Courtesy of Dr Jianguo Zhang, this image shows the result of a winning algorithm submitted to the White Matter Hyperintensities (WMH) Segmentation Grand Challenge held at MICCAI 2017.
At left is the raw FLAIR axial image; at centre is the ground truth of WMH regions annotated by a radiologist; at right is the automatic WMH segmentation result generated by the algorithm (U-Net with Two-channel Input). This algorithm, developed by an international team led by Dr Zhang, placed first among 20 Grand Challenge entries. Presentation slides from the team can be viewed here.
Further MICCAI 2017 success for the Computer Vision and Image Processing (CVIP) group at Dundee came from MSc student Mayssa Soussia, who received the Best Paper award at the 1st International Workshop on Connectomics in Neuroimaging. Congratulations to all!