Deep-learning-based clinical decision support system for gastric neoplasms in real-time endoscopy: Development and validation study.

Abstract

Authors previously established deep-learning models to predict the histopathology and invasion-depth of gastric lesions using endoscopic images. This study aimed to establish and validate a deep-learning-based clinical decision support system (CDSS) for the automated detection and classification (diagnosis and invasion-depth prediction) of gastric neoplasms in real-time endoscopy. The same 5,017 endoscopic images, which were employed to establish previous models, were used for the training data. The primary outcomes were the 1. Lesion-detection rate for the detection model and 2. Lesion-classification accuracy for the classification model. For the performance validation of lesion-detection model, 2,524 real-time procedures were tested in a randomized pilot study. Consecutive patients were allocated either to CDSS-assisted screening endoscopy or conventional screening endoscopy. The lesion-detection rate was compared between the groups. For the performance validation of lesion-classification model, a prospective multicenter external-test was conducted using 3,976 novel images from five institutions. The lesion-detection rate was 95.6% (internal-test). For the performance validation, CDSS-assisted endoscopy showed higher lesion-detection rate compared to conventional screening endoscopy, although statistically not significant (2.0% vs. 1.3%, P-value=0.21) (randomized study). The lesion-classification rate was 89.7% in the four-class classification (advanced-, early gastric cancer, dysplasia, and non-neoplasm) and 89.2% in the invasion-depth prediction (mucosa-confined or submucosa-invaded) (internal-test). For the performance validation, CDSS reached 81.5% accuracy in the four-class classification and 86.4% accuracy in the binary classification (prospective multicenter external-test). The CDSS demonstrated potential for real-clinic application and high performance in terms of lesion detection and classification of detected lesions in the stomach.