Evaluation of intercellular lipid lamellae in the stratum corneum by polarized microscopy

BACKGROUND: Intercellular lipids contain a lamellar structure that glows in polarized images. It could be expected that the intercellular lipid content be estimated from the luminance values calculated from polarized images of stratum corneum strips. Therefore, we attempted to develop a method for simple and rapid evaluation of the intercellular lipid content through a procedure. Herein, we demonstrated a relationship between the luminance value and the amount of ceramides, one of the main components of intercellular lipids. MATERIALS AND METHODS: The stratum corneum was collected from the forearm using slides with a pure rubber‐based adhesive, which did not produce unnecessary luminescence under polarizing conditions. Images were analyzed using luminance indices. The positive secondary ion peak images were obtained using the time of flight‐secondary ion mass spectrometry; the polarized and brightfield images were obtained using a polarized microscope. The ceramide and protein amount was measured by high‐performance liquid chromatography and bicinchoninic acid protein assay after microscope imaging. Images and quantitative values were used to construct evaluation models based on a convolutional neural network (CNN). RESULTS: There was a correlation between the highlighted areas of the polarized image to overlap with the area where ceramide‐derived peak was detected. Evaluation of the CNN‐based model of the polarized images predicted the amount of ceramides per unit of stratum corneum. CONCLUSION: The method proposed in the study enabled a large number of specimens to provide a simple, rapid, and efficient evaluation of the intercellular lipid content.