DEEP LEARNING BASED CLASSIFICATION OF BONE TUMORS USING IMAGE SEGMENTATION

This paper addresses the critical issue of automated cancer and tumor detection, focusing on bone cancers such as Chondrosarcoma, Ewing sarcoma, and Osteo sarcoma. With a staggering one in three individuals experiencing cancer at some point, the intersection of biomedical research and computer science becomes crucial. The study employs deep learning algorithms applied to histological images for the extraction of features specific to each type of bone tumor. In cases of metastasis, where cancer cells spread through the bloodstream or lymph system, the system processes images to identify cells in the initial stages. The proposed approach utilizes a convolutional neural network (CNN) to enhance efficiency, reduce processing time, and improve accuracy. Preprocessed images undergo classification by trained classifiers, enhancing the overall sensitivity of the system. The introduced algorithm incorporates Deep CNN layers, digital image processing, DNN models, feature extraction, classification, and MATLAB, providing a comprehensive solution for automated cancer and tumor detection. This research contributes to the advancement of medical diagnostics, offering a more accurate and efficient method for early detection and classification of bone tumors.

Keywords: Deep CNN layers, Digital image processing, DNN model, Feature extraction, Classification, MATLAB.