IBIIS-AIMI Seminar: Towards Equitable Image-Based Personalized Medicine: Causality, Confidence, and Bias Mitigation - Tal Arbel, PhD

Abstract: In current clinical practice, treatment decisions often rely on broad demographic factors and standardized markers that miss individual patient nuances. In this talk, I explore how uncertainty-aware causal deep learning—informed by medical images and demographics—can tailor treatments to a patient’s unique profile, improving the accuracy, fairness, and safety of personalized medicine. This framework is grounded in large-scale MRI data from randomized controlled trials for neurological disease treatments.

To ensure these personalized models are truly equitable and trustworthy, we must rigorously expose and mitigate hidden biases. I will highlight how fine-tuned Vision-Language Foundation Models (e.g. based on Stable Diffusion) offer interpretable, patient-specific explanations by generating precise medical image counterfactuals—powerful tools for uncovering and mitigating biases driven by spurious correlations. Building on this, I will present recent strategies for correcting calibration biases across population subgroups in Multimodal Large Language Models (MLLMs). Finally, I will offer a glimpse into the enormous potential of Agentic AI in advancing clinical care. Transitioning toward dynamic clinical agents holds immense promise for unlocking transparent reasoning and advanced explainability, ultimately  paving the way for a truly interactive and trustworthy era of clinical decision support.

About: Tal Arbel is a Professor in the Department of Electrical and Computer Engineering and an Associate Member of the School of Computer Science at McGill University, where she directs the Probabilistic Vision Group and Medical Imaging Lab within the Centre for Intelligent Machines. She is a Canada CIFAR AI Chair, a Core Member of Mila - Quebec Artificial Intelligence Institute, a Fellow of the Canadian Academy of Engineering, an Associate Member of the Goodman Cancer Institute, and a Co-Advisor for the ELLIS PhD Program.

Prof. Arbel’s research focuses on the development of probabilistic deep learning methods in computer vision and medical image analysis. Her current work is focused on causal inference, generative models, Vision-Language Models, and Agentic AI with the goal of supporting the safe, equitable, and trustworthy clinical deployment of personalized medicine. She is a recipient of the 2025 McGill Bravo Award and the 2019 ChristophePierre Award for Research Excellence. Her lab’s recent research has been recognized with several Best Paper and Best Poster awards at venues such as MICCAI and MIDL. She regularly serves on the organizing committees of major international conferences (e.g., MICCAI, MIDL, ICCV, CVPR). She is a co-founder of the online journal Machine Learning for Biomedical Imaging (MELBA), where she served as Editor-in-Chief for five years and is currently an Executive Editor.