A wearable multi-parameter electrochemical detection and signal processing system for real-time sweat analysis.

This study presents a wearable multi-parameter electrochemical detection system (WMEDS) designed for the simultaneous monitoring of sweat biomarkers. The system integrates a microcontroller, a dual-mode signal acquisition circuit, and a low-power Bluetooth module. To address the non-stationary electrochemical signals, a preprocessing approach combining Savitzky-Golay filtering with Z-score-difference detection is employed. The core algorithm incorporates wavelet hierarchical threshold denoising, which improves the average signal-to-noise ratio by 12 dB. This signal processing method effectively suppresses noise while preserving key signal features (peak retention rate of 97.3%), resulting in a smoothness index exceeding 0.99, and the root mean square remains below ± 0.5%. The efficacy of the WMEDS is demonstrated in Ca and glucose detection, indicating comparable results with those obtained from commercial electrochemical workstations. Furthermore, the WMEDS achieves highly accurate detection of Ca (sensitivity of 30.6 mV/decade) and glucose (sensitivity of 7.7 μA/mM), with quantitative errors less than 5.5% and on-body test fluctuations below 2.2%, thereby fulfilling the performance requirements for wearable sweat monitoring applications.