|
A new method to transfer the standard addition procedure for concentration determination to immunoassays with non-linear calibration curves was developed. The new method was successfully applied to simulated data and benchmarked against a state-of-the-art algorithm, showing a significantly improved performance with improvement factors between 2 and 192. The logit function was used to transform the immunoassay signal response of test samples spiked with known analyte concentrations. The relationship between logit(signal) and log-transformed estimated total analyte concentration is linear if the estimated total analyte concentration is correct. Finally, the new method was validated experimentally using different assays in varying, relevant complex matrices, such as serum, saliva, and milk. Different concentrations of testosterone and amitriptyline between 0.05 and 3.0 µg L−1 were quantified using a binding inhibition assay in combination with reflectometric interference spectroscopy (RIfS) as the transduction principle. The sample concentration was calculated using a numerical method. Samples could be quantified with recoveries between 70 and 118%. The standard addition method accounts for individual matrix interference on the immunoassay by spiking the test sample itself. Although the experiments were carried out using RIfS, the method can be applied to any immunoassay that meets the analytical requirements.
|
