Dynamic segmentation of the cerebral cortex in MR data using implicit active contours (original) (raw)

We propose a dynamic coupled-surface level set approach for the segmentation of the cerebral cortex in MR images of the human brain. An iterative scheme for the estimations of local intensity distributions is applied to compensate for artefacts within the data. Results are given for 5 MR data sets acquired on a 3 Tesla Scanner. No pre-processing is required and a constant set of parameters is used for all data sets. We also present a comparison to segmentation results created by the standard neuroimaging software BrainVoyager and show advantages of our approach.