Computational Cardiology is a research group advancing cardiovascular care through machine learning, natural language processing, and federated data infrastructure. The group designs EHR-embedded pragmatic trials and AI-driven tools for early detection, risk prediction, and personalized treatment of heart disease. Their projects include national screening programs, federated learning platforms, and cardioprotective drug trials for cancer patients.
Funding
Funding not disclosed
Founders
Product
Problem
Cardiovascular disease remains a leading cause of death, yet diagnosis and treatment often rely on fragmented data sources, delayed detection, and one-size-fits-all approaches. Researchers and clinicians lack integrated tools to leverage large-scale, multi-source data—including imaging, biomarkers, and electronic health records—for early risk assessment and personalized care.
Solution
Computational Cardiology develops and applies advanced computational methods to transform cardiovascular research and clinical practice. The group integrates machine learning, natural language processing, and federated data infrastructure to analyze diverse datasets, enabling predictive analytics, improved diagnostics, and personalized treatment strategies. They design EHR-embedded pragmatic trials that use real-world data for efficient, causally robust clinical research, and employ genetic methods to identify novel drug targets. Through collaborative projects like AI4HF and Check@Home, they deliver trustworthy AI solutions for heart failure risk assessment and national early-detection screening programs for cardiovascular, kidney, and metabolic diseases.
Target Audience
Primary beneficiaries are cardiologists, clinicians, data scientists, epidemiologists, and healthcare providers across academic medical centers and research consortia, as well as patients participating in screening and risk-assessment programs.
Features
- Federated learning platform for privacy-preserving, multi-site cardiology data analysis without centralizing sensitive patient information
- Natural language processing tools to extract structured insights from unstructured clinical narratives, improving diagnostics and automating medical coding
- AI-based electrocardiogram and medical imaging analysis for inherited cardiomyopathies and valvular heart disease
- Genetically guided drug target identification using genome-wide association studies, colocalization, and Mendelian randomization
- Design and execution of EHR-embedded pragmatic trials, including the HOVON 170 ANTICIPATE trial evaluating dexrazoxane for anthracycline-induced cardiotoxicity prevention
- Infrastructure for standardized data ingestion, harmonization, and a public data catalog for multi-source cardiology research