Using Artificial Intelligence with Endoscopic Ultrasound to Assess Early Esophageal Cancer Invasion Depth A Multicenter, Prospective, Randomized Study

What this Trial Is About

Researchers are evaluating an artificial intelligence system to assist in classifying how deeply early esophageal squamous cell carcinoma invades tissue using endoscopic ultrasound. This helps provide a basis for preoperative T staging and supports diagnosis and treatment decisions. The study is a multicenter, prospective, randomized cohort trial, comparing AI-assisted grading to routine diagnosis methods in patients with early esophageal cancer or precancerous lesions. Participants will be randomly assigned to one of two groups the AI group, which uses artificial intelligence to assist in determining the invasion depth under endoscopic ultrasound, or the control group, which receives routine diagnosis without AI assistance. Each group is expected to include about 100 patients. The study uses central randomization with single masking, where the procedure is identified as protocol A or B by the operating physician based on allocation. During the study, researchers will assess the accuracy of grading the depth of cancer invasion over two years. They will also monitor secondary outcomes such as survival rate over three years and progression-free survival over one year. Participants undergo endoscopic ultrasound and pathological examinations, with detailed records reviewed. The study will last several years, ensuring thorough evaluation and follow-up to understand the impact of AI-assisted diagnosis on early esophageal cancer staging.

AI-assisted Endoscopic Ultrasound Grading of Early Esophageal Cancer Invasion Depth: A Multicenter, Prospective, Randomized Cohort Study