2024 Program with Abstracts

Workshop: Bias in AI - Lombardy/Umbria

le 8 juin 2024 from 9h30 CST to 10h30 CST
Bias in AI
 
Moderators:
 
Dr. Geetha Menon
Dr. Young Lee-Bartlet
 
Speakers:
 
Dr. Parminder Basran, PhD, FCCPM (Cornell University, College of Veterinary Medicine)
Dr. David DeVries, MSc, PhD (London Health Sciences Centre, Department of Radiation Oncology)
Kayla O'Sullivan-Steben, MSc (McGill University, Medical Physics Unit)
Carter Kolbeck, MASc (Radformation - formely Limbus AI)
 
Abstract:
Artificial Intelligence (AI) has revolutionized the field of medical physics and the implementation of AI-based tools is expected to grow substantially over the next decade. These tools are slated to enable advanced diagnostics, treatment planning, and patient care, with medical physicists expected to play an important role in ensuring their safe and effective implementation. However, the deployment of AI systems in healthcare introduces inherent challenges related to bias. There are many types of bias unknown to those developing and implementing AI systems, which can result in unintended consequences to the patient and introduce ethical concerns. It is therefore important to be able to identify these sources of bias, understand their impact, and mitigate against them when implementing AI systems into clinical practice.
 
This session will explore the sources, impacts, and strategies to detect and mitigate bias in AI systems within the context of medical physics. The current and future uses of AI in medical physics will be discussed while framing its implementation through an inclusive, diverse, equitable, accessible (IDEA), and just lens. Additionally, current biases in healthcare and its propagation through AI models will be discussed. This session will provide attendees with the knowledge of bias in AI systems and the ability to better understand how AI systems are affected by bias. Lastly, attendees will be provided with a call to action to deepen their understanding, foster collaboration, and empower themselves to address bias effectively in clinical AI applications.
 
Learning Objectives
Attendees of the workshop will be able to:
  1. Explain the uses of artificial intelligence in medical physics and the challenges of implementing AI systems into healthcare (with a focus on bias)
  2. Identify sources of bias in clinical AI systems
  3. Evaluate the possible impacts of bias in clinical AI systems
  4. Apply mitigation strategies to combat bias
  5. Appreciate the importance of interdisciplinary collaboration involving clinicians, data scientists, and ethicists to mitigate bias