查看文章

A novel Neuro–level set algorithm is proposed and evaluated for simultaneous segmentation and grading of brain tumors. The algorithm utilizes vascular and cellular information provided by the reconstructed images of dynamic contrast susceptibility magnetic resonance images and apparent diffusion coefficient images. The reconstructed images are relative cerebral blood volume (rCBV) image, time to peak (TTP) image and percentage of base at peak (PBP) image along with apparent diffusion coefficient (ADC) maps. The proposed neural-levelset algorithm utilizes a neural block to provide the speed image for the levelset methods. In this study, two different architectures of levelset method have been implemented and their results are compared. The results show that the proposed Neuro-Levelset method performs better in differentiating the tumor, edema, necrosis in reconstructed images of dynamic contrast susceptibility magnetic resonance images and apparent diffusion coefficient MR images.