Study On The Severity Of Skin Lesions In Psoriasis Vulgaris

BackgroundAseessment of the severity of psoriasis is critical to the judgement of patients’condition, the choices of treatment, and the evaluation of treatment efficacy. Among a variety of assessment tools of psoriasis, Psoriasis Area and Severity Index, or PASI score is the most validated and detailed one. However, PASI scores are estimated subjectively, which often leads to significant inter-and intra-individual variation in scores. Thus, an objective and automatic scoring method is required urgently. Nowadays, most of the similar researches conduct computer-aided image analysis based on images of skin lesions taken by digital cameras. There is no similar researches conducting image analysis based on dermoscopic images. The aim of this study is to discuss the practicability of objective measurement of erythema in psoriasis using dermoscopic images.Methods137 patients with clinically diagnosed psoriasis vulgaris (PV) were enrolled in this study. Patients’basic information, condition of PV lesions, treatment choices and so on were noted and PASI score were assessed by a highly experienced dermatologist. One or more typical lesions were taken using a smart phone (iPhone 6s) and a Handyscope (FotoFinder). The Handyscope images included at least one image of the lesion and one of the normal skin around the lesion. Several typical images were chosen according to different erythema scores, and were analyzed primarily by Adobe Photoshop, to study the correlation between color values with the red color of the lesion background and the dilated dermal vessels. Then computer-aided image analyses were conducted to assess the erythema in the order of Color space conversion→Segmentation→SVM classification.ResultsThe red color of the lesion background and the dilated dermal vessels in the dermoscopic images of PV lesions were associated with the erythema scores. The result showed good consistency between computer-aided erythema assessment and that given by dermatologists (κ=0.687,P=0000).ConclusionsQuantitative computer-aided image analysis using dermoscopic images could be an alternative method to determine the severity of PV lesions. Dermoscopy, as an important tool in the teledermatology, may play a critical role in not only assisting dermatologists, providing standard criteria, but also building a patient-management system and a database of dermatologic images.