How Artificial Intelligence (AI) Is Helping Advance Cancer Care
Hello BRIGHT Run Family,
I hope you are doing well. We experienced very cold temperatures in January and got to clean up a lot of snow. After two relatively mild winters, this is a change but that is a part of life.
During one of our regular seminars at the Escarpment Cancer Research Institute in January, one of our colleagues mentioned that blue Monday, the third Monday of January, is often dubbed as the saddest day of the year and then we ended the seminar with a lot of hope based on progress reported by our colleagues. As I thought more about the progress, it crossed my mind that I haven’t shared an interesting research collaboration with you before.
Let’s start with a discussion on this research area. It focuses on developing clinical practice guidelines (CPGs), which are evidence‑based recommendations that help clinicians diagnose, treat, and care for patients.
Evidence isn’t simply “good or bad,” it’s ranked by levels that reflect how strong and trustworthy each scientific study is. Stronger evidence gives clinicians and policymakers more confidence when making recommendations.
Before gathering evidence, you need to define the clinical question (basically, what is the evidence for) and context (is the question important, relevant, and timely). Then you search for all potential scientific studies, screen out the irrelevant ones, and rank the remaining relevant studies.
In practice, searching for evidence produces thousands of articles, most of them irrelevant. Researchers spend enormous amounts of time manually (by critically reading) filtering these down to a small number of useful studies. The questions are: can AI help reduce this manual workload and how consistently can AI do this?
That is exactly what my colleague and collaborator Dr. Xiaomei Yao is studying. One of our datasets comes from a complex breast cancer CPG, making it an excellent test case for evaluating how well AI models can handle this very challenging task.
Over the past two years, I enjoyed learning from Dr. Yao regarding some of the complexities of CPG development. Dr. Yao enjoyed learning about the usage of AI tools in her domain. This helped both of us to critically think about our problem.
As both of us continue to learn from our joint work, we also published a protocol regarding how to systematically run our experiments (A Study Protocol for a Comprehensive Evaluation of Two Artificial Intelligence‐Based Tools in Title and Abstract Screening for the Development of Evidence‐Based Cancer Guidelines – Yao – 2025 – Cancer Innovation – Wiley Online Library). We were supported in this work by our students Ashley and Sharan, and by Dr. Jonathan Sussman (Chair, Oncology).
With hopes that AI will be able to assist with reducing the workload in developing CPGs for challenging areas in breast cancer,
Ashirbani
Dr. Ashirbani Saha is the first holder of the BRIGHT Run Breast Cancer Learning Health System Chair, a permanent research position established by the BRIGHT Run in partnership with McMaster University.
