Connection Between Tongue Signs and Bile Reflux Analysed With Artificial Intelligence

Not yet recruiting

• Conditions: Tongue; Bile Reflux; Artificial Intelligence

• Registration Number: NCT05369572
• Lead Sponsor: Shandong University

Brief Summary

By introducing artificial intelligence into Chinese medicine tongue diagnosis, we collated and collected tongue images, anxiety and depression scales and gastroscopy reports, mined and analysed the correlation between tongue images and bile reflux and anxiety and depression and constructed a prediction model to analyse the possibility of predicting bile reflux and anxiety and depression in patients based on tongue images.

Detailed Description

Firstly, after the patient signs the informed consent form, the researcher will collect pictures of the patient's tongue and obtain basic information about the patient.

Second, the patients are scored on the Anxiety and Depression Scale.

Thirdly, after the patient undergoes gastroscopy, the patient's gastroscopy report is obtained.

Finally, the patient's tongue image, information and gastroscopy report are matched to construct an artificial intelligence model of tongue image and bile reflux and anxiety and depression, and the quality of the model is assessed.

Study Timeline

• Study First Submitted: 2022-05-05
• Study First Submitted QC: 2022-05-05
• Study First Posted: 2022-05-11
• Status Verified: 2022-06
• Last Update Submitted: 2022-06-20
• Last Update Posted: 2022-06-22
• Study Start: 2022-06-30
• Primary Completion: 2024-06-30
• Study Completion: 2025-06-30

Recruitment & Eligibility

• Status: NOT_YET_RECRUITING
• Sex: All
• Target Recruitment: 1500

Inclusion Criteria

• Patients aged 18 to 80 years who wish to undergo gastroscopy.
• Patients have given their informed consent and signed the informed consent form.

Exclusion Criteria

• Serious heart, liver, kidney or other underlying illness, or mental illness.
• Patients taking anti-anxiety or depression medication within 3 months.
• Current H. pylori infection.
• History of surgery on the digestive or biliary tract.
• Peptic ulcer, malignant tumour of the digestive tract, etc.
• Patients taking bismuth or other staining medications.
• Pregnant or lactating women.

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Negative predictive values (NPV)	3 years	Negative predictive values for artificial intelligence models Negative Predictive Value = True Negative / (True Negative + False Negative) \*100%
Sensitivity	3 years	Sensitivity of artificial intelligence models Sensitivity = number of true positives / (number of true positives + number of false negatives) \* 100%.
Specificity	3 years	Specificity of Artificial Intelligence Models Specificity = number of true negatives / (number of true negatives + number of false positives)); \*100%
Positive predictive values(PPV)	3 years	Positive predictive values from artificial intelligence models Positive predictive value = true positive / (true positive + false positive) \*100%
AUC (95% CI)	3 years	area under the receiver operating characteristic curve (AUC),
Accuracy	3 years	Accuracy for artificial intelligence models Accuracy = (true positives + true negatives) / total number of subjects \* 100%

Trial Locations

Locations (1)

Qilu hosipital; 🇨🇳 Jinan, Shandong, China