Melanoma recognition by a deep learning convolutional neural network-Performance in different melanoma subtypes and localisations

Julia K Winkler; Katharina Sies; Christine Fink; Ferdinand Toberer; Alexander Enk; Teresa Deinlein; Rainer Hofmann-Wellenhof; Luc Thomas; Aimilios Lallas; Andreas Blum; Wilhelm Stolz; Mohamed S Abassi; Tobias Fuchs; Albert Rosenberger; Holger A Haenssle

doi:10.1016/j.ejca.2019.11.020

Melanoma recognition by a deep learning convolutional neural network-Performance in different melanoma subtypes and localisations

Eur J Cancer. 2020 Mar:127:21-29. doi: 10.1016/j.ejca.2019.11.020. Epub 2020 Jan 20.

Authors

Affiliations

¹ Department of Dermatology, University of Heidelberg, Heidelberg, Germany.
² Department of Dermatology and Venerology, Medical University of Graz, Graz, Austria.
³ Hospices Civils de Lyon, Department of Dermatology, Lyon Sud University Hospital, Pierre Bénite, France.
⁴ First Department of Dermatology, Aristotle University, Thessaloniki, Greece.
⁵ Public, Private and Teaching Practice, Konstanz, Germany.
⁶ Department of Dermatology, Allergology and Environmental Medicine II, Hospital Thalkirchner Street, Munich, Germany.
⁷ Faculty of Computer Science and Mathematics, University of Passau, Passau, Germany.
⁸ Department of Research and Development, FotoFinder Systems GmbH, Bad Birnbach, Germany.
⁹ Institute of Genetic Epidemiology at the Center of Statistics, University of Goettingen, Goettingen, Germany.
¹⁰ Department of Dermatology, University of Heidelberg, Heidelberg, Germany. Electronic address: holger.haenssle@med.uni-heidelberg.de.

PMID: 31972395
DOI: 10.1016/j.ejca.2019.11.020

Abstract

Background: Deep learning convolutional neural networks (CNNs) show great potential for melanoma diagnosis. Melanoma thickness at diagnosis among others depends on melanoma localisation and subtype (e.g. advanced thickness in acrolentiginous or nodular melanomas). The question whether CNN may counterbalance physicians' diagnostic difficulties in these melanomas has not been addressed. We aimed to investigate the diagnostic performance of a CNN with approval for the European market across different melanoma localisations and subtypes.

Methods: The current market version of a CNN (Moleanalyzer-Pro®, FotoFinder Systems GmbH, Bad Birnbach, Germany) was used for classifications (malignant/benign) in six dermoscopic image sets. Each set included 30 melanomas and 100 benign lesions of related localisations and morphology (set-SSM: superficial spreading melanomas and macular nevi; set-LMM: lentigo maligna melanomas and facial solar lentigines/seborrhoeic keratoses/nevi; set-NM: nodular melanomas and papillomatous/dermal/blue nevi; set-Mucosa: mucosal melanomas and mucosal melanoses/macules/nevi; set-AM_skin: acrolentiginous melanomas and acral (congenital) nevi; set-AM_nail: subungual melanomas and subungual (congenital) nevi/lentigines/ethnical type pigmentations).

Results: The CNN showed a high-level performance in set-SSM, set-NM and set-LMM (sensitivities >93.3%, specificities >65%, receiver operating characteristics-area under the curve [ROC-AUC] >0.926). In set-AM_skin, the sensitivity was lower (83.3%) at a high specificity (91.0%) and ROC-AUC (0.928). A limited performance was found in set-mucosa (sensitivity 93.3%, specificity 38.0%, ROC-AUC 0.754) and set-AM_nail (sensitivity 53.3%, specificity 68.0%, ROC-AUC 0.621).

Conclusions: The CNN may help to partly counterbalance reduced human accuracies. However, physicians need to be aware of the CNN's limited diagnostic performance in mucosal and subungual lesions. Improvements may be expected from additional training images of mucosal and subungual sites.

Keywords: Convolutional neural network; Deep learning; Dermoscopy; Melanoma; Nevi.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Aged
Case-Control Studies
Deep Learning*
Female
Follow-Up Studies
Humans
Male
Melanoma / classification*
Melanoma / diagnosis*
Middle Aged
Neural Networks, Computer*
Prognosis
ROC Curve
Retrospective Studies