• 文献检索
  • 文档翻译
  • 深度研究
  • 学术资讯
  • Suppr Zotero 插件Zotero 插件
  • 邀请有礼
  • 套餐&价格
  • 历史记录
应用&插件
Suppr Zotero 插件Zotero 插件浏览器插件Mac 客户端Windows 客户端微信小程序
定价
高级版会员购买积分包购买API积分包
服务
文献检索文档翻译深度研究API 文档MCP 服务
关于我们
关于 Suppr公司介绍联系我们用户协议隐私条款
关注我们

Suppr 超能文献

核心技术专利：CN118964589B侵权必究
粤ICP备2023148730 号-1Suppr @ 2026

文献检索

告别复杂PubMed语法，用中文像聊天一样搜索，搜遍4000万医学文献。AI智能推荐，让科研检索更轻松。

立即免费搜索

文件翻译

保留排版，准确专业，支持PDF/Word/PPT等文件格式，支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述，25分钟生成高质量综述，智能提取关键信息，辅助科研写作。

立即免费体验

基于目标检测在苏木精-伊红（H&E）染色的全玻片图像中检测中心母细胞。

Detection of centroblast cells in H&E stained whole slide image based on object detection.

作者信息

Yuenyong Sumeth, Boonsakan Paisarn, Sripodok Supasan, Thuwajit Peti, Charngkaew Komgrid, Pongpaibul Ananya, Angkathunyakul Napat, Hnoohom Narit, Thuwajit Chanitra

机构信息

Department of Computer Engineering, Faculty of Engineering, Mahidol University, Nakhon Pathom, Thailand.

Department of Pathology, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand.

出版信息

Front Med (Lausanne). 2024 Feb 7;11:1303982. doi: 10.3389/fmed.2024.1303982. eCollection 2024.

DOI:10.3389/fmed.2024.1303982
PMID:38384407
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC10879397/
Abstract

INTRODUCTION

Detection and counting of Centroblast cells (CB) in hematoxylin & eosin (H&E) stained whole slide image (WSI) is an important workflow in grading Lymphoma. Each high power field (HPF) patch of a WSI is inspected for the number of CB cells and compared with the World Health Organization (WHO) guideline that organizes lymphoma into 3 grades. Spotting and counting CBs is time-consuming and labor intensive. Moreover, there is often disagreement between different readers, and even a single reader may not be able to perform consistently due to many factors.

METHOD

We propose an artificial intelligence system that can scan patches from a WSI and detect CBs automatically. The AI system works on the principle of object detection, where the CB is the single class of object of interest. We trained the AI model on 1,669 example instances of CBs that originate from WSI of 5 different patients. The data was split 80%/20% for training and validation respectively.

RESULT

The best performance was from YOLOv5x6 model that used the preprocessed CB dataset achieved precision of 0.808, recall of 0.776, mAP at 0.5 IoU of 0.800 and overall mAP of 0.647.

DISCUSSION

The results show that centroblast cells can be detected in WSI with relatively high precision and recall.

摘要

引言

在苏木精和伊红（H&E）染色的全切片图像（WSI）中检测和计数中心母细胞（CB）是淋巴瘤分级的重要工作流程。对WSI的每个高倍视野（HPF）切片检查CB细胞数量，并与将淋巴瘤分为3级的世界卫生组织（WHO）指南进行比较。发现和计数CB细胞既耗时又费力。此外，不同的阅片者之间常常存在分歧，而且由于多种因素，即使是单个阅片者也可能无法始终如一地进行操作。

方法

我们提出了一种人工智能系统，该系统可以扫描WSI中的切片并自动检测CB。该人工智能系统基于目标检测原理工作，其中CB是唯一感兴趣的目标类别。我们在来自5名不同患者的WSI的1669个CB示例实例上训练了人工智能模型。数据分别以80%/20%的比例用于训练和验证。

结果

性能最佳的是使用预处理后的CB数据集的YOLOv5x6模型，其精度为0.808，召回率为0.776，在IoU为0.5时的平均精度均值（mAP）为0.800，总体mAP为0.647。

