Texture analysis and fracture identification of lower extremity bones X-ray images

Abstract

Lower limb bones or lower limb component related to the torso with pelvic ankle interference can be fractured. Fractures can be detected automatically take advantage x-ray images performed using feature extraction methods. Feature Extraction helpful to know existence and location of fracture with x-ray images. This research apply Gray Level Co-Occurrence Matrix (GLCM) and K-Means Clustering Algorithm to analyze texture of lower extremity bones or lower limb bones x-ray images especially on the lower leg bones (cruris) consisting of two long bones (tibia) and leg bone (fibula), as well as the kneecap bone (patella). The GLCM feature extraction process yields an image characteristic with four parameters, i.e. Contrast, Correlation, Energy, and Homogeneity done before clustering steps for identification of fractured or non-fractured (normal) bones. The results accuracy texture analysis of lower extremity bones x-ray images using GLCM Feature Extraction Method and K-Means Clustering Algorithm is 80 percent.