Objective evaluation of treatment effects on port-wine stains using L*a*b* color coordinates.

Wide variations in port-wine stains and their responses to various therapies pose a need for the development of an objective method to evaluate the effects of treatment. Several techniques such as laser Doppler, reflectance spectrometry, and tristimulus colorimetry have been used to evaluate the color of port-wine stains, but these techniques are limited by cost, small test size area, and other factors. Therefore, we developed a simple and cost-effective method of evaluating treatment results on port-wine stains using the L*a*b* color coordinate system in combination with a personal computer. For 22 patients with port-wine stains, the slide photographs were digitized using a slide scanner. L*a*b* color differences of the normal control and port-wine stain sites were obtained before and after treatment, and treatment effect (percent) was calculated. By calculating each color difference between the lesion and normal skin both before and after treatment, problems arising from different illuminating conditions during photography were minimized. The results were compared with the visual evaluation conducted by three experienced plastic surgeons. The treatment effects analyzed by L*a*b* color coordinate ranged from 4 to 95 percent, with a mean of 48.1 percent, whereas treatment effects evaluated by the plastic surgeons ranged from 15 to 92 percent, with a mean of 51.1 percent. The subjective clinical grades correlated well with the treatment effects obtained by the proposed color analysis system (correlation coefficient, 0.89). The maximum difference in the effect of treatment for a patient evaluated by the three clinicians was up to 60 percent, which means that visual judgment is very subjective and variable. The color analysis system proposed as a result of this study is very easy, objective, quantitative, cost-effective, and can be useful for the evaluation of treatment effects on colored skin lesions such as port-wine stains.