GRIN+：面向不平衡医疗数据的快速且有效的机器遗忘
GRIN+: Towards Fast Yet Effective Machine Unlearning for Imbalanced Medical Data
- Guangzhou University（广州大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
GRIN+提出一种面向不平衡医疗数据的机器遗忘框架，通过解耦遗忘知识与泛化表示、类别自适应影响力评分和方向约束更新，解决隐私-效率-效用三难困境，在多个医疗数据集上实现高效且准确的遗忘。
AI中文摘要：
随着深度学习模型成为现代医疗保健的基础，GDPR和HIPAA等隐私法规所规定的“被遗忘权”要求有效的机器遗忘（MU）技术，以从训练模型中移除敏感的患者数据。然而，现有的MU技术常常面临一个基本的“隐私-效率-效用”（PEU）三难困境，尤其是在医疗场景中，数据通常具有严重的类别不平衡和长尾分布特征。在这种情况下，由于多数类别的梯度主导，标准的遗忘方法可能无法保护关键的临床知识，或错误地删除对诊断罕见疾病至关重要的特征。为了解决这些挑战，我们提出了GRIN+，一种新颖的机器遗忘框架，专为不平衡医疗场景中的快速且精确的数据擦除而设计。GRIN+通过分析“遗忘”集和“保留”集的梯度贡献，在参数级别将遗忘特定知识与泛化表示解耦。它引入了一种类别自适应的影响力评分机制来纠正梯度主导，并采用方向约束的更新策略以防止关键临床知识的意外侵蚀。在多个医疗数据集（包括皮肤癌（ISIC）、脑肿瘤（MRI）和乳腺超声（BUSI））上的全面基准测试表明，GRIN+实现了PEU三难困境的最优平衡。实验结果显示，与现有基线相比，GRIN+在保持高诊断准确性和强大隐私性的同时，显著提高了运行效率。我们开源了GRIN+的代码和基准，以支持进一步的研究。
英文摘要：
As deep learning models become fundamental to modern healthcare, the "Right to be Forgotten" mandated by privacy regulations like GDPR and HIPAA necessitates effective machine unlearning (MU) to remove sensitive patient data from trained models. However, existing MU techniques often struggle with a fundamental "privacy-efficiency-utility" (PEU) trilemma, particularly in medical scenarios where data is frequently characterized by severe class imbalance and long-tailed distributions. In such cases, standard unlearning methods can fail to protect key clinical knowledge or mistakenly delete features essential for diagnosing rare conditions due to the gradient dominance of majority classes. To address these challenges, we propose GRIN+, a novel machine unlearning framework designed for fast and precise data erasure in imbalanced medical scenarios. GRIN+ decouples unlearning-specific knowledge from generalized representations at the parameter level by analyzing the gradient contributions of both "forget" and "retain" sets. It introduces a class-adaptive influence scoring mechanism to rectify gradient dominance and employs a direction-constrained update strategy to prevent the unintended erosion of vital clinical knowledge. Comprehensive benchmarking across multiple medical datasets, including skin cancer (ISIC), brain tumor (MRI), and breast ultrasound (BUSI), demonstrates that GRIN+ achieves an optimal balance of the PEU trilemma. Experimental results show that GRIN+ maintains high diagnostic accuracy and robust privacy while significantly enhancing runtime efficiency compared to existing baselines. We open-source the GRIN+ code and benchmarks to support further research.
