Skin Disease Diagnosis from Photographs Using Deep Learning

This work aims to study performance of different deep learning based approaches to classify skin diseases automatically from colored digital photographs. We applied recent network models, which are U-Net, Inception Version-3 (InceptionV3), Inception and Residual Network (InceptionResNetV2), VGGNet, and Residual Network (ResNet). Comparative evaluations of the results obtained by these network models indicated that automated diagnosis from digital photographs is possible with accuracy between 74% (by U-net) and 80% (by ResNet). Therefore, further studies are still required in this area to design and develop a new model by combining advantages of different network models and also to obtain higher accuracy. In addition, testing of the model should be performed with more data including more diversity to see reliability of the model.

[1]  Eduardo Valle,et al.  Deep-Learning Ensembles for Skin-Lesion Segmentation, Analysis, Classification: RECOD Titans at ISIC Challenge 2018 , 2018, ArXiv.

[2]  Mohammed Feham,et al.  M-Health: Skin Disease Analysis System Using Smartphone's Camera , 2013, ANT/SEIT.

[3]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Sergey Ioffe,et al.  Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning , 2016, AAAI.

[5]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[6]  John Paoli,et al.  Expert-Level Diagnosis of Nonpigmented Skin Cancer by Combined Convolutional Neural Networks , 2019, JAMA dermatology.

[7]  R. Hay,et al.  The global burden of skin disease in 2010: an analysis of the prevalence and impact of skin conditions. , 2014, The Journal of investigative dermatology.

[8]  Y. Fujisawa,et al.  Deep‐learning‐based, computer‐aided classifier developed with a small dataset of clinical images surpasses board‐certified dermatologists in skin tumour diagnosis , 2018, The British journal of dermatology.

[9]  Dumitru Erhan,et al.  Going deeper with convolutions , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[10]  Jürgen Schmidhuber,et al.  Highway Networks , 2015, ArXiv.

[11]  Man Jae Kim.  Development of a Portable Optical Imaging System based on a Smartphone and Image Classification using a Learning Algorithm , 2017 .

[12]  Giovanni Pellacani,et al.  Validity and Reliability of Dermoscopic Criteria Used to Differentiate Nevi From Melanoma: A Web-Based International Dermoscopy Society Study. , 2016, JAMA dermatology.

[13]  Sergey Ioffe,et al.  Rethinking the Inception Architecture for Computer Vision , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[14]  Mohammed Bennamoun,et al.  Resfeats: Residual network based features for image classification , 2016, 2017 IEEE International Conference on Image Processing (ICIP).

[15]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[16]  H. Kittler,et al.  Diagnostic accuracy of dermoscopy. , 2002, The Lancet. Oncology.

[17]  Tianfu Wang,et al.  Dense deconvolution net: Multi path fusion and dense deconvolution for high resolution skin lesion segmentation , 2018, Technology and health care : official journal of the European Society for Engineering and Medicine.

[18]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[19]  João Paulo Papa,et al.  Computational methods for pigmented skin lesion classification in images: review and future trends , 2018, Neural Computing and Applications.

[20]  Jian Sun,et al.  Convolutional neural networks at constrained time cost , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[21]  Nicholas B. MacKinnon,et al.  Vanishing point: a smartphone application that classifies acne lesions and estimates prognosis , 2016 .

[22]  João Paulo Papa,et al.  Computational methods for the image segmentation of pigmented skin lesions: A review , 2016, Comput. Methods Programs Biomed..

[23]  D. Margolis,et al.  The burden of skin disease in the United States , 2017, Journal of the American Academy of Dermatology.