深度卷积神经网络对胃病变普通内镜图像诊断的研究  

Diagnosis of routine endoscopic images of gastric lesions through a deep convolutional neural network

作　　者：张黎明[1] 张洋 王俐 王江源[1] 刘玉兰[1] Zhang Liming;Zhang Yang;Wang Li;Wang Jiangyuan;Liu Yulan(Department of Gastroenterology,Peking University People′s Hospital,Beijing 100044,China;Internet Medical Department of Aixin Life Insurance Co.,Ltd.,Beijing 100043,China)

机构地区：[1]北京大学人民医院消化科,100044 [2]爱心人寿互联网医疗部,北京100043

出　　处：《中华消化内镜杂志》2021年第10期789-794,共6页Chinese Journal of Digestive Endoscopy

基　　金：北京大学医学青年科技创新发展平台基金(BMU2018PYB014)。

摘　　要：目的通过深度卷积神经网络(convolutional neural network, CNN)技术实现胃病变内镜图像的快速、准确人工智能辅助诊断。方法收集2012—2018年北京大学人民医院1 121例胃病变的普通白光内镜图像和病理结果。胃病变图像包括消化性溃疡、早期胃癌及高级别上皮内瘤变、进展期胃癌、胃黏膜下肿瘤共4类, 另外还包括无病变正常胃黏膜图像。共17 217张图像作为训练集, 使用CNN ResNet-34模型训练分类任务, 使用全CNN DeepLabv3模型训练像素分割任务。经过训练后的CNN通过一个测试集评估诊断效能, 测试集包括237例胃病变, 共1 091张普通内镜图像。计算CNN诊断的准确率、敏感度、特异度、阳性预测值、阴性预测值。结果 CNN对于早期胃癌及高级别上皮内瘤变的诊断准确率为78.6%(33/42), 敏感度为84.4%(27/32), 特异度为60.0%(6/10), 阳性预测值87.1%(27/31), 阴性预测值54.5%(6/11);对于消化性溃疡的诊断准确率为90.4%(47/52), 敏感度为92.7%(38/41), 特异度为81.8%(9/11);对于进展期胃癌的诊断准确率为88.1%(52/59), 敏感度为91.8%(45/49), 特异度为70.0%(7/10);对于胃黏膜下肿瘤的诊断准确率为86.0%(43/50), 敏感度为89.7%(35/39), 特异度为72.7%(8/11)。所有测试集图像识别时间为42 s。结论 CNN可以作为早期胃癌及其他胃病变内镜图像的快速辅助识别方法, 识别速度快, 准确率高。Objective To develop a deep convolutional neural network(CNN)to automatically detect gastric lesions in endoscopic images.Methods A CNN-based diagnostic system was constructed based on ResNet-34 residual network structure and DeepLabv3 structure,and trained by using 17217 routine gastroscopy images.These images were from 1121 gastric lesions of five types acquired in Peking University People′s Hospital between 2012 and 2018,namely peptic ulcer(PU),early gastric cancer(EGC)and high-grade intraepithelial neoplasia(HGIN),advanced gastric cancer(AGC),gastric submucosal tumors(SMTs),and normal gastric mucosa without lesions.The trained CNN was evaluated through a test dataset that contained 1091 routine gastroscopy images of 237 gastric lesions.The accuracy,sensitivity,specificity,positive predictive value,and negative predictive value of the CNN were calculated.Results The accuracy,sensitivity,specificity,positive predictive value,and negative predictive value of the CNN-assisted diagnosis of EGC and HGIN were 78.6%(33/42),84.4%(27/32),60.0%(6/10),87.1%(27/31),and 54.5%(6/11),respectively.The accuracy,sensitivity,and specificity of CNN-assisted diagnosis of PU were 90.4%(47/52),92.7%(38/41),and 81.8%(9/11),respectively,the outcomes of AGC were 88.1%(52/59),91.8%(45/49),and 70.0%(7/10),respectively,and those of gastric SMTs were 86.0%(43/50),89.7%(35/39),and 72.7%(8/11),respectively.The CNN′s recognition time for all images of the test set was 42 seconds.Conclusion The constructed CNN system,as a rapid and accurate auxiliary diagnostic instrument,can detect not only EGC and HGIN but also other gastric lesions.

关 键 词：人工智能 胃肿瘤 消化性溃疡 诊断 卷积神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP391.41[自动化与计算机技术—控制科学与工程] R573[医药卫生—消化系统] R735.2[医药卫生—内科学]