Related Experiment Video
Updated: Nov 8, 2025
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Coronary Vessel Segmentation by Coarse-to-Fine Strategy Using U-nets
Le Nhi Lam Thuy1,2, Tan Dat Trinh1, Le Hoang Anh1
1Information Science Faculty, Sai Gon University, Vietnam.
Insights
This study introduces a novel coarse-to-fine method for segmenting coronary arteries of varying sizes in angiograms. The approach effectively extracts both primary and secondary vessels, improving coronary artery segmentation accuracy.
Area of Science:
- Medical Imaging
- Cardiovascular Imaging
- Image Segmentation
Background:
- Coronary arteries vary significantly in size and visual contrast within angiograms.
- Segmenting all coronary artery sizes with a single model is challenging due to differing properties and fragmented vessels.
Purpose of the Study:
- To develop a novel, coarse-to-fine method for accurate coronary artery extraction from angiograms.
- To address the limitations of single-model segmentation for diverse coronary artery sizes.
Main Methods:
- A U-net model was employed for initial segmentation of the main coronary artery.
- A new algorithm identified junctions between primary and secondary coronary arteries.
- A second U-net model segmented secondary coronary arteries within defined regions.
Main Results:
- The proposed method achieved a Dice coefficient of 76.40% on coronary X-ray datasets.
- The coarse-to-fine strategy demonstrated effectiveness in segmenting varied coronary artery structures.
Conclusions:
- The novel segmentation approach shows significant potential for improving coronary vessel segmentation in medical imaging.
- This method offers a promising solution for analyzing coronary arteries of different scales.
Abstract:
Each level of the coronary artery has different sizes and properties. The primary coronary arteries usually have high contrast to the background, while the secondary coronary arteries have low contrast to the background and thin structures. Furthermore, several small vessels are disconnected or broken up vascular segments. It is a challenging task to use a single model to segment all coronary artery sizes. To overcome this problem, we propose a novel segmenting method for coronary artery extraction from angiograms based on the primary and secondary coronary artery. Our method is a coarse-to-fine strategic approach for extracting coronary arteries in many different sizes. We construct the first U-net model to segment the main coronary artery extraction and build a new algorithm to determine the junctions of the main coronary artery with the secondary coronary artery. Using these junctions, we determine regions of the secondary coronary arteries (rectangular regions) for a secondary coronary artery-extracted segment with the second U-net model. The experiment result is 76.40% in terms of Dice coefficient on coronary X-ray datasets. The proposed approach presents its potential in coronary vessel segmentation.
