Two research projects involving the Health AI & Analytics Lab at the Management Science and Engineering Department at the University of Waterloo have been awarded funding in the third round of the Graham Seed Fund, which supports early‑stage, collaborative research aimed at addressing pressing challenges in healthcare delivery through applied, translational work. Both projects emerged from close collaborations between the Health AI and Analytics lab and the Waterloo Regional Health Network:
Reducing Chemotherapy Drug Wastage through Data‑Driven Operations
UWaterloo Investigators: Fatih Safa Erenay, Houra Mahmoudzadeh
Partner: Waterloo Regional Health Network
Chemotherapy drugs are supplied in vials of fixed sizes, which often do not match a patient’s prescribed dose. As a result, when only part of a vial is used and the remaining portion cannot be utilized in time, it must be discarded for safety reasons. Given the rapidly increasing cost of cancer drugs and the large number of treatments administered each year, this wastage places a significant financial burden on society, amounting to millions of dollars.
This project aims to reduce chemotherapy drug wastage by improving how cancer centres forecast drug demand, plan clinic schedules, and set inventory levels, while keeping the solution practical and avoiding added procedural complexity for frontline staff. Using historical treatment patterns and operational inputs, this project will support decisions such as better alignment of patient booking with drug preparation, safer vial sharing and batching opportunities, and more informed dose-rounding and closed-system transfer device (CSTD) planning where appropriate.
Project outcomes will be measured through reduced wastage volume and cost, and more stable drug availability. This aligns with the hospital’s mission and values by supporting high-quality, patient-centered care, improving operational efficiency and resource stewardship, and helping teams deliver more seamless, reliable treatment experiences.
A Socio-technical AI Readiness Framework for Workflow intelligence in Community Cardiac Care
UWaterloo Investigators: Sharon Ferguson, Ada Hurst, Houra Mahmoudzadeh
Partner: Waterloo Regional Health Network
Atrial fibrillation is the most common arrhythmia in Canada, with catheter ablation as a first-line therapy. Ontario patients face long wait times due to inefficiencies and variability in workflows within procedural care settings. Current workflow data are manually collected, incomplete, and error-prone, representing an opportunity for AI-enabled approaches to improve data capture and analysis.
This project explores the organizational, cultural, and practical AI-readiness conditions required for ambient AI data collection in community cardiac care, using Atrial Fibrillation ablation as a case study. This work will (1) map current ablation workflows; (2) elicit requirements for an ambient AI system with considerations for trust, usability, and privacy; (3) assess feasibility of using current language models to support future workflow intelligence system; and (4) synthesize an AI-readiness framework to guide workflow analysis in procedural care.
Using interviews, simulated procedures, and feasibility testing, the project will produce readiness factors and design guidelines for AI systems in community cardiac care that are grounded in real clinical and organizational conditions. The resulting framework aims to inform the responsible introduction of AI‑based workflow intelligence tools in cardiac care and similar procedural healthcare settings.
Together, these two projects reflect the Health AI & Analytics Lab’s focus on responsible, applied AI and analytics that address real‑world healthcare challenges by combining technical rigor with deep attention to operational, human, and organizational contexts.