无需训练即可在推理时选出关键组织切片，提升癌症诊断准确率
Test-Time Instance Selection for Improved Whole Slide Image Analysis
- 推理时动态筛选代表性组织切片，避免冗余信息干扰
- 在多个癌症分类任务中性能优于或等同于现有模型
- 可无缝接入现有模型，无需修改架构或重新训练
全幻灯片图像（WSI）分析广泛应用于癌症诊断。传统方法将百万像素级的WSI分割为小块，通过多实例学习（MIL）模型处理。然而，现有MIL模型通常处理所有切片，其中许多包含冗余或非信息性组织模式。尽管近期研究聚焦于实例选择以识别有判别力的切片并减少冗余，但这些选择模块仍需额外训练。本文提出测试时实例选择（TTIS），一种无需训练、即插即用的推理框架，可在推理阶段选择紧凑且具代表性的切片。TTIS进一步引入多视角集成策略，融合组织形态学的不同方面，增强鲁棒性。重要的是，TTIS可无缝集成到现有MIL模型中，无需重训练或结构改动，实现灵活部署。在多个基准上的大量评估表明，该方法在多种分类与亚型任务中均优于或匹配基线MIL性能。代码已开源：https://github.com/QuIIL/TTIS
原文摘要 · Abstract (English)
Whole Slide Image (WSI) analysis has been widely studied for cancer diagnosis. Conventionally, a gigapixel WSI is divided into small patches and processed by Multiple Instance Learning (MIL) models. However, existing MIL models typically process all patches, many of which contain redundant or non-informative tissue patterns. Although recent approaches have focused on instance selection to identify discriminative patches and reduce redundancy, these selection modules still require additional training. In this work, we propose Test-Time Instance Selection (TTIS), a training-free, plug-and-play framework that selects compact yet representative patches during inference. TTIS further incorporates a multi-view ensemble strategy to integrate distinct facets of tissue morphology, enhancing robustness. Importantly, TTIS can be seamlessly integrated into existing MIL models without retraining or architectural changes, enabling flexible deployment. Extensive evaluations across multiple benchmarks demonstrate that our approach improves or matches baseline MIL performance across a range of classification and subtyping tasks. Our implementation code is available at https://github.com/QuIIL/TTIS
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。