aiMCard: Transformative healthcare AI multi-Modal tool for integrated Cardiometabolic risk prediction in Qatar

RESEARCH IN CONTEXT: Cardiometabolic disease (CMD), linked to cardiovascular disease (CVD) and type 2 diabetes (T2D), is a major public health challenge. In Qatar, T2D prevalence was 14.0% in 2021, projected to rise to 28.1% by 2050, while CVD caused 33% of deaths in 2019. Current CMD risk stratification relies on conventional factors, which may not fully capture disease heterogeneity and are often based on European populations, limiting their applicability to diverse groups like Qataris. THE KEY QUESTION: There is a need for advanced, multimodal AI tools integrating EHR, clinical/biochemical data, genetics, multi-omics (transcriptomics, proteomics), imaging (ECG, MRI, iDXA), and lifestyle factors to predict CMD accurately across diverse populations. PROPOSED RESEARCH: We propose to: (I) Develop a novel AI framework for CMD prediction tailored to Qataris using 25,000 QPHI samples, (II) Validate it in a cohort of incident CMD cases, and (III) Enhance its performance using multi-ancestry biobanks. Our AI model will integrate enriched polygenic risk scores (PRS), multi-omics, clinical/biochemical data, imaging-derived body fat measures, and lifestyle factors. We will validate the framework in 300 QPHI participants with transcriptomic/proteomic profiles and calibrate it using external biobanks (e.g., UK Biobank) to ensure precision and transferability. This multimodal AI framework aims to enable earlier, more effective CMD prevention strategies in clinical practice.

Hamad Bin Khalifa University (HBKU)