(Department of Computer Science, School of Data Science, Mathematics and Informatics, IPB University, Indonesia)(2) Karlisa Priandana
(Department of Computer Science, School of Data Science, Mathematics and Informatics, IPB University, Indonesia)(3) Husin Alatas
(Department of Physics, Faculty of Mathematics and Natural Sciences, IPB University, Indonesia)(4) Renan Prasta Jenie
(Department of Digital Business, Faculty of Business and Social Science, Binawan University, Jakarta, Indonesia)*corresponding author
AbstractDiabetes remains a major global health challenge. According to the International Diabetes Federation (IDF), approximately 10.5% of the world’s population is affected, with nearly half unaware of their condition. Therefore, accurate and accessible blood glucose level (BGL) monitoring remains critical. This paper presents a noninvasive BGL estimation approach using in vivo residual infrared (IR) signals collected from patients’ fingertips, which is inherently challenging due to weak glucose-related signatures and interference from other physiological factors. To address this, we propose a tailored feature engineering strategy to uncover hidden relationships between residual signals and BGL. The proposed method improves BGL range discrimination, as indicated by increased Spearman correlations and consistent gains in Kendall and Pearson correlations, demonstrating its effectiveness. Furthermore, we propose a task-specific hybrid neural network architecture, consisting of parallel 1D convolutional and dense blocks to capture both local spectral–temporal patterns and global nonlinear relationships. A distribution-aware loss function is also introduced to reduce prediction bias toward the dataset mean, improving sensitivity across the full BGL range. For validation, the proposed model is compared with random forest (RF) and boosting-based methods, showing superior performance with a 14.11% mean absolute percentage error (MAPE). Clinical reliability is evaluated using Clarke error grid (CEG) analysis, where 84.7% of predictions fall within the accurate zone and 15.3% within the clinically acceptable zone. These results demonstrate the potential of the proposed approach for practical, low-cost, and portable noninvasive BGL monitoring.
KeywordsNoninvasive blood glucose; Residual infrared signals; Feature engineering; Hybrid deep learning; Distribution-aware loss
|
DOIhttps://doi.org/10.26555/ijain.v12i3.1970 |
Article metricsAbstract views : 118 | PDF views : 16 |
Cite |
Full Text Download
|
References
[1] “Diabetes facts and figures,” International Diabetes Federation (IDF), 2023. [Online]. Available at: https://idf.org/about-diabetes/facts-figures/.
[2] R. Khan, Z. Chua, J. Tan, Y. Yang, Z. Liao, and Y. Zhao, “From Pre-Diabetes to Diabetes: Diagnosis, Treatments and Translational Research,” Medicina (B. Aires)., vol. 55, no. 9, p. 546, Aug. 2019, doi: 10.3390/medicina55090546.
[3] J. Wu, Z. Peng, X. Zhang, W. He, and C. Yang, “Wavelet-ResNet: A deep residual network combined with wavelet transform for photoacoustic blood glucose detection,” Biomed. Signal Process. Control, vol. 97, p. 106718, Nov. 2024, doi: 10.1016/j.bspc.2024.106718.
[4] C. Xiong, Z. Ren, and T. Liu, “Quantitative blood glucose detection influenced by various factors based on the fusion of photoacoustic temporal spectroscopy with deep convolutional neural networks,” Biomed. Opt. Express, vol. 15, no. 5, p. 2719, May 2024, doi: 10.1364/BOE.521059.
[5] J. Parab, M. Sequeira, M. Lanjewar, C. Pinto, and G. Naik, “Backpropagation Neural Network-Based Machine Learning Model for Prediction of Blood Urea and Glucose in CKD Patients,” IEEE J. Transl. Eng. Heal. Med., vol. 9, pp. 1–8, 2021, doi: 10.1109/JTEHM.2021.3079714.
[6] A. Aloraynan, S. Rassel, M. R. Kaysir, and D. Ban, “Dual quantum cascade lasers for noninvasive glucose detection using photoacoustic spectroscopy,” Sci. Rep., vol. 13, no. 1, p. 7927, May 2023, doi: 10.1038/s41598-023-34912-3.
[7] S. Vanaja et al., “Non-invasive glucometer monitoring system through optical based near-infrared sensor method,” Comput. Methods Biomech. Biomed. Eng. Imaging Vis., vol. 12, no. 1, p. 12, Dec. 2024, doi: 10.1080/21681163.2024.2327423.
[8] Satish, K. Sen, and S. Anand, “Analysis of glucose‐dependent dielectric properties of aqueous‐based solution: A proof of concept,” IET Sci. Meas. Technol., vol. 15, no. 7, pp. 562–568, Sep. 2021, doi: 10.1049/smt2.12057.
[9] R. P. Jenie et al., “Non-Invasive Optical Blood Glucose Measurement based on Discrete Fourier Transform and Fast Artificial Neural Network: Fasting Normal Glucose Participants Case Study,” J. Med. Device., vol. 13, no. 4, p. 7, Jul. 2019, doi: 10.1115/1.4044336.
