-field system. T1- and T2-weighted images were made in 3 planes, arthrography performed, and T1-weighted images repeated. Unrelated cadaver shoulder joints were injected with red methylmethacrylate, frozen, and sectioned into dorsal, sagittal, and transverse sections. Representative anatomic images were compared with the CTA, MRA, and T2-weighted images.

The joint cartilage was identified on all CTA images. The glenohumeral ligaments were clearly visible on the transverse and dorsal CTA and MRA images but were not detected on the T2-weighted sequences. The biceps tendon could be evaluated on all images. The supraspinatus, infraspinatus, and subscapularis muscle tendons were most easily identified on the MRA and T2-weighted MR images.

CTA was useful for seeing articular cartilage and both CTA and MRA were useful for seeing the normal glenohumeral ligaments; however, further research is needed to examine each technique's sensitivity for diagnosing the pathology of these ligaments and any associated lesions of the capsule, tendon, and bone, and for determining the clinical importance of CTA for evaluating loose fragments or osteochondral lesions.