arXiv 2304.002162024-02-16eess.IVcs.CVcs.LG

用于病理图像诊断的跨尺度多示例学习

Cross-scale Multi-instance Learning for Pathological Image Diagnosis

Vanderbilt University（范德堡大学）；The University of North Carolina at Chapel Hill（北卡罗来纳大学教堂山分校）；Vanderbilt University Medical Center（范德堡大学医学中心）；Veterans Affairs Tennessee Valley Healthcare System（田纳西河谷退伍军人事务医疗系统）

Ruining Deng, Can Cui, Lucas W. Remedios, Shunxing Bao, R. Michael Womick, Sophie Chiron, Jia Li, Joseph T. Roland, Ken S. Lau, Qi Liu, Keith T. Wilson, Yaohong… 展开作者

Ruining Deng, Can Cui, Lucas W. Remedios, Shunxing Bao, R. Michael Womick, Sophie Chiron, Jia Li, Joseph T. Roland, Ken S. Lau, Qi Liu, Keith T. Wilson, Yaohong Wang, Lori A. Coburn, Bennett A. Landman, Yuankai Huo

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对病理图像诊断中单一尺度多示例学习忽略跨尺度信息的问题，提出跨尺度MIL（CS-MIL）算法，通过显式聚合跨尺度关系提升诊断性能，并发布可视化数据集与开源代码。

摘要：

在数字病理学中，分析高分辨率全切片图像（WSIs）时处理跨多个尺度的信息带来了重大挑战。多示例学习（MIL）是处理高分辨率图像的常见解决方案，它通过对对象包（即较小图像补丁的集合）进行分类来工作。然而，此类处理通常在WSIs的单个尺度（例如20倍放大）下进行，忽略了人类病理学家进行诊断所必需的重要跨尺度信息。在本研究中，我们提出了一种新颖的跨尺度MIL算法，以将跨尺度关系显式聚合到单个MIL网络中，用于病理图像诊断。本文的贡献有三方面：（1）提出了一种集成多尺度信息和跨尺度关系的新颖跨尺度MIL（CS-MIL）算法；（2）创建并发布了一个具有尺度特定形态学特征的玩具数据集，以检查和可视化差异化的跨尺度注意力；（3）我们简单的跨尺度MIL策略在内部和公共数据集上均展示了卓越的性能。官方实现已公开于 https://github.com/hrlblab/CS-MIL。

英文摘要：

Analyzing high resolution whole slide images (WSIs) with regard to information across multiple scales poses a significant challenge in digital pathology. Multi-instance learning (MIL) is a common solution for working with high resolution images by classifying bags of objects (i.e. sets of smaller image patches). However, such processing is typically performed at a single scale (e.g., 20x magnification) of WSIs, disregarding the vital inter-scale information that is key to diagnoses by human pathologists. In this study, we propose a novel cross-scale MIL algorithm to explicitly aggregate inter-scale relationships into a single MIL network for pathological image diagnosis. The contribution of this paper is three-fold: (1) A novel cross-scale MIL (CS-MIL) algorithm that integrates the multi-scale information and the inter-scale relationships is proposed; (2) A toy dataset with scale-specific morphological features is created and released to examine and visualize differential cross-scale attention; (3) Superior performance on both in-house and public datasets is demonstrated by our simple cross-scale MIL strategy. The official implementation is publicly available at https://github.com/hrlblab/CS-MIL.

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。