VAMPIRE: Uncovering Vessel Directional and Morphological Information from OCTA Images for Cardiovascular Disease Risk Factor Prediction

2025-07-26

AI Summary

Current cardiovascular disease (CVD) risk assessment methods suffer from two key limitations: fundus photography and OCT lack the microvascular structural resolution offered by optical coherence tomography angiography (OCTA); moreover, most approaches employ binary risk classification, neglecting the continuous pathological relationships between CVD and circulating biomarkers—thereby compromising predictive accuracy and clinical utility. To address these gaps, we introduce OCTA-CVD, the first publicly available OCTA-based fundus image dataset for CVD risk assessment. We further propose VAMPIRE, a novel multimodal model integrating a Mamba-Based Directional module—capturing dynamic vascular orientation patterns—and an Information-Enhanced Morphological module—incorporating domain-specific vascular morphology priors. VAMPIRE jointly predicts both CVD risk grades and quantitative levels of critical blood biomarkers in a multi-task learning framework. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art backbone networks, dedicated OCTA analysis methods, and ophthalmic foundation models across multiple evaluation metrics.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Data Mining & Knowledge Management: Mining of Visual, Multimedia & Multimodal Data

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Cardiovascular disease (CVD) remains the leading cause of death worldwide, requiring urgent development of effective risk assessment methods for timely intervention. While current research has introduced non-invasive and efficient approaches to predict CVD risk from retinal imaging with deep learning models, the commonly used fundus photographs and Optical Coherence Tomography (OCT) fail to capture detailed vascular features critical for CVD assessment compared with OCT angiography (OCTA) images. Moreover, existing methods typically classify CVD risk only as high or low, without providing a deeper analysis on CVD-related blood factor conditions, thus limiting prediction accuracy and clinical utility. As a result, we propose a novel multi-purpose paradigm of CVD risk assessment that jointly performs CVD risk and CVD-related condition prediction, aligning with clinical experiences. Based on this core idea, we introduce OCTA-CVD, the first OCTA dataset for CVD risk assessment, and a Vessel-Aware Mamba-based Prediction model with Informative Enhancement (VAMPIRE) based on OCTA enface images. Our proposed model aims to extract crucial vascular characteristics through two key components: (1) a Mamba-Based Directional (MBD) Module that captures fine-grained vascular trajectory features and (2) an Information-Enhanced Morphological (IEM) Module that incorporates comprehensive vessel morphology knowledge. Experimental results demonstrate that our method can surpass standard classification backbones, OCTA-based detection methods, and ophthalmologic foundation models. Our codes and the collected OCTA-CVD dataset are available at https://github.com/xmed-lab/VAMPIRE.

Problem

• Predict CVD risk using detailed OCTA vascular features
• Improve CVD risk assessment with multi-purpose prediction
• Extract vessel directional and morphological information from OCTA

Innovation

• Mamba-Based Directional Module captures vascular trajectories
• Information-Enhanced Morphological Module analyzes vessel morphology
• VAMPIRE model integrates OCTA images for CVD prediction