Tal Arbel

Biographie

Tal Arbel est professeure titulaire au Département de génie électrique et informatique de l’Université McGill, où elle dirige le groupe de vision probabiliste et le laboratoire d'imagerie médicale du Centre sur les machines intelligentes.

Elle est titulaire d'une chaire en IA Canada-CIFAR et membre associée de Mila – Institut québécois d’intelligence artificielle ainsi que du Centre de recherche sur le cancer Goodman. Les recherches de la professeure Arbel portent sur le développement de méthodes probabilistes d'apprentissage profond dans les domaines de la vision par ordinateur et de l’analyse d'imagerie médicale pour un large éventail d'applications dans le monde réel, avec un accent particulier sur les maladies neurologiques.

Elle a remporté le prix de la recherche Christophe Pierre 2019 de McGill Engineering et est Fellow à l'Académie canadienne d'ingénierie. Elle fait régulièrement partie de l'équipe organisatrice de grandes conférences internationales sur la vision par ordinateur et l'analyse d'imagerie médicale (par exemple celles de la Medical Image Computing and Computer-Assisted Intervention Society/MICCAI et de Medical Imaging with Deep Learning/MIDL, l’International Conference on Computer Vision/ICCV ou encore la Conference on Computer Vision and Pattern Recognition/CVPR). Elle est rédactrice en chef et cofondatrice de la revue Machine Learning for Biomedical Imaging (MELBA).

Étudiants actuels

Cornelius Crijnen

Collaborateur·rice de recherche - UdeM

Nima Fathi

Maîtrise recherche - McGill

Gian Favero

Maîtrise recherche - McGill

Elizabeth Laura Janes

Maîtrise recherche - McGill

Emily Kaczmarek

Doctorat - McGill

Doctorat - McGill

Mohamed Mohamed

Maîtrise recherche - McGill

Toky Raharison Ralambomihanta

Baccalauréat - McGill

Ryan Rezai

Maîtrise recherche - McGill

Parham Saremi

Maîtrise recherche - McGill

PRISM : Un modèle d'IA générative explicable pour l'imagerie médicale

Xing Shen

Maîtrise recherche - McGill

Maîtrise recherche - McGill

Billets de blogue

Image of an Xray and the DDIM process to generate counterfactual version of Xrays

1 juillet 2025

par

Amar Kumar

Anita Kriz

Mohammed Havaei

Tal Arbel

Lire l'article

Publications

Pixel Perfect MegaMed: A Megapixel-Scale Vision-Language Foundation Model for Generating High Resolution Medical Images

Zahra TehraniNasab

Medical image synthesis presents unique challenges due to the inherent complexity and high-resolution details required in clinical contexts.… (voir plus) Traditional generative architectures such as Generative Adversarial Networks (GANs) or Variational Auto Encoder (VAEs) have shown great promise for high-resolution image generation but struggle with preserving fine-grained details that are key for accurate diagnosis. To address this issue, we introduce Pixel Perfect MegaMed, the first vision-language foundation model to synthesize images at resolutions of 1024x1024. Our method deploys a multi-scale transformer architecture designed specifically for ultra-high resolution medical image generation, enabling the preservation of both global anatomical context and local image-level details. By leveraging vision-language alignment techniques tailored to medical terminology and imaging modalities, Pixel Perfect MegaMed bridges the gap between textual descriptions and visual representations at unprecedented resolution levels. We apply our model to the CheXpert dataset and demonstrate its ability to generate clinically faithful chest X-rays from text prompts. Beyond visual quality, these high-resolution synthetic images prove valuable for downstream tasks such as classification, showing measurable performance gains when used for data augmentation, particularly in low-data regimes. Our code is accessible through the project website - https://tehraninasab.github.io/pixelperfect-megamed.

2025-07-17

ArXiv (prépublication)

Prompt4Trust: A Reinforcement Learning Prompt Augmentation Framework for Clinically-Aligned Confidence Calibration in Multimodal Large Language Models

Elizabeth Laura Janes

Xing Shen

Multimodal large language models (MLLMs) hold considerable promise for applications in healthcare. However, their deployment in safety-criti… (voir plus)cal settings is hindered by two key limitations: (i) sensitivity to prompt design, and (ii) a tendency to generate incorrect responses with high confidence. As clinicians may rely on a model's stated confidence to gauge the reliability of its predictions, it is especially important that when a model expresses high confidence, it is also highly accurate. We introduce Prompt4Trust, the first reinforcement learning (RL) framework for prompt augmentation targeting confidence calibration in MLLMs. A lightweight LLM is trained to produce context-aware auxiliary prompts that guide a downstream task MLLM to generate responses in which the expressed confidence more accurately reflects predictive accuracy. Unlike conventional calibration techniques, Prompt4Trust specifically prioritizes aspects of calibration most critical for safe and trustworthy clinical decision-making. Beyond improvements driven by this clinically motivated calibration objective, our proposed method also improves task accuracy, achieving state-of-the-art medical visual question answering (VQA) performance on the PMC-VQA benchmark, which is composed of multiple-choice questions spanning diverse medical imaging modalities. Moreover, our framework trained with a small downstream task MLLM showed promising zero-shot generalization to larger MLLMs in our experiments, suggesting the potential for scalable calibration without the associated computational costs. This work demonstrates the potential of automated yet human-aligned prompt engineering for improving the the trustworthiness of MLLMs in safety critical settings. Our codebase can be found at https://github.com/xingbpshen/prompt4trust.

