Ultrasonography and articular cartilage defects in the knee: an in vitro evaluation of the accuracy of cartilage thickness and defect size assessment.

The purpose of this cadaver study was to test the accuracy of ultrasonography in measuring cartilage thickness, and the extent and depth of induced cartilage defects on the medial and lateral femoral condyles of the knee in a clinically relevant setting. With the knees maximally flexed, cartilage thickness was measured at 24 marked sites in four knees with a 10 MHz probe. The areas of measurement were then excised and the thickness measured with a calliper gauge. In another seven cadaver knees, 21 cartilage defects were produced. The defect diameter varied from 4 to 8 mm. The depths of the defects were either a partial cartilage defect (grade 2), a defect to intact subchondral bone (grade 3), or a defect involving subchondral bone (grade 4) (classification by ICRS). The limits of agreement between ultrasonography and calliper gauge measurement for cartilage thickness were chi(diff)+/-2SD(diff)=0.0+/-0.4 mm. For cartilage defect diameter, the limits of agreement between ultrasonography and the slide ruler measurement were chi(diff)+/-2SD(diff)=-0.2+/-1.0 mm. For the depths of the lesions there was a 100% agreement between radiologist and actual lesion depth for the classification into International Cartilage Repair Society (ICRS) grades 2, 3, and 4. Our conclusion is that ultrasonography is capable of measuring accurately both cartilage thickness and the extent and depth of induced cartilage defects in a cadaver model.