Medical research and technological advancements are heading towards tailored healthcare approaches that prioritize individual needs, allowing for more accurate diagnoses, more effective treatments, and better patient outcomes overall. One such approach is the use of digital twins, which involves creating a unique computational and technological representation of an individual's medical copy based on data collected from multiple edge AI technologies. Moving towards preventive and personalized healthcare requires real-time monitoring of key biomarkers directly on-body. There is still a need for robust and low-power wearable sensors that can continuously monitor multiple biomarkers in human biofluids. These sensors would have the potential to serve two primary purposes: firstly, early detection of change in concentration of biomarkers, enabling doctors to make faster treatment decisions. Secondly, the large amount of data collected from different patients in hospitals can be used to develop personalized healthcare solutions that improve patient outcomes. FET-based biosensors have emerged as promising candidates for the next generation of low-power and label-free biosensors due to their ease of miniaturization, low-power consumption, and the possibility of integration into CMOS technology. This project aims to lay the groundwork for the creation of robust and low-power multimodal biomarker ISFET-type wearable sensors that can quasi-continuously monitor multiple biomarkers in human interstitial fluid. In this work, fully-depleted silicon nanowire array Field-Effect Transistors on SOI technology have been proposed as biosensors to detect biomarkers in human biofluids, such as interstitial fluid. Two different platforms have been fabricated, one entirely at the Centre of MicroNanotechnology (CMi) of EPFL, and one at CEA-LETI in Grenoble. The fabricated SiNW FETs have demonstrated excellent performance in terms of electrical characterization, with an OFF to ON ratio larger than 6 orders of magnitude and a subthreshold swing of 80mV/decade, two key requirements to obtain an efficient biosensor. A dependable method of surface functionalization for detecting C-Reactive Protein (CRP) has been developed. This method employs antibody fragments on high-k oxides, and it has been extensively validated using various characterization techniques such as Surface Plasmon Resonance (SPR), X-Ray spectroscopy (XPS), and Quartz Crystal Microbalance with Dissipation Monitoring (QCM-D). Efficient and reliable monitoring of C-Reactive Protein in dual-gate configuration and preliminary results in constant current mode have been demonstrated. Moreover, we have proven a reliable and reproducible detection of pH levels both in buffer and human ISF-like solutions. The intrinsic double gate structure of the system has been exploited to obtain signal amplification (from Nernstian sensitivity of 59 mV/pH up to 3 V/pH), and a constant current configuration has been introduced to obtain a system with voltage output and the possibility to have fine-tuning of the output dose response through the selection of the system operating point. The results obtained in this thesis move toward real-life applications of wearable sensing, thus considering stability, reliability, and miniaturization. The proposed biosensors offer a promising solution for the detection of biomarkers in human biofluids, enabling real-time health evolution monitoring, and preventive and personalized care.