AugDiff: Diffusion based Feature Augmentation for Multiple Instance Learning in Whole Slide Image
This addresses efficiency and realism issues in feature augmentation for medical image analysis, offering a domain-specific incremental improvement over existing methods like Mixup.
The paper tackles the computational burden of image augmentation in Multiple Instance Learning (MIL) for Whole Slide Images by introducing AugDiff, a diffusion-based feature augmentation framework, which improves performance across three cancer datasets with concrete gains in metrics like AUC.
Multiple Instance Learning (MIL), a powerful strategy for weakly supervised learning, is able to perform various prediction tasks on gigapixel Whole Slide Images (WSIs). However, the tens of thousands of patches in WSIs usually incur a vast computational burden for image augmentation, limiting the MIL model's improvement in performance. Currently, the feature augmentation-based MIL framework is a promising solution, while existing methods such as Mixup often produce unrealistic features. To explore a more efficient and practical augmentation method, we introduce the Diffusion Model (DM) into MIL for the first time and propose a feature augmentation framework called AugDiff. Specifically, we employ the generation diversity of DM to improve the quality of feature augmentation and the step-by-step generation property to control the retention of semantic information. We conduct extensive experiments over three distinct cancer datasets, two different feature extractors, and three prevalent MIL algorithms to evaluate the performance of AugDiff. Ablation study and visualization further verify the effectiveness. Moreover, we highlight AugDiff's higher-quality augmented feature over image augmentation and its superiority over self-supervised learning. The generalization over external datasets indicates its broader applications.