Hyperspectral imaging aided by artificial neural networks for functional skin characterization
 (Conference Presentation)

Alexander V. Bykov; Evgeny Zherebtsov; Mikhail Kirillin; Daria Loginova; Alexey Popov; Alexander Doronin; Igor Meglinski

doi:10.1117/12.2321148

18 September 2018 Hyperspectral imaging aided by artificial neural networks for functional skin characterization (Conference Presentation)

Alexander V. Bykov, Evgeny Zherebtsov, Mikhail Kirillin, Daria Loginova, Alexey Popov, Alexander Doronin, Igor Meglinski

Author Affiliations +

Proceedings Volume 10768, Imaging Spectrometry XXII: Applications, Sensors, and Processing; 107680C (2018) https://doi.org/10.1117/12.2321148
Event: SPIE Optical Engineering + Applications, 2018, San Diego, California, United States

Abstract

We introduce a portable hand-held hyperspectral imaging system for the functional diagnostics of skin and vascular system. Hyperspectral image analysis aided by artificial neural networks (ANN) allows to reconstruct major physiological parameters of human skin nearly in real-time. The developed device provides spatial distribution of blood volume fraction, oxygenation and melanin content within skin. Special attention has been paid on the system validation and calibration using specially developed skin mimicking phantoms with confirmed optical properties. The device was built on the basis of unique hyperspectral snapshot camera utilizing a micro Fabry-Perot filter providing real spectral response in each pixel (no interpolation is used in image formation). A broadband illumination unit combined with the camera is based on the fiber-optic illuminator providing uniform distribution of light intensity and utilizes halogen lamp. The specially developed ANN algorithm was used to perform the inverse problem solution for quantitative assessment of major parameters of skin based on the measured hyperspectral images. A set of diffuse reflectance spectra of human skin imitated by the Monte Carlo method developed in-house has been used extensively for the training of ANN. The volume fraction of blood, oxygen saturation, melanin content and thickness of the epidermal layer were used variable parameters in the utilized seven-layer Monte Carlo-based skin model. The total training set contained 45,198 spectra in the range of 505–800 nm simulated with a step of 5 nm. The developed imaging system has been successfully used to perform the occlusion test measurements with healthy volunteers.

Conference Presentation

Citation Download Citation

Alexander V. Bykov, Evgeny Zherebtsov, Mikhail Kirillin, Daria Loginova, Alexey Popov, Alexander Doronin, and Igor Meglinski "Hyperspectral imaging aided by artificial neural networks for functional skin characterization (Conference Presentation)", Proc. SPIE 10768, Imaging Spectrometry XXII: Applications, Sensors, and Processing, 107680C (18 September 2018); https://doi.org/10.1117/12.2321148

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
PRESENTATION

WATCH
PRESENTATION SAVE TO MY LIBRARY

GET CITATION

KEYWORDS

Skin

Hyperspectral imaging

Artificial neural networks

Imaging systems

Monte Carlo methods

Algorithm development

Blood

Show All Keywords

Keywords/Phrases

Search In:

Publication Years