Non-invasive healthcare technologies are an important part of research and development nowadays due to the low cost and convenience offered to both healthcare receivers and providers. This work overviews the recent advances in the field of non-invasive electrochemical biosensors operating in secreted human physiological fluids, viz. tears, sweat, saliva, and urine. Described electrochemical devices are based on different electrochemical techniques, viz. amperometry, coulometry, cyclic voltammetry, and impedance spectroscopy. Challenges that confront researchers in this exciting area and key requirements for biodevices are discussed. It is concluded that the field of non-invasive sensing of biomarkers in bodily fluid is highly convoluted. Nonetheless, if the drawbacks are appropriately addressed, and the pitfalls are adroitly circumvented, the approach will most certainly disrupt current clinical and self-monitoring practices.

Bacterial infections can affect the skin, lungs, blood, and brain, and are among the leading causes of mortality globally. Early infection detection is critical in diagnosis and treatment but is a time- and work-consuming process taking several days, creating a hitherto unmet need to develop simple, rapid, and accurate methods for bacterial detection at the point of care. The most frequent type of bacterial infection is infection of the urinary tract. Here, we present a wireless-enabled, portable, potentiometric sensor for E. coli. E. coli was chosen as a model bacterium since it is the most common cause of urinary tract infections. The sensing principle is based on reduction of Prussian blue by the metabolic activity of the bacteria, detected by monitoring the potential of the sensor, transferring the sensor signal via Bluetooth, and recording the output on a laptop or a mobile phone. In sensing of bacteria in an artificial urine medium, E. coli was detected in ~4 h (237 ± 19 min; n = 4) and in less than 0.5 h (21 ± 7 min, n = 3) using initial E. coli concentrations of ~10³ and 10⁵ cells mL⁻¹, respectively, which is under or on the limit for classification of a urinary tract infection. Detection of E. coli was also demonstrated in authentic urine samples with bacteria concentration as low as 10⁴ cells mL⁻¹, with a similar response recorded between urine samples collected from different volunteers as well as from morning and afternoon urine samples.

An electronic tongue is a powerful analytical instrument based on an array of non-selective chemical sensors with a partial specificity for data gathering and advanced pattern recognition methods for data analysis. Connecting electronic tongues with electrochemical techniques for data collection has led to various applications, mostly within sensing for food quality and environmental monitoring, but also in biomedical research for the analyses of different bioanalytes in human physiological fluids. In this paper, an electronic tongue consisting of six electrodes (viz., gold, platinum, palladium, titanium, iridium, and glassy carbon) was designed and tested in authentic (undiluted, unpretreated) human saliva samples from eight volunteers, collected before and during the COVID-19 pandemic. Investigations of 11 samples using differential pulse voltammetry and a principal component analysis allowed us to distinguish between SARS-CoV-2-free and infected authentic human saliva. This work, as a proof-of-principle demonstration, provides a new perspective for the use of electronic tongues in the field of enzyme-free electrochemical biosensing, highlighting their potential for future applications in non-invasive biomedical analyses.

The prevailing form of bacterial infection is within the urinary tract, encompassing a wide array of bacteria that harness the urinary metabolome for their growth. Through their metabolic actions, the chemical composition of the growth medium undergoes modifications as the bacteria metabolize urine compounds, leading to the subsequent release of metabolites. These changes can indirectly indicate the existence and proliferation of bacterial organisms. Here, we investigate the use of an electronic tongue, a powerful analytical instrument based on a combination of non-selective chemical sensors with a partial specificity for data gathering combined with principal component analysis, to distinguish between infected and non-infected artificial urine samples. Three prevalent bacteria found in urinary tract infections were investigated, Escherichia coli, Klebsiella pneumoniae, and Enterococcus faecalis. Furthermore, the electronic tongue analysis was supplemented with ¹H NMR spectroscopy and flow cytometry. Bacteria-specific changes in compound consumption allowed for a qualitative differentiation between artificial urine medium and bacterial growth.