A comparison of readout segmented EPI and interleaved EPI in high-resolution diffusion weighted imaging.

PURPOSE: To provide a comprehensive understanding of multi-shot EPI diffusion imaging methods by comparing Readout segmented EPI (RS-EPI) and interleaved EPI (iEPI).

MATERIALS AND METHODS: RS-EPI and iEPI were compared on the same 3T scanner. A 2D navigator was used for both RS-EPI and iEPI for phase correction. Signal to noise ratio (SNR), fractional anisotropy (FA) and distortion level were compared using phantom data. Distortion reduction capability and scan efficiency were compared with different protocols with simulations. In addition, distortion reduction capability and diffusion tensor imaging performance were compared using in vivo data.

RESULTS: Our phantom data showed that the mean SNRs were 50.5, 86.6 and 45.4 for RS-EPI using GRAPPA=3, fully sampled iEPI and iEPI using GRAPPA=2 respectively. The mean FA values were 0.08, 0.05 and 0.09 for RS-EPI using GRAPPA=3, fully sampled iEPI and iEPI using GRAPPA=2 respectively. The distortion levels were 1.34mm, 1.29mm and 0.61mm for RS-EPI using GRAPPA=3, fully sampled iEPI and iEPI using GRAPPA=2 respectively. The effective echo spacing could be reduced by increasing the number of shots for both methods but more prominent for iEPI. The scan time was approximately proportional to the number of shots for both methods and RS-EPI showed a shorter scan time. Our in vivo data for distortion comparison showed consistent results with the effective echo spacing study. The mean difference of the FA and MD values between the high resolution sequences and SS-EPI was all within 7%.

CONCLUSION: For high resolution diffusion imaging, iEPI has more potential in distortion reduction than RS-EPI when increasing the number of shots. RS-EPI can achieve a reasonable SNR with a shorter scan time than iEPI. RS-EPI and iEPI have similar performance in FA and MD quantifications as well as showing structure details when using eleven shots for in vivo diffusion tensor imaging.