Deep learning-based, computer-aided classifier developed with dermoscopic images shows comparable performance to 164 dermatologists in cutaneous disease diagnosis in the Chinese population.

Publisher: Chinese Medical Association ; produced by Wolters Kluwer Country of Publication: China NLM ID: 7513795 Publication Model: Print Cited Medium: Internet ISSN: 2542-5641 (Electronic) Linking ISSN: 03666999 NLM ISO Abbreviation: Chin Med J (Engl) Subsets: MEDLINE

Publication: <2015- > : Beijing : Chinese Medical Association ; produced by Wolters Kluwer
Original Publication: Peking, Chinese Medical Assn.

Deep Learning* ; Melanoma*/diagnostic imaging ; Skin Neoplasms*/diagnostic imaging; China ; Computers ; Dermatologists ; Dermoscopy ; Humans

Background: Diagnoses of Skin diseases are frequently delayed in China due to lack of dermatologists. A deep learning-based diagnosis supporting system can facilitate pre-screening patients to prioritize dermatologists' efforts. We aimed to evaluate the classification sensitivity and specificity of deep learning models to classify skin tumors and psoriasis for Chinese population with a modest number of dermoscopic images.
Methods: We developed a convolutional neural network (CNN) based on two datasets from a consecutive series of patients who underwent the dermoscopy in the clinic of the Department of Dermatology, Peking Union Medical College Hospital, between 2016 and 2018, prospectively. In order to evaluate the feasibility of the algorithm, we used two datasets. Dataset I consisted of 7192 dermoscopic images for a multi-class model to differentiate three most common skin tumors and other diseases. Dataset II consisted of 3115 dermoscopic images for a two-class model to classify psoriasis from other inflammatory diseases. We compared the performance of CNN with 164 dermatologists in a reader study with 130 dermoscopic images. The experts' consensus was used as the reference standard except for the cases of basal cell carcinoma (BCC), which were all confirmed by histopathology.
Results: The accuracies of multi-class and two-class models were 81.49% ± 0.88% and 77.02% ± 1.81%, respectively. In the reader study, for the multi-class tasks, the diagnosis sensitivity and specificity of 164 dermatologists were 0.770 and 0.962 for BCC, 0.807 and 0.897 for melanocytic nevus, 0.624 and 0.976 for seborrheic keratosis, 0.939 and 0.875 for the "others" group, respectively; the diagnosis sensitivity and specificity of multi-class CNN were 0.800 and 1.000 for BCC, 0.800 and 0.840 for melanocytic nevus, 0.850 and 0.940 for seborrheic keratosis, 0.750 and 0.940 for the "others" group, respectively. For the two-class tasks, the sensitivity and specificity of dermatologists and CNN for classifying psoriasis were 0.872 and 0.838, 1.000 and 0.605, respectively. Both the dermatologists and CNN achieved at least moderate consistency with the reference standard, and there was no significant difference in Kappa coefficients between them (P > 0.05).
Conclusions: The performance of CNN developed with relatively modest number of dermoscopic images of skin tumors and psoriasis for Chinese population is comparable with 164 dermatologists. These two models could be used for screening in patients suspected with skin tumors and psoriasis respectively in primary care hospital.

