Automated coronary artery tree segmentation in coronary CTA using a multiobjective clustering and toroidal...

Area of Science:

• Medical Imaging
• Cardiovascular Research
• Image Analysis

Background:

• Coronary artery segmentation is crucial for diagnosing coronary artery disease.
• Challenges include image noise, low contrast, and complex vessel structures.
• Accurate segmentation aids radiologists in clinical detection.

Purpose of the Study:

• To develop an accurate method for segmenting 3D coronary artery trees from CTA images.
• To overcome segmentation challenges posed by medical image quality and anatomical complexity.

Main Methods:

• A novel framework combining multiobjective clustering and toroidal model-guided tracking.
• Integrated noise reduction, candidate region detection, and geometric feature extraction.
• Candidate regions identified via multiobjective clustering; arteries tracked using toroidal model guidance.

Main Results:

• The proposed framework demonstrates superior performance compared to existing methods.
• Achieved a mean Dice Similarity Coefficient (DSC) of 84%, Jaccard index of 74%, and Recall of 93%.
• Qualitative and quantitative results validate the method's effectiveness on CTA data.

Conclusions:

• The developed segmentation framework accurately segments coronary artery trees from CTA volumes.
• This advancement enhances the precision of 3D vascular tree segmentation for clinical applications.