Multi-modal Retinal Image Registration Using a Keypoint-Based Vessel Structure Aligning Network

TL;DR

This paper introduces a deep learning-based method for accurately aligning multi-modal retinal images, aiding diagnosis by providing pixel-level comparison of vessel structures across different imaging modalities.

Contribution

It presents an end-to-end trainable network combining vessel structure feature extraction, keypoint detection, and graph neural network-based matching, trained self-supervised on synthetic data.

Findings

Outperforms existing methods on synthetic retinal datasets.

Generalizes effectively to real macula and public fundus datasets.

Achieves higher registration accuracy in multi-modal retinal image alignment.

Abstract

In ophthalmological imaging, multiple imaging systems, such as color fundus, infrared, fluorescein angiography, optical coherence tomography (OCT) or OCT angiography, are often involved to make a diagnosis of retinal disease. Multi-modal retinal registration techniques can assist ophthalmologists by providing a pixel-based comparison of aligned vessel structures in images from different modalities or acquisition times. To this end, we propose an end-to-end trainable deep learning method for multi-modal retinal image registration. Our method extracts convolutional features from the vessel structure for keypoint detection and description and uses a graph neural network for feature matching. The keypoint detection and description network and graph neural network are jointly trained in a self-supervised manner using synthetic multi-modal image pairs and are guided by synthetically sampled ground truth homographies. Our method demonstrates higher registration accuracy as competing methods for our synthetic retinal dataset and generalizes well for our real macula dataset and a public fundus dataset.