Level set segmentation for brain region using CT scan images
Chuen Rue Ng, Norliza Mohd Noor, Omar Mohd Rijal · 2016
The segmentation of the brain region is known to be vital in many research especially in the field of neuropsychiatric disorders for both detection and diagnosis. The segmentation of the brain and the computation of brain volume proved to be vital in the detection of many brain pathologies using Computed Tomography (CT) scan images. Distance regularized level set is a variation of level set that do not requires reinitialization. Hence, automated algorithm can be developed to segment brain region and quantify brain volume automatically. In this paper, level set was used to segment the brain region. A refine level set is then carried out if the segmented results consist of objects with high or low Hounsfield Unit (HU) as compared to the threshold HU set by calculating the mean of brain region. The automated level set algorithm showed encouraging results after being compared to the ground truth volume traced manually.