Fuzong Clinical Medical College, Fujian Medical University, Fuzhou, China.
Department of Digestive Diseases, 900th Hospital of PLA, Fuzhou, China.
Endoscopy. 2024 May;56(5):334-342. doi: 10.1055/a-2252-4874. Epub 2024 Feb 27.
Inaccurate Forrest classification may significantly affect clinical outcomes, especially in high risk patients. Therefore, this study aimed to develop a real-time deep convolutional neural network (DCNN) system to assess the Forrest classification of peptic ulcer bleeding (PUB).
A training dataset (3868 endoscopic images) and an internal validation dataset (834 images) were retrospectively collected from the 900th Hospital, Fuzhou, China. In addition, 521 images collected from four other hospitals were used for external validation. Finally, 46 endoscopic videos were prospectively collected to assess the real-time diagnostic performance of the DCNN system, whose diagnostic performance was also prospectively compared with that of three senior and three junior endoscopists.
The DCNN system had a satisfactory diagnostic performance in the assessment of Forrest classification, with an accuracy of 91.2% (95%CI 89.5%-92.6%) and a macro-average area under the receiver operating characteristic curve of 0.80 in the validation dataset. Moreover, the DCNN system could judge suspicious regions automatically using Forrest classification in real-time videos, with an accuracy of 92.0% (95%CI 80.8%-97.8%). The DCNN system showed more accurate and stable diagnostic performance than endoscopists in the prospective clinical comparison test. This system helped to slightly improve the diagnostic performance of senior endoscopists and considerably enhance that of junior endoscopists.
The DCNN system for the assessment of the Forrest classification of PUB showed satisfactory diagnostic performance, which was slightly superior to that of senior endoscopists. It could therefore effectively assist junior endoscopists in making such diagnoses during gastroscopy.
不准确的 Forrest 分类可能会显著影响临床结果，尤其是在高危患者中。因此，本研究旨在开发一种实时深度卷积神经网络（DCNN）系统，以评估消化性溃疡出血（PUB）的 Forrest 分类。
回顾性收集来自中国福州第 900 医院的训练数据集（3868 张内镜图像）和内部验证数据集（834 张图像）。此外，还从其他四家医院收集了 521 张图像用于外部验证。最后，前瞻性收集了 46 段内镜视频，以评估 DCNN 系统的实时诊断性能，其诊断性能也与三位高级和三位初级内镜医生的进行了前瞻性比较。
DCNN 系统在 Forrest 分类评估方面具有良好的诊断性能，在验证数据集中的准确率为 91.2%（95%CI 89.5%-92.6%），宏平均受试者工作特征曲线下面积为 0.80。此外，DCNN 系统可以实时视频中自动判断可疑区域，并具有 92.0%（95%CI 80.8%-97.8%）的准确率。在前瞻性临床比较测试中，DCNN 系统的诊断性能优于内镜医生，且更准确和稳定。该系统有助于略微提高高级内镜医生的诊断性能，显著提高初级内镜医生的诊断性能。
用于评估 PUB 的 Forrest 分类的 DCNN 系统具有令人满意的诊断性能，略优于高级内镜医生。因此，它可以有效地帮助初级内镜医生进行胃镜检查时做出此类诊断。
