卷积神经网络实时评估白光内镜下幽门螺杆菌感染：一项前瞻性、多中心研究（附视频）
Real-Time Evaluation of Helicobacter pylori Infection by Convolution Neural Network During White-Light Endoscopy: A Prospective, Multicenter Study (With Video).
机构信息
Department of Gastroenterology, Sir Run Run Shaw Hospital, Medical School, Zhejiang University, Hangzhou, China.
West China Xiamen Hospital, Sichuan University, Xiamen, China.
出版信息
Clin Transl Gastroenterol. 2023 Oct 1;14(10):e00643. doi: 10.14309/ctg.0000000000000643.
INTRODUCTION
Convolutional neural network during endoscopy may facilitate evaluation of Helicobacter pylori infection without obtaining gastric biopsies. The aim of the study was to evaluate the diagnosis accuracy of a computer-aided decision support system for H. pylori infection (CADSS-HP) based on convolutional neural network under white-light endoscopy.
METHODS
Archived video recordings of upper endoscopy with white-light examinations performed at Sir Run Run Shaw Hospital (January 2019-September 2020) were used to develop CADSS-HP. Patients receiving endoscopy were prospectively enrolled (August 2021-August 2022) from 3 centers to calculate the diagnostic property. Accuracy of CADSS-HP for H. pylori infection was also compared with endoscopic impression, urea breath test (URT), and histopathology. H. pylori infection was defined by positive test on histopathology and/or URT.
RESULTS
Video recordings of 599 patients who received endoscopy were used to develop CADSS-HP. Subsequently, 456 patients participated in the prospective evaluation including 189 (41.4%) with H. pylori infection. With a threshold of 0.5, CADSS-HP achieved an area under the curve of 0.95 (95% confidence interval [CI], 0.93-0.97) with sensitivity and specificity of 91.5% (95% CI 86.4%-94.9%) and 88.8% (95% CI 84.2%-92.2%), respectively. CADSS-HP demonstrated higher sensitivity (91.5% vs 78.3%; mean difference = 13.2%, 95% CI 5.7%-20.7%) and accuracy (89.9% vs 83.8%, mean difference = 6.1%, 95% CI 1.6%-10.7%) compared with endoscopic diagnosis by endoscopists. Sensitivity of CADSS-HP in diagnosing H. pylori was comparable with URT (91.5% vs 95.2%; mean difference = 3.7%, 95% CI -1.8% to 9.4%), better than histopathology (91.5% vs 82.0%; mean difference = 9.5%, 95% CI 2.3%-16.8%).
DISCUSSION
CADSS-HP achieved high sensitivity in the diagnosis of H. pylori infection in the real-time test, outperforming endoscopic diagnosis by endoscopists and comparable with URT. Clinicaltrials.gov ; ChiCTR2000030724.
简介
在胃镜检查中使用卷积神经网络可能有助于在不获取胃活检的情况下评估幽门螺杆菌感染。本研究的目的是评估基于卷积神经网络的幽门螺杆菌感染计算机辅助决策支持系统（CADSS-HP）在白光内镜下的诊断准确性。
方法
使用来自浙江大学医学院附属邵逸夫医院（2019 年 1 月至 2020 年 9 月）的上消化道内镜白光检查的存档视频记录来开发 CADSS-HP。前瞻性纳入 3 个中心于 2021 年 8 月至 2022 年 8 月期间接受内镜检查的患者来计算诊断特性。还将 CADSS-HP 对幽门螺杆菌感染的诊断准确性与内镜印象、尿素呼气试验（UBT）和组织病理学进行了比较。幽门螺杆菌感染的定义为组织病理学和/或 UBT 阳性。
结果
使用 599 名接受内镜检查患者的视频记录来开发 CADSS-HP。随后，456 名患者参与了前瞻性评估，其中 189 名（41.4%）患有幽门螺杆菌感染。当阈值为 0.5 时，CADSS-HP 的曲线下面积为 0.95（95%置信区间 [CI]，0.93-0.97），灵敏度和特异性分别为 91.5%（95% CI 86.4%-94.9%）和 88.8%（95% CI 84.2%-92.2%）。CADSS-HP 的诊断灵敏度（91.5%比 78.3%；平均差异=13.2%，95% CI 5.7%-20.7%）和准确性（89.9%比 83.8%，平均差异=6.1%，95% CI 1.6%-10.7%）均高于内镜医师的内镜诊断。CADSS-HP 诊断幽门螺杆菌感染的敏感性与 UBT 相当（91.5%比 95.2%；平均差异=3.7%，95% CI -1.8%至 9.4%），优于组织病理学（91.5%比 82.0%；平均差异=9.5%，95% CI 2.3%-16.8%）。
讨论
CADSS-HP 在实时检测中对幽门螺杆菌感染的诊断具有高灵敏度，优于内镜医师的内镜诊断，与 UBT 相当。Clinicaltrials.gov；ChiCTR2000030724。
Zotero 插件