Antimicrobial resistance (AMR) represents a major global health challenge and motivates the development of rapid analytical approaches for characterizing clinical bacterial isolates. Surface-enhanced Raman spectroscopy (SERS), combined with machine learning (ML), has emerged as a promising approach for label-free spectral classification of bacterial isolates. However, the field lacks methodological consensus on how hierarchical spectral datasets should be structured, standardized, and partitioned for reliable model evaluation. This gap has led to widespread data leakage and inflated performance reports in SERS–AI studies. Here, we present a systematic benchmarking framework based on 15,000 SERS spectra acquired from 15 clinical Staphylococcus aureus isolates representing distinct resistance phenotypes (MRSA, ERSA, and SSA), evaluating twelve data-partitioning strategies across five machine learning models and three feature selection or dimensionality-reduction methods. Our results show that spectrum-level splits consistently overestimate classification accuracy due to non-independent samples, whereas isolate- or subject-level partitioning provides more realistic estimates of model generalizability. Among the evaluated models, random forest combined with Boruta feature selection produced consistently robust and interpretable performance. Overall, model performance depended more strongly on dataset organization and partitioning strategy than on the choice of machine learning algorithm.
Hydrogen peroxide (H2O2) is a relatively stable reactive oxygen species and an important biomarker of oxidative stress in human semen. Elevated H2O2 concentrations have been associated with impaired sperm function and male infertility. Therefore, accurate determination of total H2O2 levels in semen is valuable for assessing semen oxidative status. In this work, a cost-effective and sensitive electrochemical sensing platform was developed for the determination of total H2O2 levels in human semen. A lab-made flexible porous graphene electrode was modified with a three-dimensional poly(3,4-ethylenedioxythiophene) film embedded with Prussian blue analog (3D PBA@PEDOT/F-PGE). The 3D PBA@PEDOT/F-PGE transducer was fabricated via a one-step electrodeposition process, providing superior conductivity, low charge-transfer resistance, excellent structural integrity, and stable electrocatalytic activity toward H2O2 reduction. The transducer exhibited a linear detection range of 0.50 μM to 6.0 mM, a sensitivity of 390 μA mM-1 cm-1, and a detection limit of 0.18 μM. The transducer was further integrated with a paper-based microfluidic device and a battery-free near-field communication (NFC) potentiostat to construct a portable electrochemical sensing platform capable of determining H2O2 over a concentration range of 1.0 μM to 4.0 mM with good precision and long-term stability. The smartphone-controlled sensing platform was successfully applied to the determination of total H2O2 concentrations in human semen samples, providing a convenient approach for assessing semen oxidative status. This platform shows considerable potential for point-of-care monitoring of oxidative stress associated with male reproductive health and for future applications in portable electrochemical sensing technologies.
