From Event: SPIE BiOS, 2022
In this project, we propose a deep learning-based method to detect calcified tissue within coronary optical coherence tomography (OCT) images. Conventionally, diseased tissue was manually checked on a frame (Bscan)-by-frame (Bscan) basis. Based on faster region-based convolutional neural network, our proposed method can automatically detect the calcified regions from diseased coronary artery. Our method achieves promising result of mean average precision (0.74) and recall (0.79) in detecting calcified regions. The proposed method could provide valuable information for locating calcified tissue within a large volume of OCT images. It has great potential to aid the treatment of coronary artery disease.
© (7 March 2022) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Hongshan Liu, Xuesheng Li, and Yu Gan, "Automated object detection within coronary optical coherence tomography images," Proc. SPIE PC11936, Diagnostic and Therapeutic Applications of Light in Cardiology 2022, PC119360E (Presented at SPIE BiOS: 7 March 2022; Published: 7 March 2022); https://doi.org/10.1117/12.2610257.6293554326001.