Noise reduction in DEXA image based on system noise modeling

Abstract

Denoising X-ray image is an important preprocessing process to measure BMD (Bone Mineral Density) correctly. In this paper, we deal with denoising X-ray image in DEXA (Dual-Energy X-ray Absorptiometry) system. Image noise in DEXA is modeled by separating source and detector noises. Using the noise model in DEXA, a noise reduction method is proposed. The proposed denoising method consists of Wiener filter to reduce the detector noise, and logarithmic scale-wavelet BayesShrink filter to reduce the source noise. Experimental results verified the proposed DEXA noise model and showed that the proposed noise reduction method outperformed the conventional method.