Adaptive image denoising approach for low-dose computed tomography

Low-dose computed tomography (LDCT) is usually performed by reducing the power of the x-ray tube in clinical CT scanners. However, images acquired through LDCT are known to be of low-quality due to the presence of statistical noise and other related artifacts. Effective denoising techniques are required to improve the quality of LDCT images towards green and safe CT imaging. In this paper, a new method is presented to improve the so-called, non-local means (NLM) filtering for effective LDCT imaging. The proposed method incorporates a prior knowledge obtained from probabilistic atlas during the filtering process. Additional anatomical information obtained through the atlas is likely to be useful in improving the image quality using NLM filtering. The proposed method is evaluated using real data and a notable improvement in image quality improvement is achieved.