[Knee joint cartilage in magnetic resonance tomography. MR chondrovolumetry (MR-CVM) using fat-suppressed FLASH 3D sequence].

UNLABELLED

The objective of this study was to optimize the demonstration of articular cartilage with magnetic resonance tomography (MRT) and to assess its accuracy in determining the articular cartilage volume of the knee joint.

METHODS

A fat-suppressed FLASH-3D sequence was optimized on healthy volunteers. A fresh cardaveric knee joint was removed from a 82-year-old man and immediately imaged without being frozen or fixed. MRT was carried out at 1.5 T and 25 mT/m (Vision, Siemens, Erlangen, Germany) with a conventional CP knee coil. Sagittal and transverse sections were acquired perpendicular to the articular surfaces, and then, parallel to these imaging planes, anatomical sections were obtained with a diamond band saw. The volumes of the patellar, tibial and femoral cartilages were determined from both the images and the sections, using an image analysing system.

RESULTS

Signal intensities and contrast-to-noise ratios depend on the parameters used. The highest contrast between the cartilage and the periarticular tissues was obtained at a flip angle of 30 degrees (TR = 60 ms, TE = 11 ms). However, a flip angle of 60 degrees was judged to provide optimal subjective image quality. Using these parameters, the deviations between the radiological and the anatomical cartilage volumes were -4.7% in the patella, -3.1% in the tibial plateau and -4.2% in the femur.

CONCLUSIONS

The volumes of the knee-joint cartilages may be accurately determined with MRT if an appropriate pulse sequence is chosen. From a clinical point of view the differences between MR images and anatomical sections can be considered negligible. These differences may be explained on the basis that MRT delineates uncalcified cartilage only and that the calcified layer is demonstrated as part of the subchondral bone plate. However, close correlation of the thicknesses of the uncalcified and the calcified layer has been reported in the literature, so the relative distribution of articular cartilage should be accurately reflected in MR images.