Development of a cerebral aneurysm segmentation method to prevent sentinel hemorrhage

Abstract

Image segmentation being the first step is always crucial for brain aneurysm treatment planning; it is also crucial during the procedure. A robust brain aneurysm segmentation has the potential to prevent the blood leakage, also known as sentinel hemorrhage. Here, we present a method combining a multiresolution and a statistical approach in two dimensional domain to segment cerebral aneurysm in which the Contourlet transform (CT) extracts the image features, while the Hidden Markov Random Field with Expectation Maximization (HMRF-EM) segments the image, based on the spatial contextual constraints. The proposed algorithm is tested on Three-Dimensional Rotational Angiography (3DRA) datasets; the average values of segmentation accuracy, DSC, FPR, FNR, specificity, and sensitivity, are found to be 99.72%, 93.52%, 0.07%, 5.23%, 94.77%, and 99.96%, respectively.