Forget-MI: Machine Unlearning for Forgetting Multimodal Information in Healthcare Settings

2025-06-29

AI Summary

To address the security-aware unlearning requirement for patient-sensitive data in medical multimodal AI models, this paper proposes Forget-MI—the first machine unlearning framework tailored for medical multimodal models. It enables selective unlearning of both unimodal features and cross-modal joint representations without full retraining, effectively nullifying the influence of target patient data. Forget-MI introduces an unlearning-aware loss function and a gradient perturbation mechanism, jointly optimizing privacy preservation and model utility, with membership inference attack (MIA) success rate as the primary privacy metric. Experiments demonstrate that on the forgotten dataset, AUC and F1 drop by 0.221 and 0.305, respectively, while MIA success rate decreases by 0.202. Crucially, test-set performance matches that of a fully retrained model and significantly outperforms existing unlearning methods.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Security and Privacy: Security and privacy of machine learning and AI applications
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Privacy preservation in AI is crucial, especially in healthcare, where models rely on sensitive patient data. In the emerging field of machine unlearning, existing methodologies struggle to remove patient data from trained multimodal architectures, which are widely used in healthcare. We propose Forget-MI, a novel machine unlearning method for multimodal medical data, by establishing loss functions and perturbation techniques. Our approach unlearns unimodal and joint representations of the data requested to be forgotten while preserving knowledge from the remaining data and maintaining comparable performance to the original model. We evaluate our results using performance on the forget dataset, performance on the test dataset, and Membership Inference Attack (MIA), which measures the attacker's ability to distinguish the forget dataset from the training dataset. Our model outperforms the existing approaches that aim to reduce MIA and the performance on the forget dataset while keeping an equivalent performance on the test set. Specifically, our approach reduces MIA by 0.202 and decreases AUC and F1 scores on the forget set by 0.221 and 0.305, respectively. Additionally, our performance on the test set matches that of the retrained model, while allowing forgetting. Code is available at https://github.com/BioMedIA-MBZUAI/Forget-MI.git

Problem

• Removing sensitive patient data from trained multimodal healthcare models
• Preserving model performance while forgetting requested data
• Reducing Membership Inference Attack risks in healthcare AI

Innovation

• Machine unlearning for multimodal healthcare data
• Loss functions and perturbation techniques
• Preserves knowledge while forgetting sensitive data