For Manuscript Submission, Check or Review Login please go to Submission Websites List.
For the academic login, please select your country in the dropdown list. You will be redirected to verify your credentials.
Diagnostic Reliability of Dermoscopic Criteria for Detecting Malignant MelanomaSoyer H.P. · Smolle J. · Leitinger G. · Rieger E. · Kerl H.
Department of Dermatology, University of Graz, Austria
Background: Recently a new standardized terminology in dermoscopy has been provided by a Consensus Meeting held by the Committee on Analytical Morphology of the Arbeitsgemeinschaft Dermatologische Forschung in Hamburg in order to be applied to further studies in this field. Objective: In this study on 159 pigmented skin tumors including 65 melanomas, the validity of various dermoscopic criteria proposed by the Consensus Meeting for detecting melanoma was evaluated. Methods: In each lesion, a detailed clinical and dermoscopic examination, photographic documentation of the clinical and dermoscopic appearance, surgical excision and histopathologic evaluation were performed. Statistical analyses including χ2 statistics, logistic regression analysis and CART (classification and regression tree) analysis were applied. Results: The diagnosis of melanoma using dermoscopy could be obtained easily, if combinations of the following criteria were observed: whitish veil, pigment network alterations (e.g. irregular pigment network, narrow pigment network, broad pigment network), irregular extensions, black dots and gray-blue areas. Interestingly, however, clinical and dermoscopic examination to detect melanomas yielded the same results, namely a sensitivity of 94%. The combined use of both methods led to an increase in the diagnostic sensitivity of 95%, whereas the combination of clinical and dermoscopic examination with logistic regression analysis of dermoscopic criteria enabled us to detect all 65 melanomas in our data set, thus providing a sensitivity of 100%. Furthermore, an algorithm for the classification of melanoma based on the evaluation of dermoscopic criteria has been constructed by CART analysis showing that the presence of a whitish veil in combination with a pigment network inevitably indicates melanoma. Conclusion: The validity of the various dermoscopic criteria set forth by the Consensus Meeting could be demonstrated as some criteria (e.g. whitish veil, irregular pigment network, irregular extensions) were observed with a significant higher frequency in melanomas.
© 1995 S. Karger AG, Basel