A calibration-base method for the evaluation of the detection limit of an electrochemical biosensor

The method based on the confidence limits of a regression straight line is applied to the linear portion of the calibration curve of a biosensor in order to determine the limit of detection. A practical application gives a detection limit of an ethanol biosensor greater than that calculated by using the method of the signal-to-noise-ratio, S/N= 3, but the value is more reliable because it accounts for the variability in the response at analyte additions. A simple approximation of the method is also presented.