讨论

结果表明，可以在WSI中以相对较高的精度和召回率检测中心母细胞。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/eb5b5685909d/fmed-11-1303982-g0009.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/f85f352601be/fmed-11-1303982-g0001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/432e09f27900/fmed-11-1303982-g0002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/8f985f36e0b3/fmed-11-1303982-g0003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/e03e49acd802/fmed-11-1303982-g0004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/c62081007dee/fmed-11-1303982-g0005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/ba83948b45b8/fmed-11-1303982-g0006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/5c64e0f9e99f/fmed-11-1303982-g0007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/5b9c0bff96f9/fmed-11-1303982-g0008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/eb5b5685909d/fmed-11-1303982-g0009.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/f85f352601be/fmed-11-1303982-g0001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/432e09f27900/fmed-11-1303982-g0002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/8f985f36e0b3/fmed-11-1303982-g0003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/e03e49acd802/fmed-11-1303982-g0004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/c62081007dee/fmed-11-1303982-g0005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/ba83948b45b8/fmed-11-1303982-g0006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/5c64e0f9e99f/fmed-11-1303982-g0007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/5b9c0bff96f9/fmed-11-1303982-g0008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/e499/10879397/eb5b5685909d/fmed-11-1303982-g0009.jpg

相似文献

1
Detection of centroblast cells in H&E stained whole slide image based on object detection.基于目标检测在苏木精-伊红（H&E）染色的全玻片图像中检测中心母细胞。
Front Med (Lausanne). 2024 Feb 7;11:1303982. doi: 10.3389/fmed.2024.1303982. eCollection 2024.
2
PseudoCell: Hard Negative Mining as Pseudo Labeling for Deep Learning-Based Centroblast Cell Detection.伪细胞：基于深度学习的中心母细胞检测中作为伪标签的硬负样本挖掘
IEEE Open J Eng Med Biol. 2024 May 30;5:514-523. doi: 10.1109/OJEMB.2024.3407351. eCollection 2024.
3
Inter-reader variability in follicular lymphoma grading: Conventional and digital reading.滤泡性淋巴瘤分级中的阅片者间变异性：传统阅片与数字阅片
J Pathol Inform. 2013 Oct 29;4:30. doi: 10.4103/2153-3539.120747. eCollection 2013.
4
Diagnostic validation of a portable whole slide imaging scanner for lymphoma diagnosis in resource-constrained setting: A cross-sectional study.资源受限环境下用于淋巴瘤诊断的便携式全切片成像扫描仪的诊断验证：一项横断面研究。
J Pathol Inform. 2023 Jan 9;14:100188. doi: 10.1016/j.jpi.2023.100188. eCollection 2023.
5
Multilayer outperforms single-layer slide scanning in AI-based classification of whole slide images with low-burden acid-fast mycobacteria (AFB).多层扫描优于单层扫描，可基于人工智能对低负担抗酸分枝杆菌（AFB）全玻片图像进行分类。
Comput Methods Programs Biomed. 2023 Jun;234:107518. doi: 10.1016/j.cmpb.2023.107518. Epub 2023 Mar 28.
6
Deep learning supported mitoses counting on whole slide images: A pilot study for validating breast cancer grading in the clinical workflow.深度学习支持的全切片图像有丝分裂计数：在临床工作流程中验证乳腺癌分级的初步研究。
J Pathol Inform. 2023 May 4;14:100316. doi: 10.1016/j.jpi.2023.100316. eCollection 2023.
7
Multi-institutional comparison of whole slide digital imaging and optical microscopy for interpretation of hematoxylin-eosin-stained breast tissue sections.多机构比较苏木精-伊红染色乳腺组织切片的全玻片数字成像与光学显微镜解读。
Arch Pathol Lab Med. 2013 Dec;137(12):1733-9. doi: 10.5858/arpa.2012-0437-OA. Epub 2013 Aug 15.
8
Use of Deep Learning to Develop and Analyze Computational Hematoxylin and Eosin Staining of Prostate Core Biopsy Images for Tumor Diagnosis.使用深度学习开发和分析前列腺核心活检图像的计算苏木精和伊红染色，用于肿瘤诊断。
JAMA Netw Open. 2020 May 1;3(5):e205111. doi: 10.1001/jamanetworkopen.2020.5111.
9
An artificial intelligence algorithm for prostate cancer diagnosis in whole slide images of core needle biopsies: a blinded clinical validation and deployment study.一种用于经皮穿刺活检全切片图像中前列腺癌诊断的人工智能算法：一项盲法临床验证与应用研究。
Lancet Digit Health. 2020 Aug;2(8):e407-e416. doi: 10.1016/S2589-7500(20)30159-X.
10
Piloting a Deep Learning Model for Predicting Nuclear BAP1 Immunohistochemical Expression of Uveal Melanoma from Hematoxylin-and-Eosin Sections.利用苏木精-伊红染色切片构建深度学习模型预测葡萄膜黑色素瘤中核BAP1免疫组化表达
Transl Vis Sci Technol. 2020 Sep 1;9(2):50. doi: 10.1167/tvst.9.2.50. eCollection 2020 Sep.

