01
Measure
Define outcomes, cohorts, missingness, data provenance, and subgroup performance before interpreting model results.
Pediatric research
Computational research designed around fairness and clinical usefulness.
CiDrep SickKids' current research mission focuses on building interpretable evidence and tools that can help children receive earlier and more equitable care.
Flagship program
Pediatric epilepsy
The Foundation's current flagship research program integrates electronic health record, neuroimaging, genomic, and geo-coded social-determinants-of-health data from the All of Us Research Program to develop fair, interpretable machine-learning approaches for pediatric epilepsy.
Research design principles
- Clinical relevance before model complexity
- Interpretability and transparent evaluation
- Structured fairness assessment
- Responsible multimodal data integration
- Reproducibility and evidence generation
Equity-first methods
Fairness is part of the research design.
The current Foundation description states that model development is paired with structured fairness review across race, ethnicity, rurality, and insurance status.
02
Interpret
Favor methods that support clinically meaningful explanation and transparent assessment of model behavior.
03
Audit
Assess performance and error patterns across relevant population groups rather than relying on aggregate accuracy alone.