Trustworthy AI
Dataset-shift diagnostics, calibration, uncertainty estimation, bias auditing, and post-deployment monitoring for clinical models.
Research
Health data science that earns trust
ISFHDS connects methodological research with clinical and public-health needs, supporting work that is rigorous, reproducible, and designed for real-world benefit.
Research priorities
Work across the health data lifecycle
Our focus spans evidence generation, responsible model development, and translation into clinical and health-system practice.
Privacy and security
Federated learning, secure aggregation, differential privacy, and governance controls for multi-institution health data.
Clinical NLP and multimodal models
Methods that responsibly combine clinical text, imaging, signals, and structured records while keeping evaluation clinically meaningful.
Real-world evidence
Causal inference and transparent observational analyses that make assumptions, provenance, and uncertainty visible.
Digital health
Remote monitoring and medical technologies evaluated for utility, safety, generalisability, and use in everyday care.
Genomics and multi-omics
Integration of molecular and clinical data to support interpretable discovery and translational research.
Health systems
Operations research, forecasting, and decision support for resilient and efficient health services.
Health equity
Analyses that examine performance across populations and account for social determinants of health.
From methods to practice
Shared standards for useful evidence
Research becomes more valuable when methods, evaluation, and implementation can be examined together.
Reproducible by design
Document data provenance, analytical decisions, model versions, and evaluation protocols so results can be checked and extended.
Clinically grounded
Frame questions with clinicians and stakeholders, then evaluate against outcomes and workflows that matter in practice.
Responsible translation
Plan for safety, fairness, privacy, and monitoring throughout development—not only when a system is ready to deploy.
Collaborate
Bring a question to the community
Connect with ISFHDS about research exchange, working groups, training, or potential collaborations.