Quantitative monitoring ofloxacin in beef by TLC-SERS combined with machine learning analysis

Abstract

Ofloxacin is one kind of quinolone antibiotic drugs, the abuse of ofloxacin in livestock and aquaculture may bring bacterial resistance and healthy problem of people. The illegally feeding cattle with ofloxacin will help it keep health, but the sedimentation of ofloxacin could bring problem in food safety. The accurate, simple and instant monitoring ofloxacin from beef by portable sensor was of vital issue in food quality. A simple and reliable method was proposed for instant and quantitative detecting ofloxacin in beef, in which the thin-layer chromatography (TLC) -surface-enhanced Raman scattering (SERS) spectroscopy was in tandem with machine learning analysis base one principal component analysis-back propagation neural network (PCA-BPNN). The TLC plate was composed with diatomite, that was function as the stationary phase to separate ofloxacin from beef. The real beef juice was directly casted onto the diatomite plate for separating and detecting. The directly monitor ofloxacin from beef was achieved and the sensitivity down to 0.01 ppm. The PCA-BPNN was used as reliable model for quantitative predict the concentration of ofloxacin, that shown superior accuracy compared with the traditional model. The results verify that the diatomite plate TLC-SERS combined with machine-learning analysis is an effective, simple and accurate technique for detecting and quantifying antibiotic drug in meat stuff to improve the food safety.

Keywords: Artificial intelligence; Ofloxacin; PCA-BPNN; Quantitative; TLC-SERS.