arXiv 2602.140102026-08-05cs.CVcs.AI

一种适用于部署的高效计算病理学基础框架

A Deployment-Friendly Foundational Framework for Efficient Computational Pathology

Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology（电子与计算机工程系，香港科学与技术大学）；Department of Computer Science and Engineering, The Hong Kong University of Science and Technology（计算机科学与工程系，香港科学与技术大学）；Department of Pathology, Nanfang Hospital, School of Basic Medical Sciences, Southern Medical University（病理学系，南芳医院，南方医科大学）；Guangdong Province Key Laboratory of Molecular Tumor Pathology（广东省分子肿瘤病理学重点实验室）；Jinfeng Laboratory, Chongqing, China（锦峰实验室，重庆，中国）；Department of General Surgery (Colorectal Surgery), The Sixth Affiliated Hospital, Sun Yat-sen University（普外科（结直肠外科），中山大学第六附属医院）

Yu Cai, Cheng Jin, Zhengyu Zhang, Jiabo Ma, Fengtao Zhou, Yingxue Xu, Zhengrui Guo, Yihui Wang, Zhengyu Zhang, Ling Liang, Yonghao Tan, Pingcheng Dong, Du Cai, … 展开作者

Yu Cai, Cheng Jin, Zhengyu Zhang, Jiabo Ma, Fengtao Zhou, Yingxue Xu, Zhengrui Guo, Yihui Wang, Zhengyu Zhang, Ling Liang, Yonghao Tan, Pingcheng Dong, Du Cai, On Ki Tang, Chenglong Zhao, Zhijian Cen, Ying Tan, Xi Wang, Can Yang, Yali Xu, Jing Cui, Zhenhui Li, Ronald Cheong Kin Chan, Yueping Liu, Feng Gao, Xiuming Zhang, Li Liang, Hao Chen, Kwang-Ting Cheng

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

LitePath是一种高效计算病理学基础框架，通过压缩模型和自适应碎片选择器，显著降低计算成本，实现快速、节能的病理图像分析。

摘要：

病理基础模型（PFMs）通过大规模数据集和扩展架构在计算病理学中实现了稳健的泛化能力，但其显著的计算成本，尤其是针对吉像素整张滑片图像，限制了临床应用和扩展性。本文提出了LitePath，一种适用于部署的基础框架，旨在减轻模型过参数化和碎片级冗余问题。LitePath整合了LiteFM，这是从三个大型PFMs（Virchow2、H-Optimus-1和UNI2）中使用19000万个碎片进行压缩得到的紧凑模型，以及自适应碎片选择器（APS），一个用于任务特定碎片选择的轻量组件。该框架将模型参数减少28倍，将FLOPs降低403.5倍，相对于Virchow2，使其能够在低功耗边缘硬件上部署，如NVIDIA Jetson Orin Nano Super。在该设备上，LitePath每小时可处理208张切片，比Virchow2快104.5倍，且处理3000张切片消耗0.36千瓦时，比在RTX3090 GPU上Virchow2低171倍。我们使用来自四个器官的37个队列、26项任务（26项内部任务、9项外部任务和2项前瞻性任务）的15672张切片（来自9808名患者，与预训练数据不重叠）验证了准确性。LitePath在19个评估模型中排名第二，优于包括H-Optimus-1、mSTAR、UNI2和GPFM在内的较大模型，同时在平均情况下保留了Virchow2 99.71%的AUC。为了量化准确性和效率之间的平衡，我们提出了部署性分数（D-Score），定义为标准化AUC和标准化FLOP的加权几何平均值，其中LitePath实现了最高值，比Virchow2高10.64%。这些结果表明，LitePath能够在可访问的硬件上实现快速、成本效益和能耗低的病理图像分析，同时保持与最先进PFMs相当的准确性，并减少了AI部署的碳足迹。

英文摘要：

Pathology foundation models (PFMs) generalize well across computational pathology tasks but remain costly for gigapixel whole-slide image analysis. Here, we present LitePath, a deployment-friendly framework that addresses model over-parameterization and patch-level redundancy. LitePath combines LiteFM, a compact model distilled from Virchow2, H-Optimus-1 and UNI2 using 190 million patches, with the Adaptive Patch Selector for task-specific patch selection. Compared with Virchow2, LitePath uses 28x fewer parameters and 403.5x fewer FLOPs. On an NVIDIA Jetson Orin Nano Super, it processes 208 slides per hour, 104.5x faster than Virchow2, and consumes 0.36 kWh per 3,000 slides, 171x less energy than Virchow2 on an RTX 3090 GPU. We evaluated LitePath on 45 multicenter cohorts across four organs and 33 tasks, comprising 37 classification and 8 survival cohorts, including 33 internal, 10 external and 2 prospective cohorts, with 17,837 slides from 9,977 patients disjoint from the pretraining data. Among 22 PFMs, LitePath achieved the best average rank (6.56 vs. 6.58 for Virchow2), retained 99.71% of Virchow2's Macro-AUC across classification cohorts, and improved mean C-index by 2.14 percentage points across survival cohorts (71.91% vs. 69.77%). We further introduce the Deployability Score (D-Score), a weighted geometric mean of normalized task performance and FLOPs. LitePath achieved the highest D-Score (0.8455), outperforming H0-mini (0.7723) and Virchow2 (0.7297). In a randomized paired crossover study of four pathologists and 120 cases, LitePath increased diagnostic accuracy by 4.1-15.8 percentage points and reduced diagnostic time by 10.9-14.2%. These results demonstrate rapid, cost-effective and energy-efficient pathology image analysis on accessible hardware while maintaining competitive performance.

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。