Classification of established atopic dermatitis in children with the in vivo imaging methods

• Published in: Journal of biophotonics Vol. 12; no. 5; pp. e201800148 - n/a
• Main Authors: Lee, Jyh‐Hong, Shih, Yuan‐Ta, Wei, Ming‐Liang, Sun, Chi‐Kuang, Chiang, Bor‐Luen
• Format: Journal Article
• Language: English
• Published: Weinheim WILEY‐VCH Verlag GmbH & Co. KGaA 01.05.2019; Wiley Subscription Services, Inc
• Subjects: Atopic dermatitis; Children; Classification; dermal fibrosis; Dermatitis; Dermis; Eczema; epidermal hyperplasia; Fibrosis; Forsterite; harmonic generation; Harmonic generations; Image reconstruction; Immune response; Immune system; in vivo imaging; In vivo methods and tests; Mathematical models; Medical imaging; Microscopy; Parameter identification; Signal generation; Skin diseases; Statistical analysis; atopic dermatitis; in vivo imaging; dermal fibrosis; harmonic generation; epidermal hyperplasia
• DOI: 10.1002/jbio.201800148
• ISSN: 1864-063X

Atopic dermatitis (AD) is a cutaneous disease resulting from a defective barrier and dysregulated immune response. The severity scoring of atopic dermatitis (SCORAD) is used to classify AD. Noninvasive imaging approaches supplementary to SCORAD were investigated. Cr:forsterite laser‐based microscopy was employed to analyze endogenous third‐harmonic generation (THG) and second‐harmonic generation (SHG) signals from skin. Imaging parameters were compared between different AD severities. Three‐dimensional reconstruction of imaged skin layers was performed. Finally, statistic models from quantitative imaging parameters were developed for predicting disease severity. Our data demonstrate that THG signal intensity of lesional skin in AD were significantly increased and was positively correlated with AD severity. Characteristic gray level co‐occurrence matrix (GLCM) values were observed in more severe AD. In the 3D reconstruction video, individual dermal papilla and obvious fibrosis in the upper papillary dermis were easily identified. Our estimation models could predict the disease severity of AD patients with an accuracy of nearly 85%. The THG signal intensity and characteristic GLCM patterns are associated with AD severity and can serve as quantitative predictive parameters. Our imaging approach can be used to identify the histopathological changes of AD objectively, and to complement the SCORAD index, thus improving the accuracy of classifying AD severity. The prevalence of atopic dermatitis (AD) is increasing among pediatric population worldwide. Noninvasive imaging approaches supplementary to SCORAD for histopathological evaluation is required. We demonstrated here harmonic generation imaging can serve as quantitative method for predicting AD severity objectively. Our in vivo imaging approach is suitable for children and can be used to classify AD and can serve to be complementary to SCORAD for clinical purpose.