Department of Medicine · Western University
Applying AI to improve patient care and support physicians.
Western AI is a clinical research lab applying large language models and deep learning networks to enhance the quality of care, reduce the documentation burden on physicians, and advance medical research at London Health Sciences Centre.
Our mission
Operationalizing AI for everyday clinical practice.
Based in the Department of Medicine at Western University, our team tests the use of large language models in physician practices to improve the quality of care, reduce the documentation burden, and support clinical decision-making. We are committed to translating modern AI into tools that genuinely improve outcomes for patients and the working lives of physicians.
Our work spans two complementary areas: applied clinical AI deployed directly into care settings, and deep learning research that uncovers new diagnostic and predictive insight from medical data.
Improve quality of care
We build and evaluate clinical decision-support tools that help physicians make better, faster, and more consistent decisions for their patients.
Reduce physician burden
By automating documentation and streamlining workflows, we give clinicians back time and reduce the administrative load that contributes to burnout.
Advance rigorous research
Every tool we deploy is studied prospectively in real clinical settings, so adoption is grounded in evidence rather than hype.
How it works
From clinical need to evidence-based deployment.
We follow a disciplined path from problem to proof. Our research spans applied clinical AI deployed at the point of care and deep learning models that extract new insight from medical data.
Identify a clinical problem
We start with real friction points raised by physicians — documentation overload, decision-making under pressure, or gaps in diagnosis.
Build a focused AI tool
We develop targeted large language model and deep learning solutions, designed to fit cleanly into existing clinical workflows.
Evaluate prospectively
Each tool is studied in live clinical settings to measure its real impact on care quality, physician experience, and patient outcomes.
Applied clinical AI
Tools that work alongside clinicians and patients in active care settings.
AI-CLINIC
Evaluating an AI voice-to-voice model in our cardiology clinics to support physician–patient interactions.
Patient education chatbot
An interactive chatbot that helps teach patients their discharge instructions after a cardiac care unit stay.
Emergency room intake
An AI voice-to-voice model being evaluated in the emergency room to gather patient history and improve flow for emergency physicians.
Deep learning
Models that surface diagnostic and predictive insight from clinical data.
AI-CARE
Using large language models to parse cardiology physician notes, extracting clinical variables to predict how patients with arrhythmias traverse the healthcare system.
AI-PVI
Predicting outcomes of pulmonary vein isolation in patients using surface ECG and deep learning models.
AI-PICM
Using baseline and paced surface ECGs with deep learning to identify pacing-induced cardiomyopathy.
Secure data repository
Building a local Western University data repository to securely and safely train AI models within the institution.
Our team
Clinicians, scientists, and trainees working side by side.
Western AI brings together physicians, machine learning researchers, and clinical trainees in the Department of Medicine at Western University, united by a commitment to rigorous, patient-centered research.
Dr. Pavel Antiperovitch, MD, FRCPC
Principal Investigator, Cardiology
Dr. Meichen Liu, PhD
Machine Learning Scientist
Dr. Iris Liu
Clinical Researcher
Dr. Bella (Chuce) Xing, IM3
Internal Medicine Resident
Dr. Kiro Hana, M2
Medical Student Researcher
Andrea Carruthers
Research Coordinator, LHRP Research