2025-07-12

ArXiv (prépublication)

AURA: A Multi-Modal Medical Agent for Understanding, Reasoning & Annotation

Nima Fathi

Recent advancements in Large Language Models (LLMs) have catalyzed a paradigm shift from static prediction systems to agentic AI agents capa… (voir plus)ble of reasoning, interacting with tools, and adapting to complex tasks. While LLM-based agentic systems have shown promise across many domains, their application to medical imaging remains in its infancy. In this work, we introduce AURA, the first visual linguistic explainability agent designed specifically for comprehensive analysis, explanation, and evaluation of medical images. By enabling dynamic interactions, contextual explanations, and hypothesis testing, AURA represents a significant advancement toward more transparent, adaptable, and clinically aligned AI systems. We highlight the promise of agentic AI in transforming medical image analysis from static predictions to interactive decision support. Leveraging Qwen-32B, an LLM-based architecture, AURA integrates a modular toolbox comprising: (i) a segmentation suite with phase grounding, pathology segmentation, and anatomy segmentation to localize clinically meaningful regions; (ii) a counterfactual image-generation module that supports reasoning through image-level explanations; and (iii) a set of evaluation tools including pixel-wise difference-map analysis, classification, and advanced state-of-the-art components to assess diagnostic relevance and visual interpretability.

2025-07-01

arXiv (publié)

Pixel Perfect MegaMed: A Megapixel-Scale Vision-Language Foundation Model for Generating High Resolution Medical Images

Zahra Tehrani Nasab

Hujun Ni

2025-07-01

arXiv (publié)

Language-Guided Trajectory Traversal in Disentangled Stable Diffusion Latent Space for Factorized Medical Image Generation

Zahra Tehrani Nasab

2025-06-11

2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) (publié)

Exposing and Mitigating Calibration Biases and Demographic Unfairness in MLLM Few-Shot In-Context Learning for Medical Image Classification

Xing Shen

Justin Szeto

Mingyang Li

Hengguan Huang

2025-06-01

arXiv (publié)

Language-Guided Trajectory Traversal in Disentangled Stable Diffusion Latent Space for Factorized Medical Image Generation

Zahra TehraniNasab

Text-to-image diffusion models have demonstrated a remarkable ability to generate photorealistic images from natural language prompts. These… (voir plus) high-resolution, language-guided synthesized images are essential for the explainability of disease or exploring causal relationships. However, their potential for disentangling and controlling latent factors of variation in specialized domains like medical imaging remains under-explored. In this work, we present the first investigation of the power of pre-trained vision-language foundation models, once fine-tuned on medical image datasets, to perform latent disentanglement for factorized medical image generation and interpolation. Through extensive experiments on chest X-ray and skin datasets, we illustrate that fine-tuned, language-guided Stable Diffusion inherently learns to factorize key attributes for image generation, such as the patient's anatomical structures or disease diagnostic features. We devise a framework to identify, isolate, and manipulate key attributes through latent space trajectory traversal of generative models, facilitating precise control over medical image synthesis.

2025-03-30

ArXiv (prépublication)

Language-Guided Trajectory Traversal in Disentangled Stable Diffusion Latent Space for Factorized Medical Image Generation

Zahra TehraniNasab

2025-03-30

ArXiv (prépublication)

Leveraging Vision-Language Foundation Models to Reveal Hidden Image-Attribute Relationships in Medical Imaging

B. Pertzov

2025-03-30

ArXiv (prépublication)

Leveraging Vision-Language Foundation Models to Reveal Hidden Image-Attribute Relationships in Medical Imaging

B. Pertzov

Vision-language foundation models (VLMs) have shown impressive performance in guiding image generation through text, with emerging applicati… (voir plus)ons in medical imaging. In this work, we are the first to investigate the question: 'Can fine-tuned foundation models help identify critical, and possibly unknown, data properties?' By evaluating our proposed method on a chest x-ray dataset, we show that these models can generate high-resolution, precisely edited images compared to methods that rely on Structural Causal Models (SCMs) according to numerous metrics. For the first time, we demonstrate that fine-tuned VLMs can reveal hidden data relationships that were previously obscured due to available metadata granularity and model capacity limitations. Our experiments demonstrate both the potential of these models to reveal underlying dataset properties while also exposing the limitations of fine-tuned VLMs for accurate image editing and susceptibility to biases and spurious correlations.

2025-03-30

ArXiv (prépublication)

Conditional Diffusion Models are Medical Image Classifiers that Provide Explainability and Uncertainty for Free

Gian Mario Favero

2025-03-27

MIDL.io/2025/Conference (présentation orale)

openreview.net

PRISM: High-Resolution&Precise Counterfactual Medical Image Generation using Language-guided Stable Diffusion

Mohammad Havaei

Developing reliable and generalizable deep learning systems for medical imaging faces significant obstacles due to spurious correlations, da… (voir plus)ta imbalances, and limited text annotations in datasets. Addressing these challenges requires architectures robust to the unique complexities posed by medical imaging data. The rapid advancements in vision-language foundation models within the natural image domain prompt the question of how they can be adapted for medical imaging tasks. In this work, we present PRISM, a framework that leverages foundation models to generate high-resolution, language-guided medical image counterfactuals using Stable Diffusion. Our approach demonstrates unprecedented precision in selectively modifying spurious correlations (the medical devices) and disease features, enabling the removal and addition of specific attributes while preserving other image characteristics. Through extensive evaluation, we show how PRISM advances counterfactual generation and enables the development of more robust downstream classifiers for clinically deployable solutions. To facilitate broader adoption and research, we make our code publicly available at https://github.com/Amarkr1/PRISM.

2025-03-27

MIDL.io/2025/Conference (présentation orale)