xQSM-Quantitative Susceptibility Mapping with Octave Convolutional Neural Networks

14 Apr 2020  ·  Yang Gao, Xuanyu Zhu, Stuart Crozier, Feng Liu, Hongfu Sun ·

Quantitative susceptibility mapping (QSM) is a valuable magnetic resonance imaging (MRI) contrast mechanism that has demonstrated broad clinical applications. However, the image reconstruction of QSM is challenging due to its ill-posed dipole inversion process. In this study, a new deep learning method for QSM reconstruction, namely xQSM, was designed by introducing modified state-of-the-art octave convolutional layers into the U-net backbone. The xQSM method was compared with recently-proposed U-net-based and conventional regularization-based methods, using peak signal to noise ratio (PSNR), structural similarity (SSIM), and region-of-interest measurements. The results from a numerical phantom, a simulated human brain, four in vivo healthy human subjects, a multiple sclerosis patient, a glioblastoma patient, as well as a healthy mouse brain showed that the xQSM led to suppressed artifacts than the conventional methods, and enhanced susceptibility contrast, particularly in the iron-rich deep grey matter region, than the original U-net, consistently. The xQSM method also substantially shortened the reconstruction time from minutes using conventional iterative methods to only a few seconds.

PDF Abstract


  Add Datasets introduced or used in this paper

Results from the Paper

  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.


No methods listed for this paper. Add relevant methods here