arXiv 2602.086522026-02-13cs.CV

基于深度学习的固定类型预测用于数字病理学的质量保证

Deep Learning-Based Fixation Type Prediction for Quality Assurance in Digital Pathology

Institute of Pathology, Technical University of Munich, Germany（慕尼黑技术大学病理学研究所）；School of Computation, Information and Technology, Technical University of Munich, Germany（慕尼黑技术大学计算、信息与技术学院）；Munich Center for Machine Learning, Germany（慕尼黑机器学习中心）；Institute of Pathology, University of Regensburg, Regensburg, Germany（罗森堡大学病理学研究所）；Institute of Pathology and Molecular Diagnostics, University Hospital of Augsburg, Germany（埃格利希大学医院病理学与分子诊断研究所）；Digital Medicine, University Hospital of Augsburg, Germany（埃格利希大学医院数字医学）；Bavarian Cancer Research Center (BZKF), Augsburg, Germany（巴伐利亚癌症研究中心（BZKF））；Bio M {}^{\text{M}} Biotech Cluster Development GmbH, Martinsried, Germany（Bio M {}^{\text{M}} 生物技术集群发展 GmbH）；Institute of Pathology, Medical Faculty Mannheim, University Heidelberg, Germany（海德堡大学曼海姆医学学院病理学研究所）；Munich Data Science Institute, Technical University of Munich, Germany（慕尼黑数据科学研究所）

Oskar Thaeter, Tanja Niedermair, Jan E. G. Albin, Johannes Raffler, Ralf Huss, Peter J. Schüffler

备注：: 11 pages, 6 figures, 7 tables

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本研究提出一种基于深度学习的模型，利用低分辨率预扫描图像预测病理切片的固定类型，提高了高通量病理学工作流程中的质量控制效率。

摘要：

准确标注固定类型是病理实验室制片过程中的关键步骤。然而，这种手动过程容易出错，影响下游分析和诊断准确性。现有方法用于验证福尔马林固定、石蜡包埋（FFPE）和冷冻切片（FS）固定类型通常需要全分辨率全片图像（WSI），限制了在高通量质量控制中的可扩展性。我们提出了一种深度学习模型，利用低分辨率预扫描缩略图图像来预测固定类型。该模型在图林病理学研究所（n=1200，Leica GT450DX）的WSI上进行训练，并在癌症基因组图谱数据集（TCGA，n=8800，Leica AT2）的类别平衡子集以及来自奥格斯堡（n=695 [392 FFPE, 303 FS]，Philips UFS）和雷根施塔特（n=202，3DHISTECH P1000）的类别平衡数据集上进行评估。我们的模型在TCGA上实现了0.88的AUROC，比现有预扫描方法高出4.8%。它在雷根施塔特和奥格斯堡的切片上分别实现了0.72的AUROC，突显了与扫描仪引起的领域偏移相关的挑战。此外，该模型处理每张切片仅需21毫秒，比现有高倍率、全分辨率方法快400倍，实现了快速、高通量的处理。这种方法为检测标签错误提供了一种高效解决方案，无需依赖高倍率扫描，为高通量病理学工作流程中的质量控制提供了有价值的工具。未来的工作将改进并评估该模型对其他扫描类型的一般化能力。我们的发现表明，该方法可以提高数字病理学工作流程的准确性和效率，并可能扩展到其他低分辨率切片标注。

英文摘要：

Accurate annotation of fixation type is a critical step in slide preparation for pathology laboratories. However, this manual process is prone to errors, impacting downstream analyses and diagnostic accuracy. Existing methods for verifying formalin-fixed, paraffin-embedded (FFPE), and frozen section (FS) fixation types typically require full-resolution whole-slide images (WSIs), limiting scalability for high-throughput quality control. We propose a deep-learning model to predict fixation types using low-resolution, pre-scan thumbnail images. The model was trained on WSIs from the TUM Institute of Pathology (n=1,200, Leica GT450DX) and evaluated on a class-balanced subset of The Cancer Genome Atlas dataset (TCGA, n=8,800, Leica AT2), as well as on class-balanced datasets from Augsburg (n=695 [392 FFPE, 303 FS], Philips UFS) and Regensburg (n=202, 3DHISTECH P1000). Our model achieves an AUROC of 0.88 on TCGA, outperforming comparable pre-scan methods by 4.8%. It also achieves AUROCs of 0.72 on Regensburg and Augsburg slides, underscoring challenges related to scanner-induced domain shifts. Furthermore, the model processes each slide in 21 ms, $400\times$ faster than existing high-magnification, full-resolution methods, enabling rapid, high-throughput processing. This approach provides an efficient solution for detecting labelling errors without relying on high-magnification scans, offering a valuable tool for quality control in high-throughput pathology workflows. Future work will improve and evaluate the model's generalisation to additional scanner types. Our findings suggest that this method can increase accuracy and efficiency in digital pathology workflows and may be extended to other low-resolution slide annotations.

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。