本文引用的文献

1
An annotation-free whole-slide training approach to pathological classification of lung cancer types using deep learning.基于深度学习的无注释全切片肺癌类型病理分类训练方法。
Nat Commun. 2021 Feb 19;12(1):1193. doi: 10.1038/s41467-021-21467-y.
2
Clinical-grade computational pathology using weakly supervised deep learning on whole slide images.基于全切片图像的弱监督深度学习的临床级计算病理学。
Nat Med. 2019 Aug;25(8):1301-1309. doi: 10.1038/s41591-019-0508-1. Epub 2019 Jul 15.
3
Current status and progress of lymphoma research in East Asian countries: Introduction and planning.
东亚国家淋巴瘤研究的现状与进展：引言与规划
Int J Hematol. 2018 Apr;107(4):392-394. doi: 10.1007/s12185-018-2425-3. Epub 2018 Feb 28.
4
Non-Hodgkin lymphoma in South East Asia: An analysis of the histopathology, clinical features, and survival from Thailand.东南亚的非霍奇金淋巴瘤：泰国的组织病理学、临床特征及生存情况分析
Hematol Oncol. 2018 Feb;36(1):28-36. doi: 10.1002/hon.2392. Epub 2017 Mar 23.
5
Patch-based Convolutional Neural Network for Whole Slide Tissue Image Classification.用于全切片组织图像分类的基于补丁的卷积神经网络
Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit. 2016 Jun-Jul;2016:2424-2433. doi: 10.1109/CVPR.2016.266.
6
The 2016 revision of the World Health Organization classification of lymphoid neoplasms.《世界卫生组织淋巴组织肿瘤分类（2016年修订版）》
Blood. 2016 May 19;127(20):2375-90. doi: 10.1182/blood-2016-01-643569. Epub 2016 Mar 15.
7
Classification of follicular lymphoma: the effect of computer aid on pathologists grading.滤泡性淋巴瘤的分类：计算机辅助对病理学家分级的影响
BMC Med Inform Decis Mak. 2015 Dec 30;15:115. doi: 10.1186/s12911-015-0235-6.
8
Region-Based Convolutional Networks for Accurate Object Detection and Segmentation.基于区域的卷积神经网络用于精确的目标检测和分割。
IEEE Trans Pattern Anal Mach Intell. 2016 Jan;38(1):142-58. doi: 10.1109/TPAMI.2015.2437384.
9
Deep learning.深度学习。
Nature. 2015 May 28;521(7553):436-44. doi: 10.1038/nature14539.
10
Histopathological image analysis for centroblasts classification through dimensionality reduction approaches.通过降维方法对中心母细胞进行组织病理学图像分析分类。
Cytometry A. 2014 Mar;85(3):242-55. doi: 10.1002/cyto.a.22432. Epub 2013 Dec 26.