AI
AI & Explainable Machine Learning
Predictive modeling, explainable machine learning and human-centred decision support for complex clinical data.
Artificial intelligence · evidence · healthcare
Computational science for safer, smarter healthcare.
AI4Health is a collaborative research group that develops artificial intelligence, high-throughput computing, automation and visual analytics to generate rigorous, interpretable evidence from complex health data.
From real-world health data and regulatory documents to interpretable models and actionable evidence.
Evidence engine
EvidenceHealth records & real-world data
SafetyFAERS, vigilance & monographs
MethodsAI · XAI · causal methods
ScaleAutomation & high-throughput compute
AI4Health
Research
Research
What we do
From data to evidence, at scale.
Our work combines computational innovation with rigorous clinical and pharmacoepidemiologic methods. We develop models, reusable pipelines and interactive systems, then apply them to consequential healthcare questions.
HT
High-Throughput Computing & Automation
Reusable computational pipelines that scale population-based safety studies, interaction screening and repeated evidence generation.
Rx
Drug Safety & Real-World Evidence
Population-based medication safety, pharmacovigilance, drug-drug interactions and evidence generation in high-risk populations.
VA
Visual Analytics
Interactive systems that help researchers explore high-dimensional, temporal, multimorbidity and incomplete health data.
Research software in practice
We build tools—not just models.
Our published visual analytics systems span EHR exploration, medication safety, disease prediction, multimorbidity and pharmacovigilance. Published system figures are shown where available.
153publications indexed from supplied Google Scholar exports
6published visual analytics systems
9interconnected research programs
2complementary disciplinary perspectives
Evidence journey
We connect the evidence lifecycle.
A core AI4Health goal is to connect evidence sources that are often studied separately and use computational methods to make the full evidence lifecycle more scalable, reproducible and interpretable.
01Clinical trialsWhat was tested, and in whom?
02Regulatory knowledgeLabels, monographs, warnings and terminology.
03Real-world dataPopulation-based outcomes in routine care.
04Safety reportsPost-market spontaneous reporting signals.
05AI & computationScale, automate, explain and connect evidence.
06Actionable evidenceSafer prescribing and better research decisions.
Collaborate
Have a health-data or medication-safety question?
We welcome conversations about methodological collaboration, research software, high-throughput studies, AI, drug safety, trials and real-world evidence.