Brain tumor detection and segmentation using mesh-free super-diffusive model

Abstract

Brain tumor detection and segmentation is a complex and challenging task in image processing. Most of the techniques used for brain tumor detection and segmentation in the early stage are failed to locate a tumor region accurately. In order to get higher performance, a partial differential equation has been designed and has been used in tumor detection and segmentation. However, these methods take more time due to processing of mesh. In this manuscript, a mesh-free fractional partial differential equation based super-diffusive model is being proposed. The model enables to choose an arbitrary order of spatial derivative which enables to locate tumor region more accurately in the early stage. A mesh-free approach has been used to solve the proposed model to remove the dependency on the mesh. Qualitative and quantitative analysis have been done to verify the claim in early stage detection and segmentation of brain tumor. It has been found that the proposed model is efficient in solving super-diffusive process and is able to extract tumor region more accurately than state-of-the-art methods.