Chinese Medical Sciences Journal ›› 2021, Vol. 36 ›› Issue (3): 204-209.doi: 10.24920/003962

所属专题: 人工智能与精准肿瘤学

• 论著 • 上一篇    下一篇

基于深度学习算法的胃炎组织病理学诊断系统

巴伟1,王书浩2,刘灿城2,王跃峰2,石怀银1,宋志刚1,*()   

  1. 1中国人民解放军总医院医学院病理科,北京 100853,中国
    2透彻影像人工智能实验室,北京 100853,中国
  • 收稿日期:2021-06-29 接受日期:2021-08-18 出版日期:2021-09-30 发布日期:2021-08-31
  • 通讯作者: 宋志刚 E-mail:songzhg301@139.com

Histopathological Diagnosis System for Gastritis Using Deep Learning Algorithm

Wei Ba1,Shuhao Wang2,Cancheng Liu2,Yuefeng Wang2,Huaiyin Shi1,Zhigang Song1,*()   

  1. 1Department of Pathology, Chinese PLA General Hospital & Medical School, Beijing 100853, China
    2Artificial Intelligence Lab, Thorough Images,Beijing 100853, China
  • Received:2021-06-29 Accepted:2021-08-18 Published:2021-09-30 Online:2021-08-31
  • Contact: Zhigang Song E-mail:songzhg301@139.com

摘要:

目的 开发一种用于慢性胃炎病理分类的深度学习算法,并使用全切片病理图像(whole slide images,WSI)评估其性能。
方法 回顾性收集解放军总医院胃活检标本1,250例(胃炎1,128例,正常胃黏膜122例)。分别使用1,008张和100张WSIs,基于DeepLab v3(ResNet-50)架构训练和验证深度学习算法,并在142张WSIs的独立测试集上测试该算法对不同胃炎亚型的诊断效能。
结果 模型为测试集中的慢性浅表性胃炎(chronic superficial gastritis,CSuG)、慢性活动性胃炎(chronic active gastritis,CAcG)和慢性萎缩性胃炎(chronic atrophic gastritis,CAtG)作出诊断所生成的受试者工作特征(receiver operating characteristic,ROC)曲线的曲线下面积分别为0.882、0.905和0.910。深度学习算法对CSuG、CAcG和CAtG分类的敏感性和特异性分别为0.790和1.000(准确度0.880)、0.985和0.829(准确度0.901)、0.952和0.992(准确度0.986)。对三种不同类型胃炎诊断的总体准确度为 0.867。通过在 WSI 中标记算法识别的可疑区域,可以生成更为透明和可解释的诊断结果。
结论 深度学习算法使用WSI对慢性胃炎进行病理学分型具有较高的准确性。通过预先标记出不同类型胃炎的区域,深度学习算法可以作为辅助诊断工具,提高病理医生的工作效率。

关键词: 人工智能, 深度学习, 算法, 胃炎, 全切片病理图像

Abstract:

Objective To develope a deep learning algorithm for pathological classification of chronic gastritis and assess its performance using whole-slide images (WSIs).
Methods We retrospectively collected 1,250 gastric biopsy specimens (1,128 gastritis, 122 normal mucosa) from PLA General Hospital. The deep learning algorithm based on DeepLab v3 (ResNet-50) architecture was trained and validated using 1,008 WSIs and 100 WSIs, respectively. The diagnostic performance of the algorithm was tested on an independent test set of 142 WSIs, with the pathologists’ consensus diagnosis as the gold standard.
Results The receiver operating characteristic (ROC) curves were generated for chronic superficial gastritis (CSuG), chronic active gastritis (CAcG), and chronic atrophic gastritis (CAtG) in the test set, respectively.The areas under the ROC curves (AUCs) of the algorithm for CSuG, CAcG, and CAtG were 0.882, 0.905 and 0.910, respectively. The sensitivity and specificity of the deep learning algorithm for the classification of CSuG, CAcG, and CAtG were 0.790 and 1.000 (accuracy 0.880), 0.985 and 0.829 (accuracy 0.901), 0.952 and 0.992 (accuracy 0.986), respectively. The overall predicted accuracy for three different types of gastritis was 0.867. By flagging the suspicious regions identified by the algorithm in WSI, a more transparent and interpretable diagnosis can be generated.
Conclusion The deep learning algorithm achieved high accuracy for chronic gastritis classification using WSIs. By pre-highlighting the different gastritis regions, it might be used as an auxiliary diagnostic tool to improve the work efficiency of pathologists.

Key words: artificial intelligence, deep learning, algorithm, gastritis, whole-slide pathological images

基金资助: 解放军总医院医疗大数据与人工智能研发项目(2019MBD-038)

Copyright © 2018 Chinese Academy of Medical Sciences. All right reserved.
 
www.cmsj.cams.cn
京公安备110402430088 京ICP备06002729号-1  Powered by Magtech.

Supervised by National Health Commission of the People's Republic of China

9 Dongdan Santiao, Dongcheng district, Beijing, 100730 China

Tel: 86-10-65105897  Fax:86-10-65133074 

E-mail: cmsj@cams.cn  www.cmsj.cams.cn

Copyright © 2018 Chinese Academy of Medical Sciences

All right reserved.

京公安备110402430088  京ICP备06002729号-1