AI Foundations Lab

BORN’s AI Foundations Lab sits at the intersection of artificial intelligence, population-based data, and perinatal/child health. The Lab combines expertise in machine learning, health informatics, and population-based screening to improve care and outcomes for pregnant individuals and their children.

Co-led by Drs Christine Armour and Kevin Dick, and governed by CHEO’s framework for leading responsibly with AI, the AI Foundations Lab leverages this expertise along with BORN’s population-based data assets (>2 million linkable perinatal/child records for nearly every birth in Ontario since 2012) and on-prem 2-node AI compute cluster to develop and train predictive population health models. As a prescribed registry, BORN is uniquely positioned to also deploy AI models and improve care and outcomes across Ontario, leveraging Prenatal Screening Ontario’s oversight role and by building a digital, interoperable, and intelligent perinatal record for the province.

In addition to applied predictive health modelling, the AI Foundations Lab also addresses central challenges for machine learning models in perinatal and child health:

Outcome rarity: health conditions of interest are often rare by machine learning standards (prevalence <10%).
Trust and equity: integrating explainable AI and equity considerations into modelling frameworks to foster trust and equity
Data privacy and cost-effectiveness: creating integrated, privacy-first AI infrastructure through federated learning, multimodal development, and cryptography

AI

The Lab's research focuses on developing novel artificial intelligence (AI) and machine learning (ML) methodologies. Specializing primarily in creating robust, interpretable AI (e.g., XAI) solutions that address complex challenges in early life health outcomes. The Lab's work spans from theoretical frameworks to practical implementations, leveraging high-performance computing (HPC) infrastructure for large-scale health data analysis.

Population Screening

The Lab's research leverages population-wide maternal-newborn datasets to develop and validate AI-based screening methods that improve healthcare outcomes. Embedded at BORN Ontario, The Lab analyzes comprehensive registry data to uncover outcome patterns, identify risk factors, and reveal opportunities for enhanced care delivery. As a prescribed registry, BORN Ontario has the distinct capability to translate research findings into direct clinical impact.

Maternal-Newborn Healthcare

The Lab's work spans from pregnancy through early childhood, addressing critical challenges in maternal-newborn health surveillance, risk assessment, and outcome prediction.

The Lab is committed to developing equitable AI solutions that serve diverse populations and communities. Through careful consideration of bias, fairness, and accessibility, The Lab ensures our innovations benefit all families while maintaining the highest standards of privacy and ethical data use.

Predictive Algorithms