Fluorescent immune-chromatographic strip signal processing and feature selection

Yueming Gao,Tianqi Li,Chuanyang Lin,Shaoheng Pan,Mengyu Wei,Min Du

DOI: https://doi.org/10.13382/j.jemi.2015.05.005

2015-01-01

Abstract:Immunofluorescence chromatographic quantitative detection technology is an emerging technique of the point of care test (POCT).Based on the previous work of immunofluorescence quantitative chromatography detector, the wavelet algorithm was used to denoise the fluorescence signal after the A/D sampling in this paper.Then the baseline drift was removed by the median filter method.Finally, the peak height and the area were sellected as the characteristic value for the quantitative detection.The experimental data shown that the SNR improved 10dB and the SRMS reduced by 30.4%after denoising.In the quantitatively detect step, the CV were 5.1% and 0.4%, and the test error were 9.1%and 0.9%, for the peak hight and area characteristic value, respectively.The results show that the peak area of the test line as the characteristic value has better accuracy and repeatability than the peak height .