Segmentation and Automatic Identification of Vasculature in Coronary Angiograms

Area of Science:

• Medical Imaging
• Cardiovascular Diagnostics
• Image Processing

Background:

• Coronary angiography is crucial for diagnosing coronary heart disease.
• Challenges in coronary artery analysis include complex structures and image noise.
• Accurate vessel segmentation and identification are vital for diagnosis.

Purpose of the Study:

• To develop an effective preprocessing scheme for enhancing coronary angiogram quality.
• To accurately segment coronary vessels using the C-V model.
• To propose an improved adaptive tracking algorithm for automatic vascular skeleton identification.

Main Methods:

• Image preprocessing techniques: block-matching and 3D filtering, unsharp masking, contrast-limited adaptive histogram equalization, and multiscale image enhancement.
• Vessel segmentation using the C-V model for contour extraction.
• Development of an improved adaptive tracking algorithm for vascular skeleton identification.

Main Results:

• The preprocessing scheme successfully enhances vascular structures and suppresses background noise.
• The C-V model accurately extracts continuous vessel contours.
• The proposed tracking method demonstrates higher accuracy and robustness than existing methods.

Conclusions:

• The integrated approach significantly improves coronary angiogram quality and analysis.
• Accurate vessel segmentation and skeleton identification are achievable with the proposed methods.
• This technique offers a more robust and accurate tool for coronary heart disease diagnosis.