Using a Hybrid System of Coactive Adaptive Neuro-Fuzzy Inference Based on GLCM for Prostate Cancer Categorization

The prostate has a complex tissue structure and a relatively large, varied form. The intricacy of the surrounding anatomical structure makes it challenging to segment the prostate and prostate cancer. Prostate cancer is second in terms of death rates among malignancies in men. A lot of the time, MRI diagnosis uses accurate prostate zonal segmentation. State-of-the-art automatic segmentation techniques sometimes fail to produce well-contained volumetric segmentation of the prostate zones. This happens because certain prostate MRI slices, such as the base and apex slices, are more challenging to segment than others. Nevertheless, such cross-slice information is not well learned and used by current techniques. It can get over this difficulty by using important multi-scale image-based information from adjacent slices. In this work, we automate the segmentation process of prostate cancer images. We use a CANFIS classifier to identify cancer. The recommended method consists of three steps: pre-processing, feature extraction, and classification. Before inputting characteristics into the CANFIS classifier, pre-processing procedures enhance the circulation of the prostate picture. The classifier evaluates a test prostate image to determine if it is normal. The proposed system carries out segmentation and cancer diagnosis. It does this by finding that the test image is aberrant. For the purpose of identifying cancer pixels, the suggested method has a sensitivity of 96.67%, a specificity of 93.15%, and an accuracy of 97.25%.