Research Questions:
1. How do biomarkers, lifestyle factors, imaging, and multi-omic profiles contribute to the incidence, severity, and mortality of cardiovascular, cerebrovascular, renal diseases, metabolic syndrome, and their comorbidities?
2. How do these risk factor effects differ between European and Asian populations?
Aims and Objectives:
1. Integrate biomarkers, lifestyle, imaging, and multi-omic data to identify key risk factors and characterize disease onset and progression trajectories.
2. Develop and validate deep learning-based multimodal predictive models in Asian populations to assess generalizability.
3. Compare multimodal models with conventional clinical prediction tools to optimize risk stratification, identify high-risk subgroups, and track disease outcomes.
4. Apply advanced statistical methods, such as trial emulation, in prospective cohorts to generate clinically actionable evidence.
5. Quantify residual cardiovascular risk beyond traditional factors.
Scientific Rationale!
Current clinical prediction models for these diseases are limited by reliance on single biomarkers or clinical features, which cannot capture long-term disease trajectories, restricting their use in primary, secondary, and tertiary prevention. Population-specific differences in disease spectra and risk factors further limit the applicability of existing models across ethnic groups and regions.
This project will leverage UK Biobank data alongside our own cohort to develop a robust, clinically relevant prediction framework. By integrating multimodal data-including imaging, genomic, and proteomic information-this study aims to produce accurate, generalizable, and personalized risk scores, addressing gaps in previous research and enhancing precision medicine in underrepresented populations.
