z-logo
Premium
The effects of navigator distortion and noise level on interleaved EPI DWI reconstruction: a comparison between image‐ and k‐space‐based method
Author(s) -
Dai Erpeng,
Zhang Zhe,
Ma Xiaodong,
Dong Zijing,
Li Xuesong,
Xiong Yuhui,
Yuan Chun,
Guo Hua
Publication year - 2018
Publication title -
magnetic resonance in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.696
H-Index - 225
eISSN - 1522-2594
pISSN - 0740-3194
DOI - 10.1002/mrm.27190
Subject(s) - distortion (music) , computer science , undersampling , computer vision , artificial intelligence , noise (video) , iterative reconstruction , algorithm , image (mathematics) , bandwidth (computing) , telecommunications , amplifier
Purpose To study the effects of 2D navigator distortion and noise level on interleaved EPI (iEPI) DWI reconstruction, using either the image‐ or k‐space‐based method. Methods The 2D navigator acquisition was adjusted by reducing its echo spacing in the readout direction and undersampling in the phase encoding direction. A POCS‐based reconstruction using image‐space sampling function (IRIS) algorithm (POCSIRIS) was developed to reduce the impact of navigator distortion. POCSIRIS was then compared with the original IRIS algorithm and a SPIRiT‐based k‐space algorithm, under different navigator distortion and noise levels. Results Reducing the navigator distortion can improve the reconstruction of iEPI DWI. The proposed POCSIRIS and SPIRiT‐based algorithms are more tolerable to different navigator distortion levels, compared to the original IRIS algorithm. SPIRiT may be hindered by low SNR of the navigator. Conclusion Multi‐shot iEPI DWI reconstruction can be improved by reducing the 2D navigator distortion. Different reconstruction methods show variable sensitivity to navigator distortion or noise levels. Furthermore, the findings can be valuable in applications such as simultaneous multi‐slice accelerated iEPI DWI and multi‐slab diffusion imaging.

This content is not available in your region!

Continue researching here.

Having issues? You can contact us here