Comparison of GWO-SVM and Random Forest Classifiers in a LevelSet based approach for Bladder wall segmentation and characterisation using MR images

In order to characterize the bladder state and functioning, it is necessary to succeed the segmentation of its wall in MR images. In this context, we propose a computer-aided diagnosis system based on segmentation and classification applied to the Bladder Wall (BW), as a part of spina bifida disease study. The proposed system starts with the BW extraction using an improved levelSet based algorithm. Then an optimized classification is proposed using some selected features. Obtained results proves the efficiency of the proposed system, which can be significantly helpful for radiologist avoiding the fastidious manual segmentation and providing a precise idea about the spina bifida severity