Automated Coronary Optical Coherence Tomography Feature Extraction with Application to Three-Dimensional...

Area of Science:

• Cardiovascular Imaging
• Medical Image Analysis
• Computational Biology

Background:

• Coronary optical coherence tomography (OCT) provides high-resolution intravascular imaging.
• Morphological plaque features alone are insufficient for predicting major adverse cardiovascular events (MACE).
• Accurate 3D simulations require detailed arterial morphology, often manually extracted from OCT, which is time-consuming.

Purpose of the Study:

• To systematically review automated segmentation techniques for OCT data from 2016-2021.
• To focus on applications for 3D vascular reconstruction and biomechanical simulation.
• To identify areas for future innovation in automated OCT analysis.

Main Methods:

• Systematic literature review of automated segmentation techniques for coronary OCT.
• Categorization of techniques based on segmented features: lumen, artery layers, plaque, and stents.
• Analysis of methods addressing OCT imaging challenges like limited penetration depth.

Main Results:

• Numerous automated segmentation techniques have emerged to overcome manual extraction limitations.
• These techniques facilitate 3D reconstruction of coronary arteries for biomechanical simulations.
• Advances cover segmentation of lumen, layers, plaque characteristics, and stents.

Conclusions:

• Automated segmentation of OCT data is crucial for accurate 3D vascular reconstruction and biomechanical simulation.
• These advancements hold significant potential for predicting plaque progression and MACE.
• Further innovation is needed for enhanced clinical translation of these techniques.