[Value of various MR sequences using 1.5 and 3.0 Tesla in analyzing cartilaginous defects of the patella in an animal model].

PURPOSE: Comparison of MRI and macropathologic evaluation using various sequences and field strengths in the detection, localization and measurement of cartilage defects in an animal model.

MATERIALS AND

METHODS

After open creation of retropatellar cartilage defects of various widths, depths and locations in 8 cadaveric sheep knee joints, the knees were examined using a fat-suppressed (FS), proton density-weighted (PD) fast spin echo (FSE), and 2D and 3D gradient echo (GE) sequences on 1.5 T and 3.0 T MR scanners. The images were analyzed by two independent radiologists in a blinded manner, by dividing the patella into 15 virtual segments. The results were correlated with the macropathologic findings with regards to location, width, and depth of the defects.

RESULTS

The highest sensitivity (67.1 %), diagnostic accuracy (85.4 %), positive (87.3 %), and negative (84.7 %) predictive values in detecting defects were obtained using the 3.0 T FS-3D-GE sequence. The highest specificity (95.6 %) yielded the 3.0 T FS-2D-GE sequence, with the other sequences inferior by no more than 2.6 %. In general, FS-3D-GE sequences were superior to FS-2D-GE (3.0 T: p < 0.05; 1.5 T: p < 0.05) and especially to FS-PD-FSE sequences (3.0 T: p < 0.01; 1.5 T: p < 0.05). In determining the defects' widths, the 3.0 T FS-3D-GE sequence was superior to all other sequences (correct measurements: 50.0 %), with only slight superiority to the 1.5 T FS-3D-GE sequence (46.9 %, p > 0.05) but clear superiority to the other sequences (28.1 - 40.6 %, vs. 1.5 T FS-PD-FSE: p < 0.05, vs. other sequences: p > 0.05). To determine the defects' depths, the 1.5 T FS-3D-GE sequence was most reliable (correct measurements: 53.1 %), followed by the 3.0 T FS-3D-GE sequence (50.0 %, significance of difference: p > 0.05).

CONCLUSION

In detecting cartilage defects, the field strength of 3.0 Tesla was only superior to 1.5 T MRI using fat-saturated 3D- or 2D-GE-sequences but not in fat-saturated proton density-weighted SE-sequences. In determination of depth and length of the defects, the higher field strength was not advantageous.