The Research of AI Assistant Gastroscope Training

Not Applicable

Completed

• Conditions: Gastroscopy; Artificial Intelligence; Training
• Interventions: Other: Artificial intelligence assistant system

• Registration Number: NCT04682821
• Lead Sponsor: Renmin Hospital of Wuhan University

Brief Summary

In this study, we proposed a prospective study about the effectiveness of artificial intelligence system for gastroscope training in novice endoscopists. The subjects would be divided into two groups. The experimental group would be trained in painless gastroscopy with the assistance of the artificial intelligence assistant system. The artificial intelligence assistant system can prompt abnormal lesions and the parts covered by the examination (the stomach is divided into 26 parts). The control group would receive routine painless gastroscopy training without special prompts. Then we compare the gastroscopy operation score, coverage rate of blind spots in gastroscopy,check the average test score before and after training, training satisfaction, detection rate of lesions and so on between the two group.

Detailed Description

In this study, we proposed a prospective study about the effectiveness of artificial intelligence system for gastroscope training in novice endoscopists. The subjects would be divided into two groups. The experimental group would be trained in painless gastroscopy with the assistance of the artificial intelligence assistant system. The artificial intelligence assistant system can prompt abnormal lesions and the parts covered by the examination (the stomach is divided into 26 parts). The control group would receive routine painless gastroscopy training without special prompts. Then we compare the gastroscopy operation score, coverage rate of blind spots in gastroscopy,check the average test score before and after training, training satisfaction, detection rate of lesions and so on between the two group.

Study Timeline

• Status Verified: 2020-12
• Study First Submitted: 2020-12-20
• Study First Submitted QC: 2020-12-20
• Study Start: 2020-12-23
• Study First Posted: 2020-12-24
• Primary Completion: 2021-05-25
• Study Completion: 2021-06-26
• Last Update Submitted: 2021-10-23
• Last Update Posted: 2021-10-26

Recruitment & Eligibility

• Status: COMPLETED
• Sex: All
• Target Recruitment: 288

Inclusion Criteria

1. Males or females who are over 18 years old;
2. After qualified medical education and obtained the Certificate of Chinese medical practitioner;

Exclusion Criterial:

1. A doctor who has already been trained in gastroenteroscopy;
2. Doctors without qualified medical education and didn't obtaine the Certificate of Chinese medical practitioner;
3. The researcher believes that the subjects are not suitable for participating in clinical trials.

Exclusion Criteria

Not provided

Study & Design

• Study Type: INTERVENTIONAL
• Study Design: PARALLEL

Arm && Interventions

Group	Intervention	Description
with Artificial intelligence assistant system	Artificial intelligence assistant system	The experiment group would receive the training with the help of artificial intelligence assistant system in addition to the common training. The system is an non-invasive AI system which could help the endoscopists to diagnosis and monitor the blind spot during the gastroscope.

Primary Outcome Measures

Name	Time	Method
Gastroscopy operation score	three month	Using a professional gastroscopy operation scoring scale, the full score is 100 points, and the score is divided into small items. In this experiment, the effect of training between the two groups was compared by comparing the scores of gastroscopy operation in the experimental group and the control group.

Secondary Outcome Measures

Name	Time	Method
Coverage rate of blind spots in gastroscopy	three month	Evaluate the gastroscope operation videos retained by each physician during the examination, and calculate the coverage of 26 parts of the gastric mucosa in the experimental group and the control group during the examination. The calculation method is: the coverage rate of the blind area of the gastroscopy = the actual number of parts covered by the examination/26 parts of the stomach x 100%.
Check the average test score before and after training	three month	the difference between the theoretical test score after the training and the theoretical test score before the training, the calculation method: the theoretical test score after the training-the theoretical test score before the training.
Training satisfaction	three month	An AI assistant group fills out a questionnaire after training, and determines the satisfaction with AI assistant training through a grading method.
Detection rate of lesions	three month	the detection rate of lesions in the experimental group and the control group by gastroscopy. Calculation method = number of gastroscopes with detected lesions/total number of gastroscopes completed by beginner physicians x 100%.

Trial Locations

Locations (1)

Renmin Hospital of Wuhan University; 🇨🇳 Wuhan, China