Reconstruction of fetal brain MRI with intensity matching and complete outlier removal.
Bottom Line: The method incorporates novel intensity matching of acquired 2D slices and robust statistics which completely excludes identified misregistered or corrupted voxels and slices.The proposed novel EM-based robust statistics also improves the reconstruction when compared to previously proposed Huber robust statistics.The best results are obtained when thin-slice data and the correct approximation of the point spread function is used.
Affiliation: Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, UK. firstname.lastname@example.orgShow MeSH
Mentions: To correct for motion between slices we use the original scheme proposed by Rousseau et al. (2006) and Jiang et al. (2007). Stacks are first co-aligned using volumetric rigid registration and the first estimate of the volume is reconstructed. Afterwards, each slice is registered to the reconstructed volume separately. Slice-to-volume rigid registration is interleaved with super-resolution reconstruction for a fixed number of iterations, which is determined experimentally. The advantage of this simple scheme is that any existing similarity measure can be used to maximize the quality of the alignment. We chose normalized mutual information (Studholme et al., 1999), as it is independent of the scaling factors and has low sensitivity to low magnitude bias fields, which are typically present in MRI acquired at 1.5 T. An overview of this interleaved scheme is presented in Fig. 2.
Affiliation: Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, UK. email@example.com