Assessment of skin barrier function using skin images with topological data analysis

Area of Science:

• Dermatology
• Biophysics
• Medical Imaging

Background:

• Precision medicine necessitates objective skin phenotype evaluation.
• Current methods for assessing skin barrier function, like transepidermal water loss (TEWL), are inconvenient for daily clinical use.

Purpose of the Study:

• To develop a novel, image-based method for skin barrier assessment.
• To correlate topological features from skin images with TEWL measurements.

Main Methods:

• Utilized topological data analysis (TDA) on microscopic skin images.
• Extracted structural and textural features reflecting skin regularity.
• Trained machine learning models to predict TEWL using identified topological features.

Main Results:

• Identified significant correlations between topological image features and TEWL.
• Achieved good accuracy (R² = 0.524) in predicting TEWL using machine learning models.
• Demonstrated that TDA can efficiently capture relevant structural information from skin images.

Conclusions:

• Topological data analysis of skin images presents a promising, non-invasive method for assessing skin barrier function.
• This image-based approach offers a more convenient alternative to traditional TEWL measurements for clinical applications.