用语义槽聚合压缩病理图像令牌，提升大模型处理效率与诊断准确率。
TC-SSA: Token Compression via Semantic Slot Aggregation for Gigapixel Pathology Reasoning
- 通过语义槽聚合将切片令牌压缩至原序列的1.7%
- 在TCGA数据集上达78.34%准确率，优于采样基线
- 适用于高分辨率病理分析，适合医学视觉大模型研究者
将大视觉语言模型应用于计算病理学有望构建诊断辅助系统，但面临关键计算瓶颈：全切片图像（WSI）规模达百万级像素。单个WSI通常包含超过10⁵个图像块，导致序列长度超出标准Transformer架构限制。现有方法多依赖空间采样，可能丢失关键诊断信息。为此，本文提出TC-SSA（Token Compression via Semantic Slot Aggregation），一种可学习的令牌压缩框架，将图像块特征聚合为固定数量的语义槽。通过门控路由模块采用稀疏Top-2路由分配图像块至槽，并进行加权聚合，在严格令牌预算下实现全局切片覆盖。最终表示保留关键诊断信息，同时将视觉令牌数减少至原始序列的1.7%。在SlideBench(TCGA)上，模型整体准确率达78.34%，诊断子集达77.14%，优于同等令牌预算下的采样基线。该方法还推广至MIL分类任务，在TCGA-BRCA上取得95.83% AUC，TCGA-NSCLC达98.27%，PANDA为79.80%。结果表明，可学习的语义聚合在效率与诊断性能间提供有效平衡。
原文摘要 · Abstract (English)
The application of large vision-language models to computational pathology holds great promise for diagnostic assistants but faces a critical computational bottleneck: the gigapixel scale of Whole Slide Images (WSIs). A single WSI typically contains over 105 patches, creating sequence lengths that exceed the constraints of standard Transformer architectures. Existing solutions often resort to spatial sampling, which risks discarding diagnostically critical evidence. To address this, we propose TC-SSA (Token Compression via Semantic Slot Aggregation), a learnable token compression framework that aggregates patch features into a fixed number of semantic slots. A gated routing module assigns patches to slots using sparse Top-2 routing, followed by weighted aggregation, enabling global slide coverage under a strict token budget. The resulting representation retains diagnostically relevant information while reducing the number of visual tokens to 1.7% of the original sequence. On the SlideBench(TCGA), our model achieves 78.34% overall accuracy and 77.14% on the diagnosis subset, outperforming sampling-based baselines under comparable token budgets. The method also generalizes to MIL classification, reaching AUC of 95.83% on TCGA-BRCA, 98.27% on TCGA-NSCLC and 79.80% on PANDA. These results suggest that learnable semantic aggregation provides an effective trade-off between efficiency and diagnostic performance for gigapixel pathology reasoning.
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。