X-ray angiogram images enhancement by facet-based adaptive anisotropic diffusion

Area of Science:

• Medical Imaging
• Image Processing
• Computational Science

Background:

• Traditional nonlinear diffusion effectively smooths images while preserving edges.
• Existing methods face challenges with parameter sensitivity, limited enhancement range, and evolution time selection.

Purpose of the Study:

• To develop a versatile nonlinear diffusion method for enhancing angiogram images.
• To address the drawbacks of traditional nonlinear diffusion models in clinical diagnosis.

Main Methods:

• Proposed a novel anisotropic diffusion method based on the facet model.
• Used facet model fitting for noise reduction and Hessian eigenvalue sums for adaptive conductance parameter selection.
• Demonstrated adaptive noise handling and parameter adjustment throughout the diffusion process.

Main Results:

• The new method adaptively adjusts to image content, overcoming parameter sensitivity.
• It shows improved noise reduction and edge enhancement capabilities.
• The method is robust to the choice of evolution time, unlike traditional approaches.

Conclusions:

• The proposed anisotropic diffusion method offers significant improvements over traditional methods for angiogram enhancement.
• This technique enhances visual quality for better clinical diagnosis.
• The adaptive nature of the method ensures reliable performance across diverse image content.