Medizinische Universität Graz - Research portal
Selected Publication:
SHR Neuro Cancer Cardio Lipid Metab Microb
Fink, C; Blum, A; Buhl, T; Mitteldorf, C; Hofmann-Wellenhof, R; Deinlein, T; Stolz, W; Trennheuser, L; Cussigh, C; Deltgen, D; Winkler, JK; Toberer, F; Enk, A; Rosenberger, A; Haenssle, HA.
Diagnostic performance of a deep learning convolutional neural network in the differentiation of combined nevi and melanomas.
J Eur Acad Dermatol Venereol. 2019;
Doi: 10.1111/jdv.16165
[OPEN ACCESS]
Web of Science
PubMed
FullText
FullText_MUG
- Co-authors Med Uni Graz
- Hofmann-Wellenhof Rainer
- Kränke Teresa Maria
- Altmetrics:
- Dimensions Citations:
- Plum Analytics:
- Scite (citation analytics):
- Abstract:
- Deep learning convolutional neural networks (CNN) may assist physicians in the diagnosis of melanoma. The capacity of a CNN to differentiate melanomas from combined nevi, the latter representing well-known melanoma simulators, has not been investigated. To assess the diagnostic performance of a CNN when used to differentiate melanomas from combined nevi in comparison to dermatologists. In this study a CNN with regulatory approval for the European market (Moleanalyzer-Pro, FotoFinder Systems GmbH, Bad Birnbach, Germany) was used. We attained a dichotomous classification (benign, malignant) in dermoscopic images of 36 combined nevi and 36 melanomas with a mean Breslow thickness of 1.3mm. Primary outcome measures were the CNN's sensitivity, specificity, and the diagnostic odds ratio (DOR) in comparison to 11 dermatologists with different levels of experience. The CNN revealed a sensitivity, specificity, and DOR of 97.1% (95% CI [82.7%-99.6%]), 78.8% (95% CI [62.8%-89.1.3%]), and 34 (95% CI [4.8-239]), respectively. Dermatologists showed a lower mean sensitivity, specificity, and DOR of 90.6% (95% CI [84.1%-94.7%]; p=0.092), 71.0% (95% CI [62.6%-78.1%]; p=0.256), and 24 (95% CI [11.6-48.4]) (p=0.1114). Under the assumption that dermatologists use the CNN to verify their (initial) melanoma diagnosis, dermatologists achieve an increased specificity of 90.3% (95% CI [79.8%-95.6%]) at an almost unchanged sensitivity. The largest benefit was observed in "beginners", who performed worst without CNN-verification (DOR=12) but best with CNN-verification (DOR=98). The tested CNN more accurately classified combined nevi and melanomas in comparison to trained dermatologists. Their diagnostic performance could be improved if the CNN was used to confirm/overrule an initial melanoma diagnosis. Application of a CNN may therefore be of benefit to clinicians. © 2019 European Academy of Dermatology and Venereology.