Three-Tesla magnetic resonance imaging of the wrist: diagnostic performance compared to 1.5-T.

PURPOSE

To assess the diagnostic capability of a new multi-contrast high-resolution imaging protocol for the wrist at 3 T and to compare it to our standard 1.5-T protocol.

MATERIALS AND METHODS

In 10 volunteers, images were acquired at 1.5 T (standard surface coil) and 3 T (custom-made phased array coil; Gyroscan Intera; Philips Medical Systems, Best, the Netherlands). Imaging protocols consisted of coronal T1-weighted turbo spin-echo (TSE), coronal T2-weighted TSE, and proton density-weighted TSE fat-saturated (FS) sequences and sagittal T2-TSE and proton density-weighted TSE-FS sequences. Increased signal-to-noise ratio at 3 T was used to reduce measured voxel size from 0.50 x 0.50 x 3.0 mm (1.5 T) to 0.20 x 0.20 x 1.5 mm (3 T). Total examination time was approximately 25 minutes. To compare the diagnostic capability of both protocols, 4 observers assessed in consensus the visibility of 14 well-defined clinically important anatomical landmarks (origin and insertion of 6 intrinsic and extrinsic carpal ligaments, central portion of the triangular fibrocartilage complex, and ulna, triquetral, and radial attachments). Image quality and artifacts were ranked qualitatively on a 5-point scale.

RESULTS

Of 140 detectable structures, 75 were detected at 1.5 T and 126 at 3 T. Overall image quality was significantly better at 3 T (3.8 vs 4.9, P = 0.002), whereas artifact score did not differ significantly (4.3 vs 4.5, P = 0.317).

CONCLUSIONS

This is the result of investing the higher signal-to-noise ratio at 3 T into better spatial resolution; depiction of the normal anatomy of the wrist benefits significantly. Additional studies will need to be performed to determine if 3-T images will increase the accuracy of detecting abnormalities of these structures of the wrist.