Low-dose CT image denoising algorithm based on weighted encoding with sparse nonlocal regularization

Low-dose CT images frequently suffer from noise and artifacts, hindering accurate clinical diagnosis by physicians. This study introduces the concept of gray absolute correlation degree within a weighted encoding framework, which incorporates a sparse nonlocal regularization algorithm. Specifically, the nonlocal regularization-based denoising technique utilizes the gray absolute correlation degree to differentiate noise points from those in edge and smooth areas, thereby filtering out the noise. The experimental results show that the proposed algorithm can effectively suppress mixed noise in low-dose CT images.