Review and recommendations for using artificial intelligence in intracoronary optical coherence tomography analysis

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Artificial intelligence (AI) offers potential for diagnosing coronary artery disease (CAD) from intravascular optical coherent tomography (IVOCT) images.
• Existing AI models for CAD diagnosis using IVOCT require rigorous validation for clinical application.

Purpose of the Study:

• To systematically review AI-based models for CAD diagnosis using IVOCT.
• To assess the clinical utility and validation status of these AI models.
• To provide recommendations for improving AI model quality and research practices.

Main Methods:

• Systematic literature review of studies published between January 2015 and December 2024.
• Search identified 8600 studies; 629 were included after initial screening.
• 39 studies were included in the final review after quality assessment.

Main Results:

• Most AI models reviewed are not currently suitable for clinical use.
• Methodological flaws and underlying biases were identified as primary limitations.
• Significant variability exists in the validation of AI models for CAD diagnosis.

Conclusions:

• Current AI models for IVOCT-based CAD diagnosis require substantial improvement.
• Recommendations are provided to enhance model development and research practices.
• Future efforts should focus on rigorous validation to ensure clinical utility of AI tools.