Researchers from the Electronics Research Institute and Badr University present a FR4-based dual-band microwave bandpass filter sensor employing split-ring resonators for noninvasive blood glucose measurement. By tracking S-parameter shifts at 2.45 and 5.2 GHz and applying CatBoost and Random Forest models, the system correlates dielectric changes in tissue with glucose concentrations, offering a compact, low-cost alternative to invasive glucose monitoring.
Key points
FR4-based dual-band bandpass filter sensor with concentric split-ring resonators tuned at 2.45 GHz and 5.2 GHz for glucose sensing.
S-parameter (S11 and S21) shifts in resonant frequency, magnitude, and phase track glucose-dependent permittivity changes.
Integration with nanoVNA measurements and Random Forest/CatBoost classifiers achieves sensitivity up to 2.026 MHz/(mg/dL) and 0.011 dB/(mg/dL).
Why it matters:
This dual-band microwave sensor with AI analysis could revolutionize diabetes care by offering highly sensitive, noninvasive glucose monitoring without needles.
Q&A
How do split-ring resonators detect glucose?
What role does machine learning play?
How does the finger phantom model work?
Is microwave exposure safe for monitoring?
Read full article
Academy
Microwave Biosensors for Noninvasive Glucose Monitoring
Introduction
Noninvasive monitoring of blood glucose levels offers a comfortable and potentially continuous alternative to traditional finger-prick tests. Microwave biosensors leverage electromagnetic waves in the gigahertz range to probe the dielectric properties—permittivity and conductivity—of tissues or fluids. These dielectric properties change predictably with glucose concentration, enabling quantification without piercing the skin.
Principle of Operation
Split-Ring Resonators (SRRs)
An SRR consists of concentric metallic rings with narrow gaps that resonate at specific microwave frequencies when excited by an incident electromagnetic field. The resonant frequency and quality factor depend on the surrounding dielectric environment. When an SRR-based sensor is placed against the skin or a fluid sample, changes in local permittivity caused by glucose levels shift these resonances, which are detected as frequency, amplitude, or phase variations in the S-parameters (Scattering parameters).
Key advantages of SRRs include compact size, high quality factor for sharp resonance peaks, and strong localized fields around the gaps, which enhance sensitivity to small permittivity changes.
Dielectric Properties and Glucose
The dielectric constant (ε’) indicates how much electric field energy a material can store, while the loss tangent (tan δ) reflects energy dissipation. Glucose in water alters both ε’ and tan δ in proportion to concentration. Using models like Cole–Cole, one can compute frequency-dependent dielectric properties for tissue-mimicking phantoms, ensuring realistic testing with multilayer constructs representing skin, fat, muscle, blood, and bone.
Sensor Design and Measurement
- Filter Structure: The sensor integrates a dual-band microstrip bandpass filter built on a low-cost FR4 substrate, embedding three SRR cells: two concentric rings for the ISM band (2.45 GHz) and a smaller ring for the WLAN band (5.2 GHz).
- Evaluation Method: A vector network analyzer or nanoVNA measures S11 (reflection) and S21 (transmission) parameters. Frequency shifts, magnitude changes, and phase shifts are recorded as sample concentration or blood glucose levels vary.
- Phantom and Human Testing: Initial validation uses a polypropylene container with aqueous glucose solutions. Further tests employ a finger phantom model and real human fingertips, assessing safety (SAR) and repeatability.
Data Analysis with Machine Learning
Multivariate S-parameter data can be complex and influenced by noise or individual variability. Machine learning classifiers such as Random Forest and CatBoost are trained on labeled datasets linking features (frequency shift, magnitude, phase) to known glucose concentrations. These models improve prediction accuracy, enabling robust noninvasive glucose estimation even with overlapping signal patterns.
Applications in Longevity Science
Continuous, painless glucose monitoring supports better diabetes management, reducing complications and improving quality of life. For longevity enthusiasts, maintaining stable glucose levels can mitigate age-related metabolic disorders and enhance overall healthspan.
Future Directions
- Miniaturization and integration into wearable devices.
- Real-time data transmission via Bluetooth or IoT platforms.
- Extended validation across diverse populations and clinical trials.
By combining microwave sensor physics, dielectric modeling, and AI-driven data analysis, this field advances toward practical, noninvasive health monitoring technologies accessible for everyday use.
