Characterisation of black skin stratum corneum by digital macroscopic images analysis

Géraud M Azehoun-Pazou; Kokou M Assogba; Hugues Adegbidi; Antoine C Vianou

doi:10.1049/htl.2020.0057

Characterisation of black skin stratum corneum by digital macroscopic images analysis

Healthc Technol Lett. 2020 Dec 15;7(6):161-167. doi: 10.1049/htl.2020.0057. eCollection 2020 Dec.

Authors

Géraud M Azehoun-Pazou^{1

2}, Kokou M Assogba², Hugues Adegbidi³, Antoine C Vianou^{2

4}

Affiliations

¹ National University of Sciences, Technologies, Engineering and Mathematics (UNSTIM), BP 2282 Abomey, Benin.
² Laboratory of Electrical Engineering, Telecommunications and Applied Informatics (LETIA), University of Abomey-Calavi, 01 BP 2009, Abomey-Calavi, Benin.
³ Department of Dermatology and Venerology, Faculty of Health Sciences, University of Abomey-Calavi, 01 BP 188, Abomey-Calavi, Benin.
⁴ Laboratory of Thermophysical Characterization and Energetic Appropriation (Lab-CTMAE), Polytechnic School of Abomey-Calavi, 01 BP 2009, Abomey-Calavi, Benin.

Abstract

Black skin medical images generally show very low contrast. Being in a global initiative of characterisation of black skin horny layer (stratum corneum) by digital images analysis, the authors in this study proposed a four-step approach. The first step consists of differentiation between probable healthy skin regions and those affected. For that, they used an automatic classification system based on multilayer perceptron artificial neural networks. The network has been trained with texture and colour features. Best features selection and network architecture definition were done using sequential network construction algorithm-based method. After classification, selected regions undergo a colour transformation, in order to increase the contrast with the lesion region. Thirdly, created colour information serves as the basis for a modified fuzzy c-mean clustering algorithm to perform segmentation. The proposed method, named neural network-based fuzzy clustering, was applied to many black skin lesion images and they obtained segmentation rates up to 94.67%. The last stage consists in calculating characteristics. Eight parameters are concerned: uniformity, standard deviation, skewness, kurtosis, smoothness, entropy, and average pixel values calculated for red and blue colour channels. All developed methods were tested with a database of 600 images and obtained results were discussed and compared with similar works.

Keywords: 600 images; automatic classification system; biomedical optical imaging; black skin horny layer; black skin lesion images; black skin medical images; black skin stratum corneum; blue colour channels; colour features; colour transformation; created colour information; digital images analysis; digital macroscopic images analysis; feature extraction; features selection; four-step approach; fuzzy set theory; global initiative; image classification; image colour analysis; image segmentation; image texture; lesion region; medical image processing; multilayer perceptron artificial neural networks; multilayer perceptrons; named neural network-based fuzzy clustering; neural nets; obtained results; pattern clustering; probable healthy skin regions; red colour channels; selected regions; sequential network construction algorithm-based method; skin; texture.