The Relationship Between Computerized Face and Tongue Image Segmentation and Metabolic Parameters in Patients with...

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Artificial Intelligence in Medicine

Background:

• Type 2 diabetes mellitus (T2DM) management requires continuous monitoring of metabolic parameters.
• Traditional monitoring methods can be invasive or time-consuming.
• Novel, non-invasive methods for assessing T2DM status are needed.

Purpose of the Study:

• To investigate the correlation and predictive value of facial and tongue features for metabolic parameters in T2DM patients.
• To establish linear regression relationships between imaging features and metabolic indicators.
• To explore the potential of AI-driven analysis of facial and tongue characteristics for T2DM assessment.

Main Methods:

• A cohort of 269 T2DM patients was studied using AI-powered tongue imaging (XiMaLife).
• Facial and tongue imaging features were collected and analyzed using advanced machine learning algorithms.
• Image features were correlated with blood examination results, including HbA1c, GA, and other metabolic markers.

Main Results:

• Significant correlations were found between multiple facial and tongue features and key metabolic parameters (HbA1c, GA, FPG, C-peptide, insulin).
• Imaging features also showed associations with hepatic, renal, cardiac, and thyroid function indicators, as well as blood cell counts.
• Linear regression analyses indicated that facial and tongue imaging parameters partially determined metabolic indicators.

Conclusions:

• Facial and tongue features are intimately associated with the metabolic status and overall health indicators in T2DM.
• AI and machine learning analysis of these features offer a promising non-invasive approach for predicting T2DM conditions and progression.