33rd Annual Scientific Meeting proceedings

Introduction
The objective of this study was to compare the diagnostic performance of computed tomography arthrography (CTA) and 3T magnetic resonance imaging (MRI) for detecting arthroscopically created experimental cartilage defects in equine cadaver stifles.

Material and Methods
Cartilage defects were created via arthroscopy (n=74) on the femoral condyles, trochlear ridges, intertrochlear groove and the patella, with randomized distributed negative controls (n=53). MRI and CTA studies were evaluated blindly by a board certified radiologist, followed by macroscopic lesion measurement using a calliper gauge. Sensitivity and specificity for MRI and CTA were determined using arthroscopy as a reference and compared.

Results
Mean lesion size and depth were 11mm2 and 2.1mm, respectively. The overall specificity was significantly higher for CTA (66.2%) compared to MRI (51.5%), P=0.04. The sensitivity of CTA (53.2%) and MRI (65.8%) did not significantly differ (P=0.087). Lesion size resulted in a greater odds of detecting lesions with CTA (P=0.04), however, this was not the case for MRI. Sensitivity and specificity of the femoropatellar joint or the femorotibial joints separately did not differ between CTA or MRI.

Discussion/Conclusions
CTA and MRI performed in a mid-range, which was most likely attributed to the experimental study design (small lesion size). CTA had a similar sensitivity as high field MRI in detecting small artificial cartilage lesions. However, CTA showed a higher specificity than MRI, thus being more accurate in diagnosing normal cartilage. This is important in a clinical setting, where CTA may be used to excluded cartilage lesions in horses without them needing arthroscopy.