Tumor Classification in Osteosarcoma Using Convolutional Neural Network

Oesteosarcoma is the most common type of cancer that starts in the bone and it can spread through all parts of the body. Children between the age group of 10 to 30 and teenagers get affected by primary bone cancer osteosarcoma. It is a rare disease and its end stage tumors are not curable. The aim of this work is to develop a Computer Aided Diagnosis (CAD) system using Convolutional Neural Network (CNN) for Osteosarcoma tumor classification and to minimize the risk of cancer by early diagnosis using the proposed system. This research work is done on the dataset taken from Cancer Image Archive.. The proposed convolutional neural network architecture has been built with ten convolutional layers and the retrieval of accuracy range from 82.14% to 88.64% was achieved. This proposed system can assist the pathologist in diagnosing the Osteosarcoma tumor classification.