AI-based detection and classification of anomalous aortic origin of coronary arteries using coronary CT angiography...

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Anomalous aortic origin of the coronary artery (AAOCA) is a rare congenital heart defect.
• AAOCA can lead to significant morbidity and mortality, including ischemia and sudden cardiac death.
• Current diagnostic methods using coronary CT angiography (CCTA) may overlook or misclassify AAOCA.

Purpose of the Study:

• To develop and validate a fully automated artificial intelligence (AI) tool for detecting and classifying AAOCA.
• To assess the performance of the AI tool in identifying AAOCA in 3D-CCTA images.
• To evaluate the potential clinical utility of the AI tool in managing AAOCA.

Main Methods:

• Development of a fully automated AI algorithm for AAOCA detection and classification.
• Validation of the AI tool using internal and external datasets of 3D-CCTA images.
• Clinical evaluation of the AI tool's performance, including discriminatory power (AUC), sensitivity, and specificity.

Main Results:

• The AI tool achieved high discriminatory performance with an AUC ≥ 0.99.
• Sensitivity and specificity ranged from 0.95 to 0.99 across all testing datasets.
• The AI model demonstrated accurate and automated detection and classification of AAOCA.

Conclusions:

• The developed AI tool provides a fully automated and accurate method for AAOCA detection and classification in CCTA.
• This AI tool has the potential for seamless integration into clinical workflows, offering real-time alerts for high-risk anatomies.
• The tool can facilitate the analysis of large CCTA cohorts, enhancing understanding and management of AAOCA.