Deep Learning-based Trichoscopic Image Analysis and Quantitative Model for Predicting Basic and Specific Classification in Male Androgenetic Alopecia
Keywords:deep learning, androgenetic alopecia, trichoscopic image, BASP classification
Since the results of basic and specific classification in male androgenetic alopecia are subjective, and trichoscopic data, such as hair density and diameter distribution, are potential quantitative indicators, the aim of this study was to develop a deep learning framework for automatic trichoscopic image analysis and a quantitative model for predicting basic and specific classification in male androgenetic alopecia. A total of 2,910 trichoscopic images were collected and a deep learning framework was created on convolutional neural networks. Based on the trichoscopic data provided by the framework, correlations with basic and specific classification were analysed and a quantitative model was developed for predicting basic and specific classification using multiple ordinal logistic regression. A deep learning framework that can accurately analyse hair density and diameter distribution on trichoscopic images and a quantitative model for predicting basic and specific classification in male androgenetic alopecia were established.
Lolli F, Pallotti F, Rossi A, Fortuna MC, Caro G, Lenzi A, et al. Androgenetic alopecia: a review. Endocrine 2017; 57: 9-17.
Ellis JA, Sinclair R, Harrap SB. Androgenetic alopecia: pathogenesis and potential for therapy. Expert Rev Mol Med 2002; 4: 1-11.
Hamilton JB. Patterned loss of hair in man; types and incidence. Ann N Y Acad Sci 1951; 53: 708-728.
Lee WS, Ro BI, Hong SP, Bak H, Sim WY, Kim DW, et al. A new classification of pattern hair loss that is universal for men and women: basic and specific (BASP) classification. J Am Acad Dermatol 2007; 57: 37-46.
Ishino A, Takahashi T, Suzuki J, Nakazawa Y, Iwabuchi T, Tajima M. Contribution of hair density and hair diameter to the appearance and progression of androgenetic alopecia in Japanese men. Br J Dermatol 2014; 171: 1052-1059.
Hayashi S, Miyamoto I, Takeda K. Measurement of human hair growth by optical microscopy and image analysis. Br J Dermatol 1991; 125: 123-129.
Van Neste D, Trüeb RM. Critical study of hair growth analysis with computer-assisted methods. J Eur Acad Dermatol Venereol 2006; 20: 578-583.
Saraogi PP, Dhurat RS. Automated digital image analysis (TrichoScan®) for human hair growth analysis: ease versus errors. Int J Trichology 2010; 2: 5-13.
Du-Harpur X, Watt FM. What is AI? Applications of artificial intelligence to dermatology. Br J Dermatol 2020; 183: 423-430.
Cullell-Dalmau M, Otero-Vinas M, Manzo C. Research techniques made simple: deep learning for the classification of dermatological images. J Invest Dermatol 2020; 140: 507-514.e501.
Jiang YQ, Xiong JH, Li HY, Yang XH, Yu WT, Gao M, et al. Recognizing basal cell carcinoma on smartphone-captured digital histopathology images with a deep neural network. Br J Dermatol 2020; 182: 754-762.
Shelhamer E, Long J, Darrell T. Fully convolutional networks for semantic segmentation. IEEE Trans Pattern Anal Mach Intell 2017; 39: 640-651.
Yan LC, Yoshua B, Geoffrey H. Deep learning. Nature 2015; 521: 436-444.
Howard AG, Zhu M, Chen B, Kalenichenko D, Wang W, Weyand T, et al. MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. [accessed 10th June 2019]. Available from: https://arxiv.org/abs/1704.04861.
Schwarz MW, Cowan WB, Beatty JC. An experimental comparison of RGB, YIQ, LAB, HSV, and opponent color models. ACM Transactions on Graphics (TOG) 1987; 6: 123-158.
Botev A, Lever G, Barber D. Nesterov's accelerated gradient and momentum as approximations to regularised update descent. 2017 International Joint Conference on Neural Networks (IJCNN): IEEE, 2017: p. 1899-1903.
Bottou L. Large-scale machine learning with stochastic gradient descent. In: Lechevallier Y, Saporta G, editors. Proceedings of COMPSTAT'2010. Heidelberg: Physica-Verlag HD; 2010: p. 177-186.
Willmott CJ, Matsuura K. Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance. Climate Res 2005; 30: 79-82.
Yu KH, Beam AL, Kohane IS. Artificial intelligence in healthcare. Nat Biomed Eng 2018; 2: 719-731.
Shen D, Wu G, Suk HI. Deep Learning in medical image analysis. Annu Rev Biomed Eng 2017; 19: 221-248.
Urban G, Feil N, Csuka E, Hashemi K, Ekelem C, Choi F, et al. Combining deep learning with optical coherence tomography imaging to determine scalp hair and follicle counts. Lasers Surg Med 2021; 53: 171-178.
Lee S, Lee JW, Choe SJ, Yang S, Koh SB, Ahn YS, et al. Clinically applicable deep learning framework for measurement of the extent of hair loss in patients with alopecia areata. JAMA Dermatol 2020; 156: 1018-1020.
Liu W, Anguelov D, Erhan D, Szegedy C, Reed S, Fu C-Y, et al. Ssd: Single shot multibox detector. European conference on computer vision: Springer, 2016: p. 21-37.
Rushton DH, Ramsay ID, Norris MJ, Gilkes JJ. Natural progression of male pattern baldness in young men. Clin Exp Dermatol 1991; 16: 188-192.
Whiting DA. Diagnostic and predictive value of horizontal sections of scalp biopsy specimens in male pattern androgenetic alopecia. J Am Acad Dermatol 1993; 28: 755-763.
How to Cite
Copyright (c) 2021 Meng Gao, Yue Wang, Haipeng Xu, Congcong Xu, Xianhong Yang, Jin Nie, Ziye Zhang, Zhixuan Li, Wei Hou, Yiqun Jiang
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (https://creativecommons.org/licenses/by-nc/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for non-commercial purposes, provided proper attribution to the original work.