Application of artificial neural networks and biosensors to determine concentrations of mixture

Abstract

Biosensor response, in case of multi-substrate mixture, has nonlinear dependence on substrate concentrations. This work investigates the possibility to approximate this dependency with artificial neural network. Also the influence of external diffusion layer to results of multi-substrate determination was investigated. The numerically modelled biosensor response was used as experimental data. The principal components analysis was used to reduce the dimension of biosensor response. Prefered method gives acceptable acuratnes on multi-substrate determination and it can be improved by relatively large external diffusion layer.

Applying artificial neural networks to solve the inverse problem of evaluating concentrations in multianalyte mixtures from biosensor signals

“…The dimensionality reduction method we use for feedforward neural networks is the principal component analysis (PCA), which transforms the raw data into principal components (PCs) -they contain the most informative features of the dataset. This method has been used in previous papers (see [4,20]). In our research, we found that performing data normalization before PCA yielded better results.Let X be an n × p response matrix, where n is the amount of biosensor responses, and p is the amount of time samples per response.…”

“…But for three analytes, a small noise level can lead to large errors, and local optimization is still time consuming, when we have a lot of responses to analyze. Litvinas and Baronas [20] used artificial neural networks to analyze biosensor responses for four analytes. The authors only used a simple feedforward network architecture, which had a small relative prediction error, but effects of noise and alternative network architectures were untested.…”