深度学习有助于在前性随机对照研究中检测食道癌和前体病变
Shao-Wei Li1,2,3, Li-Hui Zhang4,5,6, Yue Cai1
1Department of Gastroenterology, Taizhou Hospital of Zhejiang Province Affiliated to Wenzhou Medical University, Linhai, Zhejiang 317000, China.
Science translational medicine
|April 17, 2024
概括
使用卷积神经网络 (CNN) 的新深度学习系统显著改善了在内镜检查期间检测高风险食道病变 (HrELs) 的性能. 这种人工智能辅助的方法使检测率翻了一番,显示出早期食道癌症诊断的前景.
科学领域:
- 胃肠病学 胃肠病学
- 在瘤学瘤学.
- 人工智能的人工智能
背景情况:
- 无症状食道状细胞癌 (ESCC) 和癌前病变的内镜检测仍然具有挑战性.
- 提高高风险食道病变 (HrEL) 的检测率对于早期干预和改善患者的治疗结果至关重要.
研究的目的:
- 开发和验证一个深层卷积神经网络 (CNN) 系统,用于检测食道癌和HRELs.
- 评估CNN辅助内镜在改善临床实践中的HREL检测率方面的有效性.
主要方法:
- 一项随机对照试验,涉及3117名年龄≥50岁的患者,比较CNN辅助内镜与无辅助内镜.
- 患者被分为1:1的实验 (CNN辅助) 或控制 (无辅助) 组.
- 主要终点是HRELs的检测率.
主要成果:
- 实验组的HREL检测率明显高 (1.8%),与对照组 (0.9%) 相比 (P=0.029).
- 该CNN系统表现出高诊断性能,具有89.7%的灵敏度,98.5%的特异性和98.2%的准确性来检测HREL.
- 没有报告任何不良事件,这表明CNN辅助系统的安全性.
结论:
- 开发的CNN辅助内镜系统有效地提高了内镜手术期间HRELs的检测率.
- 这种人工智能驱动的工具显示了提高食道癌的早期诊断和治疗的潜力,在食道癌查计划中发挥了宝贵的作用.
相关概念视频
Endoscopic Procedures I: Esophagogastroduodenoscopy
111
An Esophagogastroduodenoscopy (EGD) is a diagnostic procedure in which an endoscopist uses a flexible, lighted endoscope to visualize the upper gastrointestinal (GI) tract. The procedure includes visualizing the oropharynx, esophagus, stomach, and the first part of the small intestine, the duodenum.
During an EGD, the endoscope can be used to:
During an EGD, the endoscope can be used to:
111
Barrett Esophagus-I: Introduction
87
Barrett's esophagus is a medical condition where the esophageal mucosa is significantly damaged by stomach acid or other digestive fluids, often due to long-term exposure associated with gastroesophageal reflux disease (GERD). In GERD, a weakened or abnormally relaxed lower esophageal sphincter allows stomach acid to flow persistently into the esophagus.
This constant acid exposure transforms the esophagus's pink mucosal lining (stratified squamous epithelium) into a type of lining more...
This constant acid exposure transforms the esophagus's pink mucosal lining (stratified squamous epithelium) into a type of lining more...
87


