A biosensor model based on multivariate regression of corrected Cole-Cole parameters

Rapid-response biosensing systems are necessary to counteract threats due to high-consequence pathogens. A impedance immunosensor for E. coli O157∶H7 detection was developed based on multi-frequency electrical impedance scanning, screen-printed microelectrodes, and lateral-flow chromatographic assay. In order to perform more meaningful data analysis, in addition to use the correction method and extract the Cole-Cole parameters (G
0
, G
∞
, f
c
, α), a multivariate regression model for bacteria concentration versus these parameters was been explored. The results showed that G
0
was most sensitive to the cell concentrations in the range of 10
²
∼10
⁷
CFU/ml, which yielded the model determination coefficient (R
²
) to 0.8761. What's more, with more valid variable entered, R
²
increased, and was up to 0.9751 for the best model. The limit of detection for the immunosensor was able to reach 39CFU/ml and the detection time was less than five minutes.