Simultaneous detection and quantification of ciprofloxacin, doxycycline, and levofloxacin in municipal lake water via deep learning analysis of complex Raman spectra

In recent years, the misuse of antibiotics has led to severe pollution in water environments, with excessive residues in lake water damaging ecosystems and promoting the emergence of antibiotic-resistant bacteria. Therefore, rapid detection of antibiotic residues in the environment is crucial. This study introduces a novel method for the simultaneous quantification of mixed antibiotics in lake water using Surface-Enhanced Raman Scattering (SERS) combined with deep learning methods. To demonstrate the accuracy of our experiments, we tested four lake water samples collected from four distinct sampling points of an artificial lake in a municipal city in China. We independently analyzed each sample mixed with commonly used antibiotics, including ciprofloxacin, doxycycline, and levofloxacin. A non-negative elastic network was then employed to predict concentration ratios of mixed antibiotics in the lake water samples. The results showed that the established method can accurately quantify the ratios of individual antibiotics in mixed solutions at all four lake water sampling points. This approach facilitates the identification and quantification of antibiotics in lake water with simplicity and rapidity, exhibiting potential application for real-world monitoring of fluctuations of antibiotic residues in natural water systems.