Hardware and Software Codesign for RF Spectra Biosensors

In the sensor field, one of the main error factors in signal detection lies in the interface between the device and the object from which the signal is to be recorded. Therefore, there is a growing trend to reduce this contact through the adoption of various solutions. These include improving the adhesion between the contact surfaces, especially in the case of electrodes, modifying the interface itself or the device, or a complete change of approach by exploiting different physical principles. One of these approaches consists of studying electromagnetic fields and their interaction with the materials with which they come into contact, including those in the biological environment. It has been observed that this interaction varies depending on several variables, but the main one is the amount of water contained within the biological environment, which is highly susceptible to the presence of magnetic fields, as it is a polar molecule. Based on this, it is possible to develop a device capable of measuring the level of hydration in body tissue and, consequently, in the patient's body. This device could first acquire radiofrequency spectra and then use machine learning techniques to compare the variations of these spectra. In this way, it would be possible to quickly and accurately obtain an immediate comparison between spectra obtained under ideal hydration conditions and those obtained under dehydration conditions, previously saved.