Refinement analysis with clustering segmentation for port wine stains

Abstract

Although both flashlamp-pumped pulsed dye laser and vascular-targeted photodynamic therapy have demonstrated safety and efficacy for port wine stains (PWS), the clinical results of these methods may be limited because the laser parameters used are virtually identical for all PWS patients despite their variations from site to site. A method that combines digital color-image processing with clustering segmentation based on color similarity and spatial proximity is proposed to refine the severity grading of PWS lesions. The evaluation in the color spaces of CIE-L*a*b*, value of a*, and erythema index are analyzed and compared, and the severity level grading is refined with the reference of a 10-level red color bar. The proposed clustering method can be used as a detailed objective assessment for PWS with a low-cost camera.