Alvaro Bertelsen1, Paul Aljabar2, Hui Xue3, Latha Srinivasan1,4, Tayyib Hayat1,5, Joanna Allsop1, Daniel Rueckert2, Mary R. Rutherford1,5, Joseph V. Hajnal1
1Robert Steiner MRI Unit, Imaging Sciences Department, MRC Clinical Sciences Centre, Hammersmith Hospital, Imperial College London, London, UK; 2Department of computing, Imperial College London, London, UK; 3MR Research and Development, Siemens Medical Solutions USA, USA; 4Neonatal Department, Hammersmith Hospital, Imperial College London; 5Perinatal Imaging Group, MRC Clinical Sciences Centre, Hammersmith Hospital, Imperial College London
Analysis of adult human cortex from MRI is well established and has recently been extended to the developing cortex of neonates, but segmentation of in utero fetal cortex from MRI remains challenging. Although snapshot imaging with volume reconstruction (SVR) provides reliable 3D fetal brain reconstructions, this has not proved sufficient for automated cortex extraction. We have now improved the quality of SVR reconstructions using a temporally continuous parameterisation of motion and applied cortex segmentation methods we developed for use on neonates. The resulting extracted cortices are substantially complete, paving the way for more detailed cortical analysis of the fetal brain.