Improving Multi-Sensor Non-Invasive Glucose Detection through AI: A Domain Generalization Approach

Yuyang Sun; Helena Cano-Garcia; Eleonora Razzicchia; Gabriela Gutierrez; Efthymios Kallos; Oana Ancu; Alex Rhodes; Diana-Elena Motei; Richard William Alexander Mackenzie; Panagiotis Kosmas

doi:10.1109/JBHI.2026.3698255

Improving Multi-Sensor Non-Invasive Glucose Detection through AI: A Domain Generalization Approach

Yuyang Sun
, Helena Cano-Garcia
, Eleonora Razzicchia
, Gabriela Gutierrez
, Efthymios Kallos
, Oana Ancu
, Alex Rhodes
, Diana-Elena Motei
, Richard William Alexander Mackenzie
, Panagiotis Kosmas

Xi’an Jiaotong University
MediWise Medical Wireless Sensing Ltd
McGill University
University of Roehampton
Queen Mary University of London
King's College London

Research output: Contribution to journal › Article › peer-review

Abstract

Accurate glucose level monitoring is crucial in diabetes management, aiming to ensure glucose levels are within a safe range and reduce the risk of complications. Inter-patient heterogeneity is one of the most important challenges to achieving accurate non-invasive glucose monitoring. This study employs meta-forests, a novel ensemble based domain generalization approach designed to address this challenge. Our technique is applied to a dataset of 54,280 data points, collected from five subjects over 10 days, using a non-invasive system that integrates near infrared (NIR) spectroscopy, millimeter-wave (mm-wave) sensing, and temperature measurements. Moreover, we significantly enhance model interpretability by incorporating Shapley additive explanations (SHAP) analyses. Importantly, our approach leads to an accuracy for the non-invasive glucose detection system that is comparable to state-of-the art methods, achieving an average root mean square error (RMSE) of 1.10 mmol/L and a mean absolute percentage error (MAPE) of 10.32% in subject-specific experiments.

Original language	English
Pages (from-to)	(In-Press)
Number of pages	14
Journal	IEEE Journal of Biomedical and Health Informatics
Volume	(In-Press)
Early online date	29 Jun 2026
DOIs	https://doi.org/10.1109/JBHI.2026.3698255
Publication status	E-pub ahead of print - 29 Jun 2026
Externally published	Yes

Bibliographical note

© 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Copyright © and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders.

This document is the author’s post-print version, incorporating any revisions agreed during the peer-review process. Some differences between the published version and this version may remain and you are advised to consult the published version if you wish to cite from it.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Machine learning
domain generalization
millimeter wave radar
infrared sensing
glucose prediction
diabetes
explainable artificial intelligence
interpretability

Access to Document

10.1109/JBHI.2026.3698255

Mackenzie2026AAMAccepted author manuscript, 2.11 MB

Cite this

Sun, Y., Cano-Garcia, H., Razzicchia, E., Gutierrez, G., Kallos, E., Ancu, O., Rhodes, A., Motei, D.-E., Mackenzie, R. W. A., & Kosmas, P. (2026). Improving Multi-Sensor Non-Invasive Glucose Detection through AI: A Domain Generalization Approach. IEEE Journal of Biomedical and Health Informatics, (In-Press), (In-Press). Advance online publication. https://doi.org/10.1109/JBHI.2026.3698255

@article{06529930c9224ffbae2eddf1a13e07aa,

title = "Improving Multi-Sensor Non-Invasive Glucose Detection through AI: A Domain Generalization Approach",

abstract = "Accurate glucose level monitoring is crucial in diabetes management, aiming to ensure glucose levels are within a safe range and reduce the risk of complications. Inter-patient heterogeneity is one of the most important challenges to achieving accurate non-invasive glucose monitoring. This study employs meta-forests, a novel ensemble based domain generalization approach designed to address this challenge. Our technique is applied to a dataset of 54,280 data points, collected from five subjects over 10 days, using a non-invasive system that integrates near infrared (NIR) spectroscopy, millimeter-wave (mm-wave) sensing, and temperature measurements. Moreover, we significantly enhance model interpretability by incorporating Shapley additive explanations (SHAP) analyses. Importantly, our approach leads to an accuracy for the non-invasive glucose detection system that is comparable to state-of-the art methods, achieving an average root mean square error (RMSE) of 1.10 mmol/L and a mean absolute percentage error (MAPE) of 10.32\% in subject-specific experiments.",

keywords = "Machine learning, domain generalization, millimeter wave radar, infrared sensing, glucose prediction, diabetes, explainable artificial intelligence, interpretability",

author = "Yuyang Sun and Helena Cano-Garcia and Eleonora Razzicchia and Gabriela Gutierrez and Efthymios Kallos and Oana Ancu and Alex Rhodes and Diana-Elena Motei and Mackenzie, \{Richard William Alexander\} and Panagiotis Kosmas",

note = "{\textcopyright} 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Copyright {\textcopyright} and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders. This document is the author{\textquoteright}s post-print version, incorporating any revisions agreed during the peer-review process. Some differences between the published version and this version may remain and you are advised to consult the published version if you wish to cite from it. ",

year = "2026",

month = jun,

day = "29",

doi = "10.1109/JBHI.2026.3698255",

language = "English",

volume = "(In-Press)",

pages = "(In--Press)",

journal = "IEEE Journal of Biomedical and Health Informatics",

issn = "2168-2194",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Improving Multi-Sensor Non-Invasive Glucose Detection through AI

T2 - A Domain Generalization Approach

AU - Sun, Yuyang

AU - Cano-Garcia, Helena

AU - Razzicchia, Eleonora

AU - Gutierrez, Gabriela

AU - Kallos, Efthymios

AU - Ancu, Oana

AU - Rhodes, Alex

AU - Motei, Diana-Elena

AU - Mackenzie, Richard William Alexander

AU - Kosmas, Panagiotis

N1 - © 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Copyright © and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders. This document is the author’s post-print version, incorporating any revisions agreed during the peer-review process. Some differences between the published version and this version may remain and you are advised to consult the published version if you wish to cite from it.

PY - 2026/6/29

Y1 - 2026/6/29

N2 - Accurate glucose level monitoring is crucial in diabetes management, aiming to ensure glucose levels are within a safe range and reduce the risk of complications. Inter-patient heterogeneity is one of the most important challenges to achieving accurate non-invasive glucose monitoring. This study employs meta-forests, a novel ensemble based domain generalization approach designed to address this challenge. Our technique is applied to a dataset of 54,280 data points, collected from five subjects over 10 days, using a non-invasive system that integrates near infrared (NIR) spectroscopy, millimeter-wave (mm-wave) sensing, and temperature measurements. Moreover, we significantly enhance model interpretability by incorporating Shapley additive explanations (SHAP) analyses. Importantly, our approach leads to an accuracy for the non-invasive glucose detection system that is comparable to state-of-the art methods, achieving an average root mean square error (RMSE) of 1.10 mmol/L and a mean absolute percentage error (MAPE) of 10.32% in subject-specific experiments.

AB - Accurate glucose level monitoring is crucial in diabetes management, aiming to ensure glucose levels are within a safe range and reduce the risk of complications. Inter-patient heterogeneity is one of the most important challenges to achieving accurate non-invasive glucose monitoring. This study employs meta-forests, a novel ensemble based domain generalization approach designed to address this challenge. Our technique is applied to a dataset of 54,280 data points, collected from five subjects over 10 days, using a non-invasive system that integrates near infrared (NIR) spectroscopy, millimeter-wave (mm-wave) sensing, and temperature measurements. Moreover, we significantly enhance model interpretability by incorporating Shapley additive explanations (SHAP) analyses. Importantly, our approach leads to an accuracy for the non-invasive glucose detection system that is comparable to state-of-the art methods, achieving an average root mean square error (RMSE) of 1.10 mmol/L and a mean absolute percentage error (MAPE) of 10.32% in subject-specific experiments.

KW - Machine learning

KW - domain generalization

KW - millimeter wave radar

KW - infrared sensing

KW - glucose prediction

KW - diabetes

KW - explainable artificial intelligence

KW - interpretability

UR - https://www.scopus.com/pages/publications/105043681957

U2 - 10.1109/JBHI.2026.3698255

DO - 10.1109/JBHI.2026.3698255

M3 - Article

SN - 2168-2194

VL - (In-Press)

SP - (In-Press)

JO - IEEE Journal of Biomedical and Health Informatics

JF - IEEE Journal of Biomedical and Health Informatics

ER -

Improving Multi-Sensor Non-Invasive Glucose Detection through AI: A Domain Generalization Approach

Abstract

Bibliographical note

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this