Author:
Guimarães Pedro,Batista Ana,Zieger Michael,Kaatz Martin,Koenig Karsten
Abstract
AbstractThe diagnostic possibilities of multiphoton tomography (MPT) in dermatology have already been demonstrated. Nevertheless, the analysis of MPT data is still time-consuming and operator dependent. We propose a fully automatic approach based on convolutional neural networks (CNNs) to fully realize the potential of MPT. In total, 3,663 MPT images combining both morphological and metabolic information were acquired from atopic dermatitis (AD) patients and healthy volunteers. These were used to train and tune CNNs to detect the presence of living cells, and if so, to diagnose AD, independently of imaged layer or position. The proposed algorithm correctly diagnosed AD in 97.0 ± 0.2% of all images presenting living cells. The diagnosis was obtained with a sensitivity of 0.966 ± 0.003, specificity of 0.977 ± 0.003 and F-score of 0.964 ± 0.002. Relevance propagation by deep Taylor decomposition was used to enhance the algorithm’s interpretability. Obtained heatmaps show what aspects of the images are important for a given classification. We showed that MPT imaging can be combined with artificial intelligence to successfully diagnose AD. The proposed approach serves as a framework for the automatic diagnosis of skin disorders using MPT.
Publisher
Springer Science and Business Media LLC
Reference37 articles.
1. Weidinger, S., Beck, L. A., Bieber, T., Kabashima, K. & Irvine, A. D. Atopic Dermatitis. in Kanerva’s Occupational Dermatology 1, 201–212 (Springer Berlin Heidelberg, 2012).
2. Drucker, A. M. et al. The Burden of Atopic Dermatitis: Summary of a Report for the National Eczema Association. J. Invest. Dermatol. 137, 26–30 (2017).
3. Gustafson, E., Pacheco, J., Wehbe, F., Silverberg, J. & Thompson, W. A Machine Learning Algorithm for Identifying Atopic Dermatitis in Adults from Electronic Health Records. Proc. - 2017 IEEE Int. Conf. Healthc. Informatics, ICHI 2017 83–90, 10.1109/ICHI.2017.31 (2017).
4. Georgakoudi, I. & Quinn, K. P. Optical imaging using endogenous contrast to assess metabolic state. Annu Rev Biomed Eng 14, 351–367 (2012).
5. Lakowicz, J. R. Principles of Fluorescence Spectroscopy. (Springer, 2007).
Cited by
31 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献