Osteoporosis diagnosis using steerable pyramid decomposition and fractional Brownian motion

Osteoporosis is a skeletal disease characterized by a decrease in bone mass and density. The objective of this paper is the diagnosis of osteoporosis disease based on X-Ray images using Wavelet decomposition and fractal analysis. The method lies on a preprocessing phase followed by Wavelet decomposition of X-Ray images into sub-bands using steerable pyramid decomposition. Features are then extracted using fractional Brownian motion analysis. To separate the Osteoporosis population (OP) from Control cases (CT), we use the Support vector machines (SVM) classifier. Our proposed method shows a good diagnosis quality achieving an accuracy rate of 93%.