OPTIMIZING PHASE CORRECTION FOR MULTI-SHOT DWI

点击翻译按钮获取中文摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. INTRODUCTION : Multi-shot techniques, such as PROPELLER (1) and self-navigated interleaved spiral (SNAILS) (2), have demonstrated great utility for high resolution diffusion-weighted imaging (DWI). However, little is known about how trajectory design parameters affect their navigation capabilities or how to optimize trajectories for multi-shot DWI. Recently, a conjugate gradient (CG) method that corrects for phase variation induced by physiologic motion was introduced and helps to improve the image quality of interleaved DWI quite dramatically. Moreover, this technique can be easily combined with the SENSE (3) reconstruction for parallel imaging. With the CG method we can evaluate the phase correction capability relative to different readout strategies and their trajectory design parameters in the presence of k-space undersampling and phase map errors. This allows us to optimize empirically the trajectories for multi-shot DWI. In this study, we demonstrate that k-space trajectories that oversample the center of k-space have more favorable properties for phase correction than critically sampled trajectories. Well-balanced oversampling of the center k-space increases the tolerance for trajectory imperfection (e.g. due to eddy currents) and errors in phase map, thus offers more robustness for multi-shot DWI. RESULTS: We show a representative set of images reconstructed for different trajectories under various conditions. The first row corresponds to the case with perfect phase maps; the second row corresponds to the case with 5% error in the phase maps; and the third row corresponds to the case with only half k-space data. Surprisingly, even with perfect phase information, the conventional spiral (? = 1) image has high residual artifacts, which is likely due to gridding errors. It compares reconstruction errors for various trajectories. It is clear that an increasing oversampling factor at the center of k-space improves the trajectory¿s tolerance to imperfect phase map and k-space trajectory. Even with a reduction factor of 2, the image with ? = 4 demonstrates good quality, while severe aliasing artifacts can be seen in EPI and conventional spiral images. ACKNOWLEDGMENTS: NIH-1R01NS35959, NIH-1R01EB002771, Lucas Foundation, Center of Advanced MR Technology of Stanford (NCRR P41 RR 09784) REFERENCES: 1) Pipe JG, et al. Magn Reson Med 2002; 47:42-52. 2) Liu C, et al. Magn Reson Med 2004; 52:1388-1396. 3) Pruessmann KP et al. Magn Reson Med 2001; 46:638-651. 4) Butts K, et al. Magn Reson Med 1996; 35:763-770