Clinical Scorecard: Smart Analysis of Tongue and Facial Images for Noninvasive Prediction of Glucolipid Metabolic Disorders

At a Glance

Key Highlights

• Glucolipid metabolic disorders are a major global health threat with increasing prevalence in China.
• Traditional Chinese Medicine (TCM) visual diagnostic methods, including tongue and facial image analysis, offer noninvasive diagnostic potential.
• Integration of artificial intelligence and machine learning enhances the accuracy of tongue and facial image-based diagnosis.

Guideline-Based Recommendations

Diagnosis

• Use fasting blood glucose ≥ 7.0 mmol/L, 2-hour postprandial glucose ≥ 11.0 mmol/L, or HbA1c ≥ 6.5% for type 2 diabetes diagnosis.
• Diagnose dyslipidemia with total cholesterol ≥ 6.21 mmol/L, LDL cholesterol ≥ 4.1 mmol/L, triglycerides ≥ 2.3 mmol/L, or non-HDL cholesterol ≥ 4.9 mmol/L.
• Diagnose MASLD based on upper abdominal ultrasound showing enhanced liver echogenicity and related features.
• Employ standardized tongue and facial image collection protocols using TFDA-1 Instrument for noninvasive assessment.

Management

• Early intervention and long-term health management are crucial for patients with glucose and lipid metabolism disorders.
• Incorporate TCM visual diagnostic methods alongside conventional biochemical testing to enhance disease assessment.

Monitoring & Follow-up

• Regular monitoring of biochemical markers including ALT, AST, γ-GT, ALP, FBG, BUN, CREA, UA, CHO, TG, HDL, and LDL.
• Use standardized tongue and facial image analysis to track disease progression and response to treatment.

Risks

• High prevalence of dyslipidemia and hypertension increases risk of atherosclerotic cardiovascular disease (ASCVD).
• Delayed diagnosis and intervention may lead to increased morbidity and mortality.

Patient & Prescribing Data

615 patients undergoing routine health examinations, including 297 with glucolipid metabolic disorders and 318 controls

Noninvasive tongue and facial image analysis combined with biochemical data can improve diagnostic accuracy and support early intervention strategies.

Clinical Best Practices

• Ensure patients fast for at least 30 minutes before tongue and facial image collection to avoid contamination.
• Use standardized equipment and controlled environment settings for image acquisition to maintain quality and consistency.
• Combine TCM visual diagnostic techniques with modern biochemical and imaging assessments for comprehensive evaluation.
• Apply machine learning algorithms for precise segmentation and analysis of tongue and facial images to enhance diagnostic performance.

Related Resources & Content

• International Clinical Practice Guidelines for Traditional Chinese Medicine
• Shanghai University of Traditional Chinese Medicine Affiliated Baoshan Hospital Study Protocol