Zhao S, Xie B, Li Y, Zhao X, Kuang Y, Su J, et al. Smart identification of psoriasis by images using convolutional neural networks: a case study in China. J Eur Acad Dermatol Venereol 2020; 34:518524. doi: 10.1111/jdv.15965. (PMID: 10.1111/jdv.15965)
Bobadilla F, Wortsman X, Munoz C, Segovia L, Espinoza M, Jemec GB. Pre-surgical high resolution ultrasound of facial basal cell carcinoma: correlation with histology. Cancer Imaging 2008; 8:163172. doi: 10.1102/1470-7330.2008.0026. (PMID: 10.1102/1470-7330.2008.0026)
Christenson LJ, Borrowman TA, Vachon CM, Tollefson MM, Otley CC, Weaver AL, et al. Incidence of basal cell and squamous cell carcinomas in a population younger than 40 years. JAMA 2005; 294:681690. doi: 10.1001/jama.294.6.681. (PMID: 10.1001/jama.294.6.681)
Hauschild A, Egberts F, Garbe C, Bauer J, Grabbe S, Hamm H, et al. Melanocytic nevi. J Dtsch Dermatol Ges 2011; 9:723734. doi: 10.1111/j.1610-0387.2011.07741.x. (PMID: 10.1111/j.1610-0387.2011.07741.x)
Minagawa A. Dermoscopy-pathology relationship in seborrheic keratosis. J Dermatol 2017; 44:518524. doi: 10.1111/1346-8138.13657. (PMID: 10.1111/1346-8138.13657)
Jackson JM, Alexis A, Berman B, Berson DS, Taylor S, Weiss JS. Current understanding of seborrheic keratosis: prevalence, etiology, clinical presentation, diagnosis, and management. J Drugs Dermatol 2015; 14:11191125.
Armstrong AW. Psoriasis. JAMA Dermatol 2017; 153:956doi: 10.1001/jamadermatol.2017.2103. (PMID: 10.1001/jamadermatol.2017.2103)
Daniyal M, Akram M, Zainab R, Munir N, Shah SMA, Liu B, et al. Progress and prospects in the management of Psoriasis and developments in Phyto-therapeutic modalities. Dermatol Ther 2019; 32:e12866doi: 10.1111/dth.12866. (PMID: 10.1111/dth.12866)
Napolitano M, Caso F, Scarpa R, Megna M, Patrì A, Balato N, et al. Psoriatic arthritis and psoriasis: differential diagnosis. Clin Rheumatol 2016; 35:18931901. doi: 10.1007/s10067-016-3295-9. (PMID: 10.1007/s10067-016-3295-9)
LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015; 521:436444. doi: 10.1038/nature14539. (PMID: 10.1038/nature14539)
Kumar A, Kim J, Lyndon D, Fulham M, Feng D. An ensemble of fine-tuned convolutional neural networks for medical image classification. IEEE J Biomed Health Inform 2017; 21:3140. doi: 10.1109/JBHI.2016.2635663. (PMID: 10.1109/JBHI.2016.2635663)
Mai SM, Sheha MA, Sharawy A. Automatic detection of melanoma skin cancer using texture analysis. Int J Comput Appl 2012; 42:2226. doi: 10.5120/5817-8129. (PMID: 10.5120/5817-8129)
Han SS, Park GH, Lim W, Kim MS, Na JI, Park I, et al. Deep neural networks show an equivalent and often superior performance to dermatologists in onychomycosis diagnosis: automatic construction of onychomycosis datasets by region-based convolutional deep neural network. PLoS One 2018; 13:e0191493doi: 10.1371/journal.pone.0191493. (PMID: 10.1371/journal.pone.0191493)
Vestergaard ME, Macaskill P, Holt PE, Menzies SW. Dermoscopy compared with naked eye examination for the diagnosis of primary melanoma: a meta-analysis of studies performed in a clinical setting. Br J Dermatol 2008; 159:669676. doi: 10.1111/j.1365-2133.2008.08713.x. (PMID: 10.1111/j.1365-2133.2008.08713.x)
Rosendahl C, Tschandl P, Cameron A, Kittler H. Diagnostic accuracy of dermatoscopy for melanocytic and nonmelanocytic pigmented lesions. J Am Acad Dermatol 2011; 64:10681073. doi: 10.1016/j.jaad.2010.03.039. (PMID: 10.1016/j.jaad.2010.03.039)
Braun RP, Ludwig S, Marghoob AA. Differential diagnosis of seborrheic keratosis: clinical and dermoscopic features. J Drugs Dermatol 2017; 16:835842.
Errichetti E, Stinco G. Dermoscopy in general dermatology: a practical overview. Dermatol Ther (Heidelb) 2016; 6:471507. doi: 10.1007/s13555-016-0141-6. (PMID: 10.1007/s13555-016-0141-6)
