Main Article Content
Detecting and identifying bone cancers in Magnetic Resonance Imaging (MRI) is a difficult and crucial job in medical imaging. MRI provides information about anatomical structures and possibly diseased tissues. Therefore, this research presents a novel approach for segmenting and categorizing MRI bone Neoplasms. This study encompasses the phases of preprocessing, segmentation, feature extraction, feature selection, and picture classification. In the preprocessing step, Speckles and white Gaussian noise are removed from the provided MRI images using the Distribution based Adaptive Filtering (DAF) approach. It smooths the picture by reducing noise and boosting the image's intensity. Clustering and label creation are conducted in the segmentation phase to forecast the Neoplasm portion. The Neighboring Cellular Automata (NCA) model is presented for clustering. The clustered picture is then assigned names such as Background (BG), border region, Gray Matter (GM), and White Matter (WM). The segmented image's characteristics are retrieved using the Differential Binary Pattern (DBP) approach. After the feature vectors have been extracted, the firefly optimization method is used to identify the best features. After choosing the feature set, the Pointing Kernel Classifier (PKC) is used to categorize aberrant and normal bone pictures and the types of bone cancers. The performance of the suggested technique is assessed using sensitivity, specificity, precision, correction rate, and positive and negative likelihood.