Optimal multi-phase three-dimensional fast imaging with steady-state free precession dynamic MRI and its clinical application to the diagnosis of hepatocellular carcinoma

Purpose: To determine the appropriate acquisition parameters for three-dimensional fast imaging with steady-state free precession (3D-FISP), to clarify the superiority of 3D-FISP to two-dimensional fast low-angle shot (2D-FLASH) on phantom study, and to clarify the clinical usefulness of 3D-FISP in diagnosing hepatocellular carcinoma (HCC). Materials and Methods: 3D-FISP images with varying flip angles were compared by using a phantom. Signal-to-noise ratios (SNRs) and contrast-enhancement ratios (CERs) were compared for the four two-dimensional fast low-angle shot (2D-FLASH) sequences and 3D-FISP sequences in a phantom. The optimal 3D-FISP dynamic study was compared with plain, post-contrast MR sequences used to study 78 HCC cases and analyzed according to histological grade. The 3D-FISP image obtained 30 sec after gadopentetate dimeglumine (Gd) administration was also compared with CT hepatic angiography (CTHA). Results: A 25° flip angle and double-dose Gd administration were appropriate for 3D-FISP dynamic study. CER was the highest with 3D-FISP, and SNR was higher in 3D-FISP than in 2D-FLASH images in a phantom with high Gd concentration. Among the 105 lesions, 103 (98%) were depicted on 3D-FISP images. The detection rate of HCC on 3D-FISP was higher than 95% for each histological grade. The vascularity of the tumors as determined by CTHA findings was correctly diagnosed on 3D-FISP in 80% of cases. Conclusion: In phantom study, 3D-FISP with double-dose Gd injection showed higher contrast than 2D-FLASH as a sequence for liver dynamic study. In clinical study, 3D-FISP is useful in the detection of HCC, regardless of tumor vascularity and histological grade.