PURPOSE
To investigate the cause of the trilaminar appearance within hyaline cartilage observed on magnetic resonance (MR) images obtained with a fat-suppressed three-dimensional spoiled gradient-recalled sequence.
MATERIALS AND METHODS
The knees of three asymptomatic volunteers were imaged with a fat-suppressed, three-dimensional, spoiled gradient-recalled sequence. The field of view, number of phase-encoding steps, and phase-encoding direction were varied. On each image, the thickness of the patellar and trochlear cartilage was measured in millimeters and divided by the pixel dimension, which effectively expressed the thickness as the number of pixels. Finally, the number of pixels was compared with the number of alternating hyperintense and hypointense lines depicted.
RESULTS
The number of truncation lines increased as pixel dimension was reduced by either decreasing the field of view or increasing the number of phase-encoding steps. The accuracy for predicting more than three lines with use of an anteroposterior phase-encoding direction varied between 83% and 92%. The appearance of the cartilage was altered when phase- and frequency-encoding directions were exchanged, but truncation lines were still evident.
CONCLUSION
The trilaminar appearance depicted within hyaline cartilage on MR images obtained with this sequence is predominantly attributable to truncation artifact rather than to histologic zonal anatomy.