The goal of this clinical trial is to verify the auxiliary role of the artificial intelligence (AI) system in mediastinal endoscopic ultrasound (EUS) scans. The main questions it aims to answer are as follows:
1. A comparison of the image recognition accuracy between the AI system and EUS endoscopists. 2. Whether the AI system can improve the quality of mediastinal scans for EUS endoscopists.
Participants will conduct mediastinal EUS with or without the assistance of the AI system.
None
Mediastinal lesions have previously undergone surgery or radiotherapy and chemotherapy;
Mental illness, drug addiction, inability to express themselves or other diseases that may affect follow-up.
Parallel
Artificial intelligence device
DeviceThis AI device can provide prompts for standard sites of pancreatic scans and anatomical structures. It may also suggest the next operation methods.
The completeness for standard station scanning
The primary outcome measure was the completeness for standard station scanning of EUS by endoscopists in the two groups. This was calculated as the number of stations successfully scanned divided by the total number of stations that should have been scanned.
The completeness of anatomical structure scanning
It calculated as the number of anatomical structures successfully scanned divided by the total number of structures that should have been scanned.
Procedure time
It defined as the observation time only, excluding the time spent on puncture, elastography, contrast-enhanced ultrasound, or other lesion observation or treatment.
The incidence of adverse events
The incidence of adverse events in both groups during the entire research process.
Click on the lead sponsor to view all of their trials.
The Third Xiangya Hospital of Central South University
Lead sponsor
ZhuZhou Central Hospital
Collaborator
The First Affiliated Hospital of University of South China
Collaborator
The First People's Hospital of Chenzhou
Collaborator
the Affiliated Nanhua Hospital, University of South China
Collaborator
