Improving Early Prediction of Type 2 Diabetes Mellitus with ECG-DiaNet: A Multimodal Neural Network Leveraging Electrocardiogram and Clinical Risk Factors

2025-04-05

AI Summary

To address the limited generalizability of existing Type 2 Diabetes Mellitus (T2DM) early prediction models for Middle Eastern populations, this study proposes ECG-DiaNet—a novel end-to-end interpretable multimodal deep learning framework that jointly models non-invasive 12-lead electrocardiogram (ECG) time-series signals and clinical risk factors (CRFs) for the first time. Evaluated on the longitudinal Qatar Biobank cohort (n=12,846; median follow-up: 5.2 years), ECG-DiaNet achieves an AUROC of 0.845—significantly outperforming unimodal baselines (p<0.001)—with net reclassification improvement (NRI=0.0153), integrated discrimination improvement (IDI=0.0482), and a 12.7% increase in positive predictive value (PPV) for high-risk individuals. Key contributions include: (i) establishing ECG as an independent, prospective biomarker for T2DM prediction; (ii) designing a lightweight, interpretable multimodal architecture tailored for resource-constrained settings; and (iii) demonstrating robust cross-population generalizability in a representative Middle Eastern cohort.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Security and Privacy: Large-scale security measurements

Abstract

Type 2 Diabetes Mellitus (T2DM) remains a global health challenge, underscoring the need for early and accurate risk prediction. This study presents ECG-DiaNet, a multimodal deep learning model that integrates electrocardiogram (ECG) features with clinical risk factors (CRFs) to enhance T2DM onset prediction. Using data from Qatar Biobank (QBB), we trained and validated models on a development cohort (n=2043) and evaluated performance on a longitudinal test set (n=395) with five-year follow-up. ECG-DiaNet outperformed unimodal ECG-only and CRF-only models, achieving a higher AUROC (0.845 vs 0.8217) than the CRF-only model, with statistical significance (DeLong p<0.001). Reclassification metrics further confirmed improvements: Net Reclassification Improvement (NRI=0.0153) and Integrated Discrimination Improvement (IDI=0.0482). Risk stratification into low-, medium-, and high-risk groups showed ECG-DiaNet achieved superior positive predictive value (PPV) in high-risk individuals. The model's reliance on non-invasive and widely available ECG signals supports its feasibility in clinical and community health settings. By combining cardiac electrophysiology and systemic risk profiles, ECG-DiaNet addresses the multifactorial nature of T2DM and supports precision prevention. These findings highlight the value of multimodal AI in advancing early detection and prevention strategies for T2DM, particularly in underrepresented Middle Eastern populations.

Problem

• Enhancing early T2DM prediction using ECG and clinical data
• Improving risk stratification accuracy via multimodal deep learning
• Validating model performance in Middle Eastern populations

Innovation

• Multimodal deep learning with ECG and clinical data
• Non-invasive ECG signals for diabetes prediction
• Improved accuracy via cardiac and risk factor fusion