A Surface-Enhanced Raman Spectroscopy Platform Integrating Dual Signal Enhancement and Machine Learning for Rapid Detection of Veterinary Drug Residues in Meat Products

The detection and quantification of veterinary drug residues in meat remain a significant challenge due to the complex background interference inherent to the meat matrix, which compromises the stability and accuracy of spectroscopic analysis. This study introduces an advanced label-free surface-enhanced Raman spectroscopy (SERS) platform for the precise identification and quantification of veterinary drugs. By employing a dual enhancement strategy involving sodium borohydride activation and calcium ion-deuterium oxide guidance, this platform achieves the efficient capture and signal amplification of drug molecules on highly active nanoparticles. High-quality SERS spectra were obtained for carprofen, doxycycline hydrochloride, chloramphenicol, and penicillin G sodium salt, enabling accurate classification and interference suppression. In addition, the application of machine learning algorithms, including PCA-LDA, heatmap, and decision tree modeling, allows for accurate differentiation of mixed drug samples. Quantitative analyses in meat samples were achieved through Raman intensity ratios and multivariate curve resolution-alternate least-squares (MCR-ALS) analysis, with results showing high consistency with high-performance liquid chromatography (HPLC) measurements. These findings highlight the potential of this SERS-based platform for accurate and rapid detection of multicomponent veterinary drug residues in complex food matrices.