Multiclass plaque segmentation of coronary artery disease patients using an enhanced model
Related Experiment Video
Updated: Jul 10, 2026
Identifying Coronary Artery Calcification on Non-gated Computed Tomography Scans
Published on: August 28, 2018
Ting Chen1, Xing Wang2, Xinliu He1
1School of Medical Imaging, Department of Medical Technology, Tianjin Medical University, Tianjin, China.
Insights
Automated analysis of intravascular optical coherence tomography (IVOCT) data using an enhanced DeepLabV3+ model improves plaque segmentation for coronary artery disease (CAD) patients, aiding personalized treatment.
Area of Science:
- Cardiovascular Imaging
- Artificial Intelligence in Medicine
- Medical Image Analysis
Background:
- Coronary artery plaque analysis is crucial for personalized treatment in coronary artery disease (CAD).
- Intravascular optical coherence tomography (IVOCT) offers detailed plaque characterization.
Purpose of the Study:
- To develop automated methods for rapid and accurate analysis of IVOCT data.
- To enable cardiologists in providing tailored treatments for CAD patients.
Main Methods:
- Integration of two coordinate image types with an enhanced DeepLabV3+ model.
- Direct three-class segmentation for fibrous, lipid, and calcified plaques.
- Data split into 7:2:1 for training, validation, and testing.
Main Results:
- The enhanced DeepLabV3+ model achieved plaque-level F1 scores of 0.855 (fiber), 0.850 (lipid), and 0.551 (calcified).
Conclusions:
- The study demonstrates potential for improved accuracy and efficiency in CAD plaque segmentation.
- Enhanced automated analysis can lead to more personalized and effective patient treatments.
Background:
The analysis of coronary artery plaques in coronary artery disease (CAD) patients is instrumental in enabling cardiologists to administer personalized treatments and improve patient prognosis. Intravascular optical coherence tomography (IVOCT) provides precise insights into plaque characteristics.
Purpose:
Enable cardiologists to provide personalized treatment for patients with CAD; the objective is to explore methods for swiftly and accurately automating the analysis of IVOCT data.
Methods:
The study employs a methodology that integrates two types of coordinate images and enhances the DeepLabV3+ model. This approach facilitates direct three-class segmentation results for fibrous, lipid, and calcified plaques. Our experimental data is split into training, validation, and test sets in a ratio of 7:2:1.
Results:
The primary findings of the research indicate that the enhanced DeepLabV3+ model achieves F1 scores of 0.855, 0.850, and 0.551 for fiber, lipid, and calcified plaque detection at the plaque level, respectively.
Conclusions:
This study lies in its potential to improve the accuracy and efficiency of plaque segmentation in CAD patients, ultimately benefiting cardiologists by enabling more personalized and effective treatments.
More Related Videos
13:07Optical Coherence Tomography Based Biomechanical Fluid-Structure Interaction Analysis of Coronary Atherosclerosis Progression
Published on: January 15, 2022
06:18Intravascular Ultrasound Image-Based Finite Element Modeling Approach for Quantifying In Vivo Mechanical Properties of Human Coronary Artery
Published on: December 6, 2024
Related Concept Videos
01:25Imaging Studies for Cardiovascular System VI: Calcium -Scoring CT
01:30Coronary Artery Disease I: Introduction