Multimodal MR Brain Image Segmentation using Hybrid Optimization Techniques for Earlier Prognosis of Tumor

Segmentation of medical images is the process of dividing an image into regions that have some characteristics that are homogeneous and consistent. Medical image segmentation for brain images is an important process in the early diagnosis of brain tumors, brain abnormalities, and their treatment planning. When performed by medical specialists, segmentation of Magnetic Resonance Imaging (MRI) takes a long time. This paper needs to propose some new hybrid approaches for image segmentation and classification for early detection and treatment of brain tumors. The conventional fuzzy c-means clustering (FCM) method is susceptible to artifacts; the local spatial information is frequently presented to an objective function to enhance the FCM method's reliability for segmenting the images. To overcome this limitation, the authors proposed the Fast and Robust FCM (FRFCM), an optimized FCM technique that utilizes morphological restoration (MR) and membership filtration which is considerably quicker and much more reliable than classical FCM. Following the clustering procedure, the new Social Spider Optimization (SSO) swarm algorithm relies on the co - operative features of the social spider is employed to obtain the best outcome. The novel integration of FRFCM and SSO has been affirmed through evaluating on numerous different BraTS datasets, yielding improved performance metric values.