Age-dependent changes in epidermal architecture explored using an automated image analysis algorithm on in vivo...

Area of Science:

• Dermatology
• Medical Imaging
• Computational Biology

Background:

• Reflectance confocal microscopy (RCM) enables noninvasive in vivo visualization of epidermal structure.
• Manual analysis of RCM images for cellular parameters is time-consuming and prone to error.
• Automated cell identification is needed for efficient RCM image analysis.

Purpose of the Study:

• To develop and validate an automated algorithm for cell identification in RCM images.
• To analyze epidermal architecture changes in children and adults using the automated method.
• To quantify skin maturation and aging processes through cellular and structural parameters.

Main Methods:

• Region-of-interest and individual cell identification using Sato and Gabor filters.
• Post-processing for improved cell detection and outlier removal.
• Application to 5345 RCM images from children and adults; calculation of cell area, perimeter, density, and epidermal layer thicknesses using deep learning.

Main Results:

• Epidermal keratinocyte size increases with age in children and adults.
• Skin maturation is a dynamic process throughout childhood and adulthood.
• Stratum Corneum and supra-papillary epidermis thickness increase with age, faster in children.

Conclusions:

• The automated methodology efficiently analyzes large RCM datasets for skin physiology parameters.
• The findings confirm dynamic skin maturation in childhood and aging in adulthood.
• This approach facilitates large-scale studies on skin development and aging.