Case Study: Responsible AI in Medical Imaging with PACS-AI

This panel will discuss the groundbreaking PACS-AI project, which aims to deploy AI models across Canadian health care. The PACS-AI solution, developed over years of multidisciplinary research, leverages the pre-existing architecture of health care centers to allow researchers and physicians to gain insights from AI models in real time. This panel will present PACS-AI and discuss the responsible AI approach which was followed throughout the development of this solution.

Practitioners will be able to draw some useful lessons from this case study on the key steps of crafting a responsible AI strategy in a heavily regulated environment. Innovative measures implemented by the research team behind PACS-AI include exploratory interviews with patients to find out their expectations, a personalized training program for ‘champion physicians’ and the release of an open-source model to facilitate the dissemination of AI in health care and foster collaboration between the scientific community and developers.

What you will learn:

• Understanding the specific challenges when deploying AI solutions in public health care systems.
• How to develop an AI governance strategy that involves multiple internal and external stakeholders (e.g. researchers, physicians, engineers, academics, etc.).
• AI is ready to transform medical imaging for the benefit of patients and health professionals.