NeCA: 3D Coronary Artery Tree Reconstruction from Two 2D Projections via Neural Implicit Representation

TL;DR

NeCA tackles the problem of reconstructing a 3D coronary artery tree from only two X-ray angiography projections by employing a self-supervised neural implicit representation. It combines a multiresolution hash encoder with a residual MLP to encode 3D coordinates into occupancy values, and leverages a differentiable cone-beam forward projector to enforce projection consistency without requiring 3D ground truth. Evaluated on RCA and LAD data from a public CCTA dataset, NeCA (and especially NeCA with orthogonal views) outperforms a supervised 3D U-Net baseline across multiple topology and connectivity metrics, demonstrating strong potential for per-subject optimization with limited data. The approach offers a path toward accurate 3D vascular reconstruction in clinical settings with reduced data requirements, albeit with longer inference times and considerations for real-world attenuation variability and imaging conditions.

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 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.

Figures (8)

• Figure 1: An example of two projections generated from RCA and LAD data.
• Figure 2: The proposed NeCA model (stages 2--5). The multiresolution hash encoder illustrates an example of $2$ resolution levels (coloured in green and blue) from fine to coarse resolution for one sampled point (in black).
• Figure 3: The results of all six metrics every 100 iterations for two RCA example data points ($R_1$ and $R_2$) using our NeCA model with two clinical-angle projections.
• Figure 4: The quantitative results of our NeCA model over two LAD example data points ($L_2$ and $L_3$) every 100 iterations with respect to all 6 metrics and evaluated with 2 clinical-angle projections.
• Figure 5: Five qualitative results of 3D RCA reconstruction. From left to right: five RCA data points $R_{1,2,3,4,5}$. From top to bottom: the reconstruction results from our NeCA model, NeCA ($90\degree$), 3D U-Net model, and the corresponding ground truth (GT).