医学
鞋跟
接收机工作特性
皮肤病科
活检
放射科
内科学
解剖
作者
Linda Tognetti,Sofia Lo Conte,L. Simone,Aimilios Lallas,Elvira Moscarella,Roberta Giuffrida,John Paoli,Emi Dika,Ignazio Stanganelli,Serena Magi,Maria Concetta Fargnoli,Caterina Longo,Gianluca Nazzaro,Paolo Broganelli,Francesco Lacarrubba,Mariano Suppa,Jean‐Luc Perrot,Philipp Tschandl,Harald Kittler,Élisa Cinotti
摘要
Abstract Background Melanoma of the sole is an aggressive rare form, often diagnosed late. Plantar atypical nevi (pAN) are frequently misdiagnosed as plantar early melanomas (pEM) and therefore excised. Our aim was to develop a clinical‐dermoscopic risk‐scoring model to help discriminate these plantar atypical melanocytic lesions (pAMLs). Materials and Methods We collected 490 pAMLs (98 pEM, 392 pAN) paired with histopathological diagnosis, dermoscopic and clinical image, maximum lesion diameter, plantar location and age and sex of the patient from 17 European centres. This plantar dataset was grouped into training (261), validation (174) and testing (55 pAMLs) subsets. European participants (104 dermatologists, 56 residents) performed a blinded tele‐dermoscopic test, including intuitive diagnosis, pattern analysis, rating of case difficulty, diagnostic confidence assessment and management decision. Results A total of 2887 dermoscopic evaluations were obtained. The iDScore_plantar model gave an average area under the receiver operating characteristic curve of 0.95 (against 0.77 for pattern analysis). It was composed of the sum of five scores ( S ) for the following items: maximum diameter 8–12 ( S = 1)/>12 mm ( S = 5); age 40–50 ( S = 2 )/>50 years ( S = 5); location on heel ( S 4) or on toes/plantar eminence ( S = 2); asymmetry of colours ( S = 2) and/or asymmetry of structures ( S = 1). ‘Long/short follow‐up, biopsy, excision’ decisions were matched with four risk ranges: no risk ( S = 0–3), low‐medium risk ( S = 4–8), medium‐high risk ( S = 9–12) and very high risk ( S = 13–17). By applying the model, participants would have reduced the number of misdiagnosed pAN and the number of pAN excised by −25.5% and −27.7%, respectively, and would have increased the number of correctly diagnosed pEM by +18.5%, the number of pEM recommended for surgical excision by +8.5% and the number of pEM recommended directly for surgical excision instead of biopsy by +16.15%. Conclusions The iDScore_plantar model proved to be a simple scoring tool to help clinicians in assigning a progressive risk of malignancy to pAMLs.
科研通智能强力驱动
Strongly Powered by AbleSci AI