Salivary Impedance-Based Glucose Sensing. A Hardware-Assisted Approach and Performance Evaluation

Abstract

Non-invasive glucose monitoring faces significant reliability hurdles due to electrode polarization and low-amplitude signal interference in salivary sensing. This paper presents an integrated mixed-signal platform using the ADuCM350 system-on-chip to evaluate the feasibility of salivary bio-impedance analysis for glucose estimation. The methodology implements an isolated 4-wire Kelvin technique and hardware-accelerated DFT processing to mitigate contact impedance and systematic errors.

While the system demonstrated high accuracy in characterizing Randles equivalent circuits, preliminary measurements with hydrogen peroxide (𝑯𝟐𝑶𝟐) solutions revealed significant challenges in electrochemical stability and signal-to-noise ratios. These results provide critical insights into the limitations of current portable architectures and highlight the need for further optimization in electrode-analyte interface stability for reliable non-invasive diagnostics.