Signal Processing Algorithm for Pre-processing the Surface Plasmon Resonance Signal Response

Abstract

Surface plasmon resonance (SPR) is a versatile optical bio-sensing technique which has an ability to detect antibody-antigen molecular binding. In this work, we present a data processing algorithm that can process and analyze the data output from SPR equipment. The SPR data output is typically a non-periodic square wave, an indicator that a biological substance is captured, with continuous noises. To remove the outliers and smoothen the data, moving average and Savitzky-Golay Filter were employed. Then, a change point detection method and polynomial regression were used to isolate the buffer data as baseline and give baseline prediction which is later used to calculate and quantify the response. From this study, the algorithm is expected to give an accurate baseline prediction and response calculation. Based on the results, the algorithm was able to detect the SPR signal response (change point detection) with an error below 15%. Thus, this algorithm would enable the researcher to analyze and interpret the SPR data much faster and simpler.