Multimodal Lead-Specific Modeling of ECG for Low-Cost Pulmonary Hypertension Assessment

2025-03-03

arXiv.org

AI Summary

Early screening for pulmonary hypertension (PH) in resource-limited settings is hindered by reliance on centralized 12-lead electrocardiography (ECG), while portable 6-lead ECG—despite its accessibility—suffers from severe scarcity of labeled data, impeding robust model development. Method: We propose a Lead-Specific Ensemble Multimodal Variational Autoencoder (LS-EMVAE), treating each ECG lead as an independent modality. It hierarchically fuses lead-specific and shared representations via mixture-of-experts and product-of-experts mechanisms, integrated within a transfer learning paradigm: pretraining on 12L-ECG followed by fine-tuning on 6L-ECG. Contribution/Results: Evaluated on 892 PH detection and 691 PH subtyping cases, LS-EMVAE significantly outperforms all baselines. Notably, its 6L-ECG performance matches that of 12L-ECG, demonstrating clinical feasibility of low-cost, portable PH screening and subtyping—enabling scalable deployment in underserved regions.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Pulmonary hypertension (PH) is frequently underdiagnosed in low- and middle-income countries (LMICs) primarily due to the scarcity of advanced diagnostic tools. Several studies in PH have applied machine learning to low-cost diagnostic tools like 12-lead ECG (12L-ECG), but they mainly focus on areas with limited resources, overlooking areas with no diagnostic tools, such as rural primary healthcare in LMICs. Recent studies have shown the effectiveness of 6-lead ECG (6L-ECG), as a cheaper and portable alternative in detecting various cardiac conditions, but its clinical value for PH detection is not well proved. Furthermore, existing methods treat 12L-/6L-ECG as a single modality, capturing only shared features while overlooking lead-specific features essential for identifying complex cardiac hemodynamic changes. In this paper, we propose Lead-Specific Electrocardiogram Multimodal Variational Autoencoder (LS-EMVAE), a model pre-trained on large-population 12L-ECG data and fine-tuned on task-specific data (12L-ECG or 6L-ECG). LS-EMVAE models each 12L-ECG lead as a separate modality and introduces a hierarchical expert composition using Mixture and Product of Experts for adaptive latent feature fusion between lead-specific and shared features. Unlike existing approaches, LS-EMVAE makes better predictions on both 12L-ECG and 6L-ECG at inference, making it an equitable solution for areas with limited or no diagnostic tools. We pre-trained LS-EMVAE on 800,000 publicly available 12L-ECG samples and fine-tuned it for two tasks: 1) PH detection and 2) phenotyping pre-/post-capillary PH, on in-house datasets of 892 and 691 subjects across 12L-ECG and 6L-ECG settings. Extensive experiments show that LS-EMVAE outperforms existing baselines in both ECG settings, while 6L-ECG achieves performance comparable to 12L-ECG, unlocking its potential for global PH screening in areas without diagnostic tools.

Problem

• Early pulmonary hypertension assessment in decentralized clinical settings
• Overcoming data scarcity for 6-lead ECG model development
• Improving multimodal fusion of ECG leads for PH detection

Innovation

• Hierarchical expert fusion mechanism for adaptive latent fusion
• Transfer learning from 12-lead to 6-lead ECG data
• Latent representation alignment loss for coherence improvement