Impedantie metingen voor de in situ en in vivo detectie en characterisatie van biofilms

Given the rise of implant-associated infections in healthcare and persistent biofouling in industry, the significant economic and health burden they cause is only expected to increase. These contaminations are often caused by bacterial biofilms, which are surface-associated microbial communities encased in a protective matrix. They are hard to eradicate and even harder to diagnose quickly, therefore early and reliable biofilm detection remains a major unmet need. Current reference methods are slow, require invasive sampling, are label-dependent, and not suited for continuous, in-situ monitoring. This gap delays treatment, increases costs, and worsens outcomes. This PhD addresses that problem by implementing electrochemical impedance spectroscopy (EIS) as a rapid, label-free, non-destructive method for biofilm detection and identification. First, we mapped how experimental choices (electrodes, configurations, analyses) shape impedimetric responses, clarifying links between biofilm traits and spectra. Using commercial microelectrode arrays (MEAs), we reveal a robust signature of biofilm growth: a low-frequency |Z| decrease paired with a high-frequency increase. Equivalent-electrical-circuit (EEC) modelling connects these spectral shifts to specific electrical components, with non-ideal interfacial capacitance emerging as most informative. Decision tree ensemble models including these EEC-derived features improved the species- and strain-level classification accuracy by 2-11\% over raw spectra alone, reaching accuracies up to 96 %. To enable miniaturised, sensitive sensing, we introduce sputtered iridium oxide film (SIROF) electrodes. Cyclic-voltammetry pre-conditioning lowered interfacial impedance and enabled detection within 24 h. This allowed the SIROF sensors to discriminate between Staphylococcus aureus and Pseudomonas aeruginosa biofilms. Here the electrode size tunes its sensitivity and performance. For translation of this technology into a clinical environment, we integrated these SIROF electrodes with titanium fracture fixation plates. In an in vivo rabbit fracture-related infection model, the device detects S. aureus within 24 h—faster than reported clinical methods. This data in combination with a simple predictive model reaches up to 88% accuracy, underscoring the value of data-driven EIS. Overall, this work moves EIS from an in vitro concept to implant-compatible in vivo sensing, providing a foundation for earlier intervention and improved outcomes in biofilm-associated infections. Further optimisation of this technology will support standardisation and clinical deployment.