Skin Imaging Group, Combination of Tradtional and Western Medicine Dermatology; Skin Imaging Group, China International Exchange and Promotion Association for Medical and Healthcare; Skin Imaging Group, Chinese Society of Dermatology, Subcommittee on Dermatologic Surgery, China Dermatologist Association; Skin Imaging Equipment Group, Committee on Skin Disease and Cosmetic Dermatology, China Association of Medical Equipment. Dermoscopic characteristics of basal cell carcinoma: a Chinese expert consensus statement (2019) (in Chinese). Chin J Dermatol 2019; 52:371377. doi: 10.3760/cma.j.issn.0412-4030.2019.06.001. (PMID: 10.3760/cma.j.issn.0412-4030.2019.06.001)
National Clinical Research Center for Skin and Immune Diseases; Huaxia Shin Image and Artificial Intelligence Cooperation, China International Exchange and Promotion Association for Medical and Healthcare; Skin Imaging Group, Dermatology Branch of China International Exchange and Promotion Association for Medical and Healthcare; Skin Imaging Group, Combination of Traditional and Western Medicine Dermatology; Subcommittee on Dermatologic Surgery, China Dermatologist Association; Skin Imaging Group (Preparing), Chinese Society of Dermatology, et al. Dermoscopic features of solar lentigo, seborrheic keratosis and lichen planus-like keratosis: an expert consensus statement (in Chinese). Chin J Dermatol 2019; 52:878883. doi: 10.35541/cjd.20190480.20. (PMID: 10.35541/cjd.20190480.20)
Qiao JJ, Zou XB, Dong HT, Xin LL. The diagnosis of dermoscopy in pigmented nevus (in Chinese). Chin J Leprosy Skin Dis 2017; 33:6569.
Lallas A, Kyrgidis A, Tzellos TG, Apalla Z, Karakyriou E, Karatolias A, et al. Accuracy of dermoscopic criteria for the diagnosis of psoriasis, dermatitis, lichen planus and pityriasis rosea. Br J Dermatol 2012; 166:11981205. doi: 10.1111/j.1365-2133.2012.10868.x. (PMID: 10.1111/j.1365-2133.2012.10868.x)
Fujisawa Y, Otomo Y, Ogata Y, Nakamura Y, Fujita R, Ishitsuka Y, et al. Deep learning-based, computer-aided classifier developed with a small dataset of clinical images surpasses board-certified dermatologists in skin tumor diagnosis. Br J Dermatol 2018; 179:373381. doi: 10.1111/bjd.16924. (PMID: 10.1111/bjd.16924)
Panchal G, Ganatra A, Kosta YP, Panchal D. Behaviour analysis of multilayer perceptrons with multiple hidden neurons and hidden layers. Int J Comput Theory Eng 2011; 3:332337. doi: 10.7763/IJCTE.2011.V3.328. (PMID: 10.7763/IJCTE.2011.V3.328)
Esteva A, Kuprel B, Novoa RA, Ko J, Swetter SM, Blau HM, et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature 2017; 542:115118. doi: 10.1038/nature21056. (PMID: 10.1038/nature21056)
Shin HC, Roth HR, Gao M, Lu L, Xu Z, Nogues I, et al. Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning. IEEE Trans Med Imaging 2016; 35:12851298. doi: 10.1109/tmi.2016.2528162. (PMID: 10.1109/tmi.2016.2528162)
Szegedy C, Vanhoucke V, Ioffe S, Shlens J, Wojna Z. Rethinking the Inception Architecture for Computer Vision. IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, USA; 2016. p. 2818–2826.
Haenssle HA, Fink C, Schneiderbauer R, Toberer F, Buhl T, Blum A, et al. Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists. Ann Oncol 2018; 29:18361842. doi: 10.1093/annonc/mdy166. (PMID: 10.1093/annonc/mdy166)
Zhang X, Wang S, Liu J, Tao C. Towards improving diagnosis of skin diseases by combining deep neural network and human knowledge. BMC Med Inform Decis Mak 2018; 18: (Suppl 2): 59doi: 10.1186/s12911-018-0631-9. (PMID: 10.1186/s12911-018-0631-9)
Han SS, Kim MS, Lim W, Park GH, Park I, Chang SE. Classification of the clinical images for benign and malignant cutaneous tumors using a deep learning algorithm. J Invest Dermatol 2018; 138:15291538. doi: 10.1016/j.jid.2018.01.028. (PMID: 10.1016/j.jid.2018.01.028)
Tschandl P, Argenziano G, Razmara M, Yap J. Diagnostic accuracy of content-based dermatoscopic image retrieval with deep classification features. Br J Dermatol 2018; 181:155165. doi: 10.1111/bjd.17189. (PMID: 10.1111/bjd.17189)
Brinker TJ, Hekler A, Utikal JS, Grabe N, Schadendorf D, Klode J, et al. Skin cancer classification using convolutional neural networks: systematic review. J Med Internet Res 2018; 20:e11936doi: 10.2196/11936. (PMID: 10.2196/11936)
Navarrete-Dechent C, Dusza SW, Liopyris K, Marghoob AA, Halpern AC, Marchetti MA. Automated dermatological diagnosis: hype or reality? J Invest Dermatol 2018; 138:22772279. doi: 10.1016/j.jid.2018.04.040. (PMID: 10.1016/j.jid.2018.04.040)
Maron RC, Weichenthal M, Utikal JS, Hekler A, Berking C, Hauschild A, et al. Systematic outperformance of 112 dermatologists in multiclass skin cancer image classification by convolutional neural networks. Eur J Cancer 2019; 119:5765. doi: 10.1016/j.jaad.2017.08.016. (PMID: 10.1016/j.jaad.2017.08.016)
Hekler A, Utikal JS, Enk AH, Hauschild A, Weichenthal M, Maron RC, et al. Superior skin cancer classification by the combination of human and artificial intelligence. Eur J Cancer 2019; 120:114121. doi: 10.1016/j.ejca.2019.07.019. (PMID: 10.1016/j.ejca.2019.07.019)
Brinker TJ, Hekler A, Enk AH, Berking C, Haferkamp S, Hauschild A, et al. Deep neural networks are superior to dermatologists in melanoma image classification. Eur J Cancer 2019; 119:1117. doi: 10.1016/j.ejca.2019.05.023. (PMID: 10.1016/j.ejca.2019.05.023)
Brinker TJ, Hekler A, Enk AH, Klode J, Hauschild A, Berking C, et al. Deep learning outperformed 136 of 157 dermatologists in a head-to-head dermoscopic melanoma image classification task. Eur J Cancer 2019; 113:4754. doi: 10.1016/j.ejca.2019.04.001. (PMID: 10.1016/j.ejca.2019.04.001)
Brinker TJ, Hekler A, Enk AH, Klode J, Hauschild A, Berking C, et al. A convolutional neural network trained with dermoscopic images performed on par with 145 dermatologists in a clinical melanoma image classification task. Eur J Cancer 2019; 111:148154. doi: 10.1016/j.ejca.2019.02.005. (PMID: 10.1016/j.ejca.2019.02.005)
Jun G, Shukui Q, Jun L, Tongyu L, Lu S, Xiaohong C, et al. Chinese guidelines on the diagnosis and treatment of melanoma (2015 edition) (in Chinese). Chin Clin Oncol 2016; 5:57doi: 10.21037/cco.2015.12.02. (PMID: 10.21037/cco.2015.12.02)
Marchetti MA, Codella NCF, Dusza SW, Gutman DA, Helba B, Kalloo A, et al. Results of the 2016 international skin imaging collaboration international symposium on biomedical imaging challenge: comparison of the accuracy of computer algorithms to dermatologists for the diagnosis of melanoma from dermoscopic images. J Am Acad Dermatol 2018; 78:270277. doi: 10.1016/j.jaad.2017.08.016. (PMID: 10.1016/j.jaad.2017.08.016)
Wang S, Liu J, Zhu C, Shu C, Zhou H, Xie F, et al. Comparison of diagnostic performance of dermatologists versus deep convolutional neural network for dermoscopic images of pigmented nevus and seborrheic keratosis (in Chinese). Chin J Dermatol 2018; 51:486489. doi: 10.3760/cma.j.issn.0412-4030.2018.07.002. (PMID: 10.3760/cma.j.issn.0412-4030.2018.07.002)
Ding X, Wang T, Shen Y, Wang X, Zhou C, Tian S, et al. Prevalence of psoriasis in China: a population-based study in six cities. Eur J Dermatol 2012; 22:663667. doi: 10.1684/ejd.2012.1802. (PMID: 10.1684/ejd.2012.1802)
Li CX, Shen CB, Xue K, Shen X, Jing Y, Wang ZY, et al. Artificial intelligence in dermatology: past, present, and future. Chin Med J 2019; 132:20172020. doi: 10.1097/CM9.0000000000000372. (PMID: 10.1097/CM9.0000000000000372)

Date Created: 20200823 Date Completed: 20210514 Latest Revision: 20210514

20231215

PMC7478660

10.1097/CM9.0000000000001023

32826613