Multivariate regression model of impedance of normal and chemically irritated skin shows predictive ability

Author

Åberg, P. ; Nicander, I. ; Geladi, P. ; Ollmar, S.

Author_Institution

Med. Eng., Karolinska Institutet, Huddinge, Sweden

Volume

3

fYear

2001

fDate

2001

Firstpage

2208

Abstract

We present predictive models that can foresee how skin will react when exposed to chemicals. Skin impedance spectra, 31 frequencies between 1 and 1000 kHz at five depth settings, were collected before and after application of chemicals on volar forearms of volunteers. Tegobetaine and sodium lauryl sulphate were used to induce the irritations. The spectra were filtered using orthogonal signal correction (OSC). The relation between skin impedance of normal and chemically irritated skin was modelled using partial least squares regression (PLS). The predictive ability of this model is demonstrated for two irritants, and additional studies are required to establish this property for other chemicals.

Keywords

bioelectric phenomena; electric impedance measurement; least mean squares methods; physiological models; prediction theory; skin; statistical analysis; 1 to 1000 kHz; bio-impedance detection; chemically irritated skin; five depth settings; impedance; irritants; multivariate regression model; normal skin; orthogonal signal correction; partial least squares regression; predictive ability; skin impedance spectra; sodium lauryl sulphate; tegobetaine; volar forearms; volunteers; Chemical engineering; Data mining; Frequency; Laser sintering; Least squares methods; Multivariate regression; Predictive models; Principal component analysis; Skin; Surface impedance;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 2001. Proceedings of the 23rd Annual International Conference of the IEEE

ISSN

1094-687X

Print_ISBN

0-7803-7211-5

Type

conf

DOI

10.1109/IEMBS.2001.1017210

Filename

1017210