Machine Unlearning for Medical Imaging

TL;DR

Machine unlearning in medical imaging addresses the right to be forgotten by removing the influence of specific training samples from pretrained models. The authors compare exact unlearning with approximate methods, specifically random relabeling and saliency unlearning, on TissueMNIST and CheXpert using ResNet backbones. Results show approximate methods can match retention and forgetting performance but degrade generalization on unseen test data, with larger forget sets exacerbating the problem; they also reveal potential easy/hard sample biases and require extra hyperparameter tuning. The findings indicate that while machine unlearning is promising for medical imaging, current approximate algorithms need substantial improvements in generalization, fairness, and efficiency to be practically viable.

Abstract

Machine unlearning is the process of removing the impact of a particular set of training samples from a pretrained model. It aims to fulfill the "right to be forgotten", which grants the individuals such as patients the right to reconsider their contribution in models including medical imaging models. In this study, we evaluate the effectiveness (performance) and computational efficiency of different unlearning algorithms in medical imaging domain. Our evaluations demonstrate that the considered unlearning algorithms perform well on the retain set (samples whose influence on the model is allowed to be retained) and forget set (samples whose contribution to the model should be eliminated), and show no bias against male or female samples. They, however, adversely impact the generalization of the model, especially for larger forget set sizes. Moreover, they might be biased against easy or hard samples, and need additional computational overhead for hyper-parameter tuning. In conclusion, machine unlearning seems promising for medical imaging, but the existing unlearning algorithms still needs further improvements to become more practical for medical applications.