arXiv 2503.230212025-03-29cs.CV

组织检测对数字病理学中诊断人工智能算法的影响

The impact of tissue detection on diagnostic artificial intelligence algorithms in digital pathology

Karolinska Institutet（卡罗林斯卡学院）；Stavanger University Hospital（斯塔万格大学医院）；University of Stavanger（斯塔万格大学）；University of Bergen（卑尔根大学）；Synlab；Medical University of Lodz（罗兹医科大学）；University of Helsinki（赫尔辛基大学）；Mehiläinen Oy（梅希莱宁公司）；Mehiläinen Länsi-Pohja Hospital（梅希莱宁兰西波赫亚医院）；Aarhus University Hospital（奥胡斯大学医院）；Aarhus University（奥胡斯大学）；Griffith University（格里菲斯大学）；SciLifeLab（瑞典生命科学实验室）

Sol Erika Boman, Nita Mulliqi, Anders Blilie, Xiaoyi Ji, Kelvin Szolnoky, Einar Gudlaugsson, Emiel A. M. Janssen, Svein R. Kjosavik, José Asenjo, Marcello Gamba… 展开作者

Sol Erika Boman, Nita Mulliqi, Anders Blilie, Xiaoyi Ji, Kelvin Szolnoky, Einar Gudlaugsson, Emiel A. M. Janssen, Svein R. Kjosavik, José Asenjo, Marcello Gambacorta, Paolo Libretti, Marcin Braun, Radzislaw Kordek, Roman Łowicki, Kristina Hotakainen, Päivi Väre, Bodil Ginnerup Pedersen, Karina Dalsgaard Sørensen, Benedicte Parm Ulhøi, Lars Egevad, Kimmo Kartasalo

备注：: 25 pages, 2 tables, 3 figures, 1 supplementary figure

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本研究比较了经典与AI组织检测方法对前列腺癌Gleason分级下游任务的影响，发现AI检测能显著减少完全漏检样本，凸显了稳健组织检测对诊断AI临床性能的重要性。

摘要：

组织检测是大多数数字病理学应用中至关重要的第一步。分割算法的细节很少被报道，并且缺乏研究调查糟糕的分割算法带来的下游影响。如果在对临床应用中的标本进行诊断分析时排除了具有诊断价值的部分，忽视组织检测质量可能会成为下游性能的瓶颈，并危及患者安全。本研究旨在确定下游任务的性能是否对组织检测方法敏感，并比较经典方法和基于AI的组织检测的性能。为此，我们使用两种不同的组织检测算法：阈值法（经典）和UNet++（AI），训练了一个用于全切片图像（WSIs）中前列腺癌Gleason分级的AI模型。共有在5台数字病理学扫描仪上扫描的33,823张WSIs被用于训练组织检测AI模型。下游的Gleason分级算法使用来自13个临床站点、在13台不同扫描仪上扫描的70,524张WSIs进行训练和测试。当从基于阈值的组织检测切换到基于AI的组织检测时，完全未被检测到的组织样本从116个（0.43%）减少到22个（0.08%），这表明对于避免在具有异常外观的切片上出现完全失败，AI模型可能比经典模型更可靠。在两种算法都能检测到组织的切片上，未观察到整体Gleason分级性能有显著差异。然而，在3.5%的恶性切片中观察到依赖于组织检测的AI分级在临床上的显著变异，这凸显了稳健的组织检测对于诊断AI实现最佳临床性能的重要性。

英文摘要：

Tissue detection is a crucial first step in most digital pathology applications. Details of the segmentation algorithm are rarely reported, and there is a lack of studies investigating the downstream effects of a poor segmentation algorithm. Disregarding tissue detection quality could create a bottleneck for downstream performance and jeopardize patient safety if diagnostically relevant parts of the specimen are excluded from analysis in clinical applications. This study aims to determine whether performance of downstream tasks is sensitive to the tissue detection method, and to compare performance of classical and AI-based tissue detection. To this end, we trained an AI model for Gleason grading of prostate cancer in whole slide images (WSIs) using two different tissue detection algorithms: thresholding (classical) and UNet++ (AI). A total of 33,823 WSIs scanned on five digital pathology scanners were used to train the tissue detection AI model. The downstream Gleason grading algorithm was trained and tested using 70,524 WSIs from 13 clinical sites scanned on 13 different scanners. There was a decrease from 116 (0.43%) to 22 (0.08%) fully undetected tissue samples when switching from thresholding-based tissue detection to AI-based, suggesting an AI model may be more reliable than a classical model for avoiding total failures on slides with unusual appearance. On the slides where tissue could be detected by both algorithms, no significant difference in overall Gleason grading performance was observed. However, tissue detection dependent clinically significant variations in AI grading were observed in 3.5% of malignant slides, highlighting the importance of robust tissue detection for optimal clinical performance of diagnostic AI.

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。