UHN devises AI chatbot for breast cancer patients

TORONTO — People living with metastatic breast cancer (MBC) — an advanced form of breast cancer where cancer has spread to other parts of the body — often face ongoing treatments and uncertainty. This raises complex questions about treatment decisions, symptom management and future planning.

To help address these challenges, researchers at The Institute for Education Research (TIER) at UHN have developed an artificial intelligence (AI) chatbot that offers personalized, on-demand support and resources for people living with MBC.

The Artificial Intelligence Patient Librarian (AIPL) provides conversational guidance and tailored recommendations, suggesting helpful resources based on individual user questions.

To evaluate how well the chatbot works, researchers conducted a three-phase study. Dr. Yvonne Leung (pictured), lead author of the study, is an educational investigator at TIER, a research scientist at UHN’s de Souza Institute, an adjunct lecturer in the Department of Psychiatry at the University of Toronto, and a teaching assistant professor at Northeastern University, Toronto.

In the three phases, researchers did the following:

• They first collaborated with MBC patients to adapt cancer education materials for use with the chatbot;
• Next, they trained the AI to scan and recommend information from more than 100 reliable online resources chosen by expert patient librarians;
• They then tested the AIPL with 36 MBC patients through surveys and focus groups.

The participants found the chatbot easy to use and helpful for finding quick answers, especially for those who may be recently diagnosed.

However, many participants were already well informed and actively involved in their care. These users sought more in-depth medical information, emotional support and guidance for managing relationships and planning for the future.

Some also expressed interest in using the chatbot to connect with peers or explore alternative treatments and clinical trials.

This study shows how AI can help close gaps in cancer care by offering reliable and timely support for those navigating complex conditions.

Future versions of the chatbot will aim to provide even more personalized help, including a digital companion that shares lived experiences. This could lead to more emotionally supportive care for people with advanced cancer.

As this technology evolves, tools such as the AIPL could transform how patients interact with their care — promoting connection, empowerment and autonomy throughout their care journey.

The development and evaluation of the AIPL was conducted in collaboration with the M. Lau Breast Centre at UHN’s Princess Margaret Cancer Centre, leveraging their expertise in patient education and support.

This work was supported by UHN Foundation and Pfizer Global Medical Grants.