Quantifying facial skin erythema more precisely by analyzing color channels of The VISIA Red images

Background: The VISIA Red images were developed to document and measure facial skin erythema, but diffuse erythema cannot be fully segmented by the VISIA system due to the automatic thresholding segmentation method. Moreover, topical area analysis is not available in the system. Materials and methods: Erythema severity degrees of 20 simulated Red images were designated 1-20 with 1-20 inflammatory lesions for each, respectively. The RGB channel mean values of each simulated image were acquired by ImageJ and relative intensity of red values calculated. Results: The relative intensity of red values positively correlate to erythema severity with a coefficient of 0.999345 (p<0.001). We also proposed a method for calibration when pustules were present in the erythema area. The method was proved by mathematical reasoning and verified by certified dermatologists. Conclusion: We demonstrated a simple and more precise method to quantify and compare facial skin erythema by analyzing the RGB channel values of the VISIA Red images. Our method brings convenience for erythema evaluation in dermatological studies.