Using an artificial neural network to determine electrical properties of epithelia

Abstract

The present study introduces a new approach for modeling electrical properties of epithelia. Artificial neural networks (ANNs) are used to estimate key parameters that otherwise can only be measured directly by applying complex and time-consuming laboratory methods. Assuming an electrical model equivalent to an epithelial layer, an ANN can be trained to learn the relation between these parameters and experimentally obtained impedance spectra. We demonstrate that even with a naive ANN our approach reduces the error rate of parameter estimation to less than 20 per cent. Successful test runs provide a proof of concept. © 2010 Springer-Verlag Berlin Heidelberg.