[10] Y. Yang et al., “Noninvasive Blood Glucose Detection System With Infrared Pulse Sensor and Hybrid Feature Neural Network,” IEEE Sens. J., vol. 24, no. 8, pp. 13385–13394, Apr. 2024, doi: 10.1109/JSEN.2024.3373048.
[11] M. Naresh, V. S. Nagaraju, S. Kollem, J. Kumar, and S. Peddakrishna, “Non-invasive glucose prediction and classification using NIR technology with machine learning,” Heliyon, vol. 10, no. 7, p. e28720, Apr. 2024, doi: 10.1016/j.heliyon.2024.e28720.
[12] G. Han, S. Chen, X. Wang, J. Wang, H. Wang, and Z. Zhao, “Noninvasive blood glucose sensing by near-infrared spectroscopy based on PLSR combines SAE deep neural network approach,” Infrared Phys. Technol., vol. 113, p. 103620, Mar. 2021, doi: 10.1016/j.infrared.2020.103620.
[13] L. M. Pires and J. Martins, “Design and Implementation of a Low-Power Device for Non-Invasive Blood Glucose,” Designs, vol. 8, no. 4, p. 63, Jun. 2024, doi: 10.3390/designs8040063.
[14] Y. Zheng et al., “Highly Sensitive Perovskite Photoplethysmography Sensor for Blood Glucose Sensing Using Machine Learning Techniques,” Adv. Sci., vol. 11, no. 43, pp. 1–10, Nov. 2024, doi: 10.1002/advs.202405681.
[15] S. V. K. R. Rajeswari and P. Vijayakumar, “Development of sensor system and data analytic framework for non-invasive blood glucose prediction,” Sci. Rep., vol. 14, no. 1, p. 9206, Apr. 2024, doi: 10.1038/s41598-024-59744-7.
[16] T. Kitazaki et al., “Glucose emission spectra through mid-infrared passive spectroscopic imaging of the wrist for non-invasive glucose sensing,” Sci. Rep., vol. 12, no. 1, p. 20558, Nov. 2022, doi: 10.1038/s41598-022-25161-x.
[17] J. Hanna et al., “Wearable flexible body matched electromagnetic sensors for personalized non-invasive glucose monitoring,” Sci. Rep., vol. 12, no. 1, p. 14885, Sep. 2022, doi: 10.1038/s41598-022-19251-z.
[18] S. Kiani and P. Rezaei, “Microwave substrate integrated waveguide resonator sensor for non-invasive monitoring of blood glucose concentration: Low cost and painless tool for diabetics,” Measurement, vol. 219, p. 113232, Sep. 2023, doi: 10.1016/j.measurement.2023.113232.
[19] S. Harnsoongnoen and B. Buranrat, “Advances in a Microwave Sensor-Type Interdigital Capacitor with a Hexagonal Complementary Split-Ring Resonator for Glucose Level Measurement,” Chemosensors, vol. 11, no. 4, p. 257, Apr. 2023, doi: 10.3390/chemosensors11040257.
[20] C. G. Juan et al., “On the Selectivity of Planar Microwave Glucose Sensors with Multicomponent Solutions,” Electronics, vol. 12, no. 1, p. 191, Dec. 2022, doi: 10.3390/electronics12010191.
[21] A. Kandwal et al., “Designing Highly Sensitive Microwave Antenna Sensor with Novel Model for Noninvasive Glucose Measurements,” Prog. Electromagn. Res., vol. 176, pp. 129–141, 2023, doi: 10.2528/PIER22113002.
[22] T. Igbe, O. W. Samuel, J. Li, F. Kulwa, A. Kandwal, and Z. Nie, “Inspection of EEG Signals for Noninvasive Blood Glucose Monitoring in Prediabetes Diagnosis,” in 2023 IEEE International Symposium on Medical Measurements and Applications (MeMeA), Jun. 2023, pp. 1–5, doi: 10.1109/MeMeA57477.2023.10171941.
[23] A. Othmane, I. Youkana, L. Kahloul, and S. Bourekkache, “AI-Based Prediction for Glucose Levels: A Comparative Study of Machine Learning and Deep Learning Approaches,” 2024, pp. 231–245, doi: 10.2991/978-94-6463-496-9_18.
[24] X. Huang et al., “Comparison of feature learning methods for non-invasive interstitial glucose prediction using wearable sensors in healthy cohorts: a pilot study,” Intell. Med., vol. 4, no. 4, pp. 226–238, Nov. 2024, doi: 10.1016/j.imed.2024.05.002.
[25] M. F. Rabby, Y. Tu, M. I. Hossen, I. Lee, A. S. Maida, and X. Hei, “Stacked LSTM based deep recurrent neural network with kalman smoothing for blood glucose prediction,” BMC Med. Inform. Decis. Mak., vol. 21, pp. 1–15, 2021, doi: 10.1186/s12911-021-01462-5.
[26] A. Gade, V. Vijaya Baskar, and J. Panneerselvam, “Hybrid model with optimal features for non-invasive blood glucose monitoring from breath biomarkers,” Biomed. Signal Process. Control, vol. 88, p. 105036, Feb. 2024, doi: 10.1016/j.bspc.2023.105036.
[27] J. M. Lee, R. Pop-Busui, J. M