Bone Cancer Detection Using Feature Extraction with Classification Using K-Nearest Neighbor and Decision Tree Algorithm

Abstract

The malignant cells that cannot be controlled from spreading throughout the body is Cancer. Among which the cancer occurs in bone is their type. It is malignant disease occur in bone of human body where their growth cant be controlled from growing. This bone cancer is very critical of all the cancer types since the malignant cells are not identified at their earlier stage and it is the major challenge. Bone cancer is highly common for children and teenagers. For earlier detection of this cancer the correlation of medical imaging has been adapted with image processing and machine learning techniques where maximum accuracy can be obtained similarly even for bone cancer. This paper proposes the detection of bone cancer from the dataset taken from clinical dataset. Here the proposed design comprises of 2 phases in predicting the disorder with higher accuracy. The first stage is extracting the feature of segmented bone image using Gray-Level Co-occurrence Matrix (GLCM) method is applied to extract the features in terms of statistical texture-based and the second phase is classification of extracted feature using K-NN with decision tree algorithm. The simulation results show the enhanced classification results and extracted output with higher accuracy.