Non-contact dermatoscope system for automated detection of suspicious skin lesions

Hannoversches Zentrum für Optische Technologien (HOT), Gottfried Wilhelm Leibniz Universität Hannover; 2 Institut für Produktentwicklung und Gerätebau (IPeG), Gottfried Wilhelm Leibniz Universität Hannover; 3 Institut für Quantenoptik (IQO), Gottfried Wilhelm Leibniz Universität Hannover

dierk.fricke@hot.uni-hannover.de

Abstract

The incidence of skin cancer has constantly increased over the last years. The deadliest form of skin cancer is the malignant melanoma, for which if its early detected, survival rates are high. To distinguish a melanoma from a benign naevus, certain optical criteria need to be evaluated. These allow to estimate the risk whether a skin lesion is a benign or malignant skin change. However, current state dermatoscopy is time consuming when a multitude of skin lesions need to be inspected by the physician. In this work, we present a prototype of an automated skin screening device based on dermatoscopy. To be able to obtain information from various skin depths, optical polarizers were integrated into the system. Also, a novel light source was developed and optimized to additionally gain Information by using algorithms for blood- and melanin-contrast enhancement. In future, our device is intended to enable a fast and reliable diagnosis of melanoma by an automated detection of suspicious skin lesions in dermatoscopic images. Hence, the physician only needs to inspect a much smaller number of lesions which are indicated by the device leading to a drastically reduced inspection time.

Keywords

Beleuchtung Bildverarbeitung Medizinische Anwendungen der Optik
Download PDF
@inproceedings{dgao118-b18, title = {Non-contact dermatoscope system for automated detection of suspicious skin lesions}, author = {D. Fricke, A. Wolf, M. Wollweber, U. Morgner, B. Roth}, booktitle = {DGaO-Proceedings, 118. Jahrestagung}, year = {2017}, publisher = {Deutsche Gesellschaft für angewandte Optik e.V.}, issn = {1614-8436}, note = {Talk B18} }
118. Annual Conference of the DGaO · Dresden · 2017