From Radio Waves to Wellness: Radar and Machine Learning for Noncontact Dehydration Assessment
IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and Biology · 2026-09 · 10 (3) · 397-409
DOI: 10.1109/jerm.2026.3686633 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and Biology, 2026-09. Verfasst von Jose Maria Sosa und 5 weiteren Autor:innen. Die Arbeit nennt 48 Literaturangaben. Gefördert von: King Abdullah University of Science and Technology, KAUST Center of Excellence for Smart Health.
Abstract (Originalsprache)
This paper introduces a novel, noninvasive, and contactless method for dehydration measurement using a frequency-modulated continuous wave (FMCW) radar operating at center frequencies of 10 GHz and 77 GHz. The proposed framework begins with a customized preprocessing technique designed to isolate the signal corresponding to the target of interest (ToI) from the received radar data, ensuring that the extracted features reliably reflect physiological indicators of hydration status. To capture the complex, nonlinear patterns associated with dehydration, a wavelet scattering transform (WST)-based feature extraction method was employed, which provides robust input for machine learning models. These features are then used to train and evaluate various classifiers, the best one being neural networks (NN), which accurately classify protocol sessions. A key challenge in nonclinical dehydration assessment is the absence of biochemical ground truth. To overcome the challenge, we developed and annotated a novel dataset—the Fasting Dataset—collected under controlled conditions from individuals who observed fasting. This dataset captures radar measurements at different physiological stages, offering a valuable reference for hydration-level analysis. Validation using this custom dataset demonstrates the initial feasibility of the proposed methodology for real-time, noninvasive, and contactless dehydration monitoring in remote healthcare settings.
Förderung
- King Abdullah University of Science and Technology
- KAUST Center of Excellence for Smart Health
Zitieren
APA
Sosa, J. M., Kotte, V. V., Ahmed, S., Rahman, M. M. U., Alouini, M., & Al-Naffouri, T. Y. (2026). From Radio Waves to Wellness: Radar and Machine Learning for Noncontact Dehydration Assessment. IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and Biology, 10(3), 397-409. https://doi.org/10.1109/jerm.2026.3686633
Vancouver
Sosa JM, Kotte VV, Ahmed S, Rahman MMU, Alouini M, Al-Naffouri TY. From Radio Waves to Wellness: Radar and Machine Learning for Noncontact Dehydration Assessment. IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and Biology. 2026;10(3):397-409. doi:10.1109/jerm.2026.3686633
BibTeX
@article{Sosa2026,
author = {Sosa, Jose Maria and Kotte, Vijith Varma and Ahmed, Sajid and Rahman, Muhammad Mahboob Ur and Alouini, Mohamed-Slim and Al-Naffouri, Tareq Y.},
title = {From Radio Waves to Wellness: Radar and Machine Learning for Noncontact Dehydration Assessment},
journal = {IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and Biology},
year = {2026},
volume = {10},
number = {3},
pages = {397--409},
doi = {10.1109/jerm.2026.3686633}
}
EUVIMED