Related Experiment Video
Updated: Jun 3, 2025
Three-Dimensional Printing of a Complex Aortic Anomaly
Published on: November 1, 2018
NeCA: 3D Coronary Artery Tree Reconstruction from Two 2D Projections via Neural Implicit Representation
Yiying Wang1, Abhirup Banerjee1,2, Vicente Grau1
1Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, Oxford OX3 7DQ, UK.
Insights
This study introduces NeCA, a self-supervised deep learning method for 3D coronary artery tree reconstruction from two 2D projections. NeCA effectively preserves vessel topology and connectivity without needing 3D data for training.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Cardiovascular Medicine
Background:
- Cardiovascular diseases (CVDs) are a leading global health concern.
- 2D X-ray invasive coronary angiography (ICA) is standard for CVD assessment but struggles with 3D vessel geometry interpretation.
- Limited projections in ICA restrict 3D coronary tree reconstruction.
Purpose of the Study:
- To develop a self-supervised deep learning method for 3D coronary artery tree reconstruction from limited 2D projections.
- To overcome the limitations of interpreting 3D coronary vessel geometry from 2D invasive coronary angiography.
- To enable accurate 3D reconstruction without requiring 3D ground truth data.
Main Methods:
- Proposed NeCA, a self-supervised deep learning approach utilizing neural implicit representation.
- Employed a multiresolution hash encoder and a differentiable cone-beam forward projector layer.
- Validated the method on a dataset derived from coronary computed tomography angiography (CCTA).
Main Results:
- NeCA achieved promising performance in 3D coronary artery tree reconstruction.
- Demonstrated effective preservation of vessel topology and branch connectivity.
- Outperformed supervised deep learning models without requiring 3D ground truth or large datasets.
Conclusions:
- NeCA offers a novel self-supervised solution for 3D coronary artery reconstruction from limited 2D ICA data.
- The method shows potential for improving 3D visualization in cardiac interventions.
- NeCA's ability to work without extensive training data is a significant advantage.
Abstract:
Cardiovascular diseases (CVDs) are the most common health threats worldwide. 2D X-ray invasive coronary angiography (ICA) remains the most widely adopted imaging modality for CVD assessment during real-time cardiac interventions. However, it is often difficult for the cardiologists to interpret the 3D geometry of coronary vessels based on 2D planes. Moreover, due to the radiation limit, often only two angiographic projections are acquired, providing limited information of the vessel geometry and necessitating 3D coronary tree reconstruction based only on two ICA projections. In this paper, we propose a self-supervised deep learning method called NeCA, which is based on neural implicit representation using the multiresolution hash encoder and differentiable cone-beam forward projector layer, in order to achieve 3D coronary artery tree reconstruction from two 2D projections. We validate our method using six different metrics on a dataset generated from coronary computed tomography angiography of right coronary artery and left anterior descending artery. The evaluation results demonstrate that our NeCA method, without requiring 3D ground truth for supervision or large datasets for training, achieves promising performance in both vessel topology and branch-connectivity preservation compared to the supervised deep learning model.
