Using AI analysis of eye images to identify Traditional Chinese Medicine constitutions

Diagnostic Performance of an Artificial Intelligence Driven Model for Traditional Chinese Medicine Constitution Classification Using Ophthalmic Imaging

This project will test whether an AI model can use eye images to identify biased TCM constitutions in adults aged 18–60 with chronic diseases.

Quick facts

• Study type: Observational
• Enrollment: 1024 (estimated)
• Ages: 18 Years to 45 Years
• Sex: All
• Sponsor: Second Affiliated Hospital of Nanchang University (Academic / other)
• Drugs / interventions: chemotherapy, immunotherapy
• Locations: 1 site (Nanchang, Jiangxi)
• Trial ID: NCT07127939 on ClinicalTrials.gov

What this trial studies

This observational project will develop and validate a multimodal deep-learning model that uses standardized ophthalmic images to classify TCM constitutions. Investigators will collect eye photographs from adults aged 18–60 with chronic systemic diseases at the Second Affiliated Hospital of Nanchang University and label data using expert TCM classification and questionnaires. Model performance will be quantified on held-out validation sets using standard diagnostic metrics and compared with expert classifications. The aim is to create a fully automated, high-throughput tool that could support earlier prediction and more personalized management in integrated Chinese and Western medicine.

Who should consider this trial

Good fit: Adults aged 18–60 with chronic systemic diseases (for example cardiovascular disease, diabetes, chronic respiratory disease, or malignancy) who can attend imaging at the study site are ideal candidates.

Not a fit: People younger than 18 or older than 60, those without chronic systemic disease, individuals who cannot obtain clear ophthalmic images, or those unable to travel to the Nanchang site are unlikely to receive benefit from this project.

Why it matters

Potential benefit: If successful, this could provide patients an objective, scalable eye-image tool to determine TCM constitution and support earlier, personalized care.

How similar studies have performed: Related AI work has been applied to tongue and pulse analysis, but ocular-based constitution classification is largely novel and has not yet shown widespread clinical success.

Eligibility criteria

Inclusion Criteria:

1. Age: Participants aged between 18 and 60 years;
2. Chronic Systemic Diseases: Defined according to the standards of the United States National Center for Chronic Disease Prevention and Health Promotion and the "Healthy China 2030" planning outline issued by the State Council. These are characterized as conditions lasting one year or longer that require continuous medical care, limit daily activities, or both. The four major chronic diseases in China include:Cardiovascular and cerebrovascular diseases,Malignant neoplasms,Diabetes mellitus and Chronic respiratory diseases;
3. Cardiovascular and Cerebrovascular Diseases: This category includes:Hypertension: Stable hypertension without acute episodes, defined as systolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, or currently undergoing antihypertensive treatment.Coronary Heart Disease: Diagnosed based on a history of coronary heart disease or confirmed through imaging evidence such as coronary angiography or CT angiography.

   Heart Failure: Classified as New York Heart Association (NYHA) functional class II-IV with a left ventricular ejection fraction \<40%.Stroke: Diagnosed based on a history of ischemic or hemorrhagic stroke or confirmed through imaging evidence such as MRI or CT scans;
4. Malignant Neoplasms: Participants with newly diagnosed, recurrent, metastatic malignant tumors, or those currently receiving specific treatments (e.g., chemotherapy, radiotherapy, immunotherapy), confirmed by histological or cytological examinations;
5. Diabetes Mellitus: Participants meeting at least one of the following criteria:

   Fasting blood glucose ≥7.0 mmol/L (126 mg/dL),2-hour post-oral glucose tolerance test blood glucose ≥11.1 mmol/L (200 mg/dL),Hemoglobin A1c ≥6.5%,Random blood glucose ≥11.1 mmol/L (200 mg/dL) accompanied by typical symptoms (e.g., polydipsia, polyuria, weight loss);
6. Chronic Respiratory Diseases: This category includes:Chronic Obstructive Pulmonary Disease (COPD): Post-bronchodilator FEV₁/FVC \<0.70 and FEV₁ \<80% of the predicted value, accompanied by at least one clinical symptom such as chronic cough, sputum production, or dyspnea.Asthma: Reversible airflow limitation, defined as a post-bronchodilator FEV₁ increase of ≥12% and ≥200 ml, accompanied by at least one clinical symptom such as recurrent wheezing, dyspnea, chest tightness, or cough;
7. Healthy Subjects: Individuals who have not experienced any systemic diseases within the past year that require continuous medical care, limit daily activities, or both;
8. Clear Diagnosis Based on Traditional Chinese Medicine and Western Medicine: Participants must have a definitive diagnosis established through both Traditional Chinese Medicine and Western Medicine methodologies;
9. Absence of Other Ocular or Systemic Organic Diseases: Participants should have no other ocular diseases and/or systemic organic lesions that significantly affect the acquisition of ocular imaging.

Exclusion Criteria:

1. Incomplete Clinical Data Supporting Diagnosis: Participants for whom the clinical data necessary to substantiate the diagnosis are incomplete;
2. Significant Opacification of Other Ocular Optical Systems: Individuals presenting with pronounced obscurations in other ocular optical systems (e.g., severe corneal opacification, nuclear cataract grade IV or higher, severe vitreous hemorrhage or opacification) that impede the completion of comprehensive ocular examinations;
3. Acute Systemic Organic Diseases: Patients concurrently suffering from acute systemic organic conditions (such as acute infections or decompensated organ failure) that prevent participation in the constitution identification procedures;
4. Pregnancy or Lactation: Individuals who are pregnant or currently breastfeeding.

Where this trial is running

Nanchang, Jiangxi

• The Second Affiliated Hospital of Nanchang University — Nanchang, Jiangxi, China (Recruiting)

Study contacts

• Study coordinator: Jiang Xiong

Conditions Ophthalmic Imaging; Deep Learning; Artificial Intelligence; Traditional Chinese Medicine Constitution; Traditional Chinese medicine constitution