Use of information visualization to enhance the sensitivity and selectivity of biosensors

 

Osvaldo N. Oliveira Jr.

Instituto de Física de São Carlos, USP, Brazil

chu@ifsc.usp.br

 

Sensing and biosensing are increasingly relying on detection methods that generate large amounts of data, as in the case of electrochemical and impedance spectroscopy techniques. Most of the sensors and biosensors, however, normally employ only a minute part of these data, but this scenario may change rapidly in the next few years with the realization that information visualization methods may assist in distinguishing very similar samples. In this lecture, an overview will be presented of multidimensional projections to enhance the sensitivity of sensors and biosensors fabricated with nanostructured organic films, namely Langmuir-Blodgett (LB) and layer-by-layer (LbL) films. With the multidimensional projections, data points taken with impedance or electrochemical measurements are visualized in a 2D plot, where the similarity among samples is depicted in terms of distances calculated with several methods. Examples will be given of biosensors based on field-effect devices in which crosstalk effects could be eliminated using information visualization, in addition to highly sensitive biosensors employing impedance measurements that allow for detection of tropical diseases. A roadmap will be discussed with possible ways to employ information visualization for optimizing biosensors and clinical diagnosis systems.