A Framework for Automated Quantification of Calcified Coronary Artery from Intravascular Optical Coherence Tomography Images

Intravascular Optical Coherence Tomography (OCT) has emerged as a powerful imaging modality for assessing the morphological characteristics of coronary arteries. Quantification of calcified coronary arteries from OCT images is crucial for evaluating the severity and progression of coronary artery disease. However, in current practice, OCT images are interpreted manually which is time-consuming, subjective, and prone to inter- and intra-observer variability. To address these limitations, we propose a framework for automated quantification of calcified coronary arteries from OCT images. By leveraging deep learning techniques, the proposed framework automatically segments lumen and calcified plaque from OCT images. Subsequently, comprehensive morphological analysis of lumen and calcified plaque is performed using advanced image processing algorithms, allowing for retrieval of various dimensions of corresponding structures. Following that, essential shape measurements are derived to ensure adequate characterization of calcified coronary arteries. The efficacy of the proposed framework was validated on a clinical dataset. Extensive experiments have demonstrated high accuracy and consistency of quantitative results estimated by the proposed framework against manual analysis with relative errors of less than 10%. The proposed framework holds great potential to extend its application to characterization of other non-calcified plaques and arteries, aiding in clinical intervention and translational research using OCT.