DETECTION AND FEATURE EXTRACTION OF MRI AND CT IMAGES USING MEDICAL IMAGES

Abstract

An efficient procedure for specifically defining the tumor boundary present in an input MRI picture is proposed in this article. Various mammogram photographs were taken and examined for the comparative analysis. CA segmentation and other existing algorithms such as Otsu's thresholding and canny edge detection were used to model brain MRI images. CA segmentation is the best choice out of any of these segmentation processes. Its simplicity over a single slice and lower susceptibility to initialization, reliability in terms of computation time, robustness against diverse and heterogeneous tumor forms, computational efficiency, and ease of use are all factors. In oncologic imaging, segmentation of brain tumors on diagnostic photographs is important for cancer management and surveillance. With the advent of image driven surgical methods, it is becoming more common. Outlining the brain tumor contour, which is normally performed manually on contrast enhanced T1-weighted MRI in current clinical practice, is a critical phase in preparing spatially localized radiotherapy. Cellular Automaton-based seeded tumor segmentation is used to segment solid brain tumors in this article. It aids physicians and researchers in the preparation of radio surgery and the evaluation of treatment reaction. The findings show that the collected pictures may be used to make an accurate diagnosis.