Study on the Effectiveness of Gastroscope Operation Quality Control Based on Artificial Intelligence Technology

Completed

• Conditions: Gastric Cancer

• Registration Number: NCT04384575
• Lead Sponsor: Peking University

Brief Summary

This study aims to construct a real-time quality monitoring system based on artificial intelligence technology.

Detailed Description

Gastroscopy plays an important role in the detection and diagnosis of upper gastrointestinal diseases. It is necessary for endoscopists to operate gastroscope according to the standardized process, in order to avoid missing early lesions. However, with the rapid increase in the number of endoscopies, the workload of endoscopists increases further. High workload reduces the quality of endoscopy, resulting in incomplete observation of anatomical parts that are easy to be missed in the process of gastroscopy. There are significant differences in the operation level of different endoscopists. Therefore, carrying out artificial intelligence methods has good academic research and practical value for improving the quality of endoscopic diagnosis and treatment.

Artificial intelligence devices need to use a large number of endoscopic images, based on this, we intends to collect endoscopic image data from our hospitals for training and validation of the model.

Study Timeline

• Study Start: 2020-02-22
• Study First Submitted: 2020-05-08
• Study First Submitted QC: 2020-05-08
• Study First Posted: 2020-05-12
• Primary Completion: 2022-01-13
• Study Completion: 2022-05-01
• Status Verified: 2023-11
• Last Update Submitted: 2023-11-28
• Last Update Posted: 2023-11-29

Recruitment & Eligibility

• Status: COMPLETED
• Sex: All
• Target Recruitment: 1570

Inclusion Criteria

1. Patiens aged 18 years or above undergoing gastroscopy；
2. Be able to read, understand and sign informed consent；

Exclusion Criteria

1. Patients with absolute contraindications to endoscopy examination;
2. pregnant women;
3. previous history of gastric surgery;
4. the researcher considers that the subject is not suitable for clinical trial.

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Sensitivity	2020.2.22-2020.7.1	number of images in which AI correctly diagnosed positive/all images with positive
Specificity	2020.2.22-2020.7.1	number of images in which AI correctly diagnosed negative/all images negative
Accuracy	2020.2.22-2020.7.1	Calculate the accuracy of AI's judgment on images

Trial Locations

Locations (1)

Beijing Cancer Hospital; 🇨🇳 Beijing, Haidian, China