XOCT: Enhancing OCT to OCTA Translation via Cross-Dimensional Supervised Multi-Scale Feature Learning

2025-09-09

AI Summary

OCT angiography (OCTA) suffers from high motion sensitivity, expensive hardware, and limited capability of existing deep learning methods to accurately reconstruct inter-layer vascular structures and fine capillary details. To address these challenges, we propose the Cross-dimensional Supervision and Multi-scale Feature fusion Network (CSMF-Net). Our method introduces hierarchical 2D en-face projections as cross-dimensional supervision signals to guide learning of layer-specific vascular representations. It further integrates multi-scale feature extraction, channel-wise reweighting, and segmentation-weighted z-axis averaging to achieve high-fidelity end-to-end OCT-to-OCTA translation. Evaluated on the OCTA-500 dataset, CSMF-Net significantly improves structural fidelity and clinical utility of synthesized OCTA images—particularly enhancing vessel continuity, contrast, and layer separation in en-face projections—outperforming state-of-the-art methods. This advancement enhances OCTA accessibility and diagnostic value without requiring additional hardware or motion compensation.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Feature Construction/Reformulation
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Optical Coherence Tomography Angiography (OCTA) and its derived en-face projections provide high-resolution visualization of the retinal and choroidal vasculature, which is critical for the rapid and accurate diagnosis of retinal diseases. However, acquiring high-quality OCTA images is challenging due to motion sensitivity and the high costs associated with software modifications for conventional OCT devices. Moreover, current deep learning methods for OCT-to-OCTA translation often overlook the vascular differences across retinal layers and struggle to reconstruct the intricate, dense vascular details necessary for reliable diagnosis. To overcome these limitations, we propose XOCT, a novel deep learning framework that integrates Cross-Dimensional Supervision (CDS) with a Multi-Scale Feature Fusion (MSFF) network for layer-aware vascular reconstruction. Our CDS module leverages 2D layer-wise en-face projections, generated via segmentation-weighted z-axis averaging, as supervisory signals to compel the network to learn distinct representations for each retinal layer through fine-grained, targeted guidance. Meanwhile, the MSFF module enhances vessel delineation through multi-scale feature extraction combined with a channel reweighting strategy, effectively capturing vascular details at multiple spatial scales. Our experiments on the OCTA-500 dataset demonstrate XOCT's improvements, especially for the en-face projections which are significant for clinical evaluation of retinal pathologies, underscoring its potential to enhance OCTA accessibility, reliability, and diagnostic value for ophthalmic disease detection and monitoring. The code is available at https://github.com/uci-cbcl/XOCT.

Problem

• Enhances OCT to OCTA translation for retinal vasculature
• Addresses motion sensitivity and high costs in OCTA acquisition
• Improves reconstruction of intricate vascular details across layers

Innovation

• Cross-Dimensional Supervision for layer-aware reconstruction
• Multi-Scale Feature Fusion with channel reweighting
• Segmentation-weighted z-axis averaging for en-face projections