基于多示例学习的整张切片图像分类的伪包混合增强
Pseudo-Bag Mixup Augmentation for Multiple Instance Learning-Based Whole Slide Image Classification
AI总结：
本文提出一种用于基于多示例学习的整张切片图像分类的伪包混合数据增强方案，通过实现尺寸与语义对齐提升模型训练效率，显著增强了分类、泛化性能及对遮挡和标签噪声的鲁棒性。
摘要：
鉴于对千兆像素图像进行建模的特殊情况，多示例学习（MIL）已成为整张切片图像（WSI）分类的最重要框架之一。在当前实践中，大多数 MIL 网络在训练时经常面临两个不可避免的问题：i) WSI 数据不足，以及 ii) 神经网络固有的样本记忆倾向。这些问题可能阻碍 MIL 模型进行充分且高效的训练，抑制分类模型在 WSI 上性能的持续提升。受 Mixup 基本思想的启发，本文提出了一种新的伪包混合（PseMix）数据增强方案来改进 MIL 模型的训练。该方案通过伪包将适用于一般图像的 Mixup 策略推广到特殊的 WSI，以便应用于基于 MIL 的 WSI 分类。在伪包的配合下，我们的 PseMix 实现了 Mixup 策略中关键的尺寸对齐和语义对齐。此外，它被设计为一种高效且解耦的方法，既不涉及耗时的操作，也不依赖 MIL 模型的预测。我们专门设计了对比实验和消融研究来评估 PseMix 的有效性和优势。实验结果表明，PseMix 通常能协助最先进的 MIL 网络刷新其在 WSI 上的分类性能。此外，它还能在特殊测试场景中提升 MIL 模型的泛化性能，并增强其对图像块遮挡和标签噪声的鲁棒性。源代码可在 https://github.com/liupei101/PseMix 获取。
英文摘要：
Given the special situation of modeling gigapixel images, multiple instance learning (MIL) has become one of the most important frameworks for Whole Slide Image (WSI) classification. In current practice, most MIL networks often face two unavoidable problems in training: i) insufficient WSI data and ii) the sample memorization inclination inherent in neural networks. These problems may hinder MIL models from adequate and efficient training, suppressing the continuous performance promotion of classification models on WSIs. Inspired by the basic idea of Mixup, this paper proposes a new Pseudo-bag Mixup (PseMix) data augmentation scheme to improve the training of MIL models. This scheme generalizes the Mixup strategy for general images to special WSIs via pseudo-bags so as to be applied in MIL-based WSI classification. Cooperated by pseudo-bags, our PseMix fulfills the critical size alignment and semantic alignment in Mixup strategy. Moreover, it is designed as an efficient and decoupled method, neither involving time-consuming operations nor relying on MIL model predictions. Comparative experiments and ablation studies are specially designed to evaluate the effectiveness and advantages of our PseMix. Experimental results show that PseMix could often assist state-of-the-art MIL networks to refresh their classification performance on WSIs. Besides, it could also boost the generalization performance of MIL models in special test scenarios, and promote their robustness to patch occlusion and label noise. Our source code is available at https://github.com/liupei101/PseMix.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
