Statistical image reconstruction for low-dose CT using nonlocal means-based regularization. Part II: An adaptive...

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computational Imaging

Background:

• Low-dose X-ray computed tomography (CT) imaging reduces radiation exposure but increases projection data noise.
• Conventional filtered back-projection (FBP) methods produce noisy images with artifacts at low doses.
• Nonlocal means (NLM) filtering can reduce noise but may not fully eliminate artifacts in severely degraded images.

Purpose of the Study:

• To develop and evaluate a novel NLM-regularized statistical image reconstruction scheme for low-dose CT.
• To introduce spatially variant filtering parameters adaptive to local image characteristics.
• To demonstrate the superiority of the adaptive method over conventional techniques for low-dose CT.

Main Methods:

• Proposed a NLM-regularized statistical image reconstruction scheme.
• Developed a novel strategy for spatially variant filtering parameters adaptive to local image characteristics.
• Evaluated the method using low-contrast phantoms and clinical patient data.

Main Results:

• The proposed adaptive NLM-regularized statistical image reconstruction effectively suppresses noise-induced artifacts.
• Spatially adaptive filtering parameters are necessary for optimal reconstruction across the entire field of view.
• The adaptive method significantly improves the quality of reconstructed CT images from low-dose acquisitions.

Conclusions:

• The adaptive NLM-regularized statistical image reconstruction method offers superior performance for low-dose CT.
• Spatial adaptivity in filtering parameters is crucial for enhancing image quality in low-dose CT.
• This approach holds significant potential for improving diagnostic accuracy in low-dose CT imaging.