Intelligent Electrochemical Sensing: Machine Learning-Powered Multidimensional Fingerprinting for Simultaneous...

Antibiotic residues in environmental, food, and biological systems have become a serious global public health concern. Signal overlap occurs among structurally similar antibiotics, posing a significant challenge to the accurate identification and quantification of multicomponent antibiotics in complex systems. Here, we propose an intelligent electrochemical strategy that integrates a single-electrode four-channel platform with machine learning for the simultaneous detection of six antibiotics (amoxicillin, chloramphenicol, ciprofloxacin, enrofloxacin, norfloxacin, and ofloxacin). By modulating the scan direction and solution pH, multidimensional fingerprint signals were obtained via square wave voltammetry (SWV) at a glassy carbon electrode modified with poly(p-aminobenzenesulfonic acid), HKUST-1, and Au-Pt nanoparticles. Machine learning algorithms were then applied to decode these complex signals and extract discriminative features. The multilayer perceptron (MLP) achieved a classification accuracy of 99.75% for the six antibiotics in PBS, while the convolutional neural network (CNN) showed excellent regression performance with R² > 0.9990 for concentration prediction. The trained CNN model was further validated in complex real samples, including milk, environmental water, and serum, maintaining R² > 0.9850. Even in serum, the prediction accuracy for six antibiotics remained above 0.9964, with detection limits reaching the picomolar level. In addition, the electrocatalytic mechanisms of the antibiotics were investigated through electrochemical analysis, combined with conceptual density functional theory (CDFT) calculations. Overall, this work establishes a new paradigm for intelligent electrochemical sensing through the integration of electrode engineering, signal acquisition, and artificial intelligence, providing a promising strategy for rapid and accurate detection of multiple trace pollutants in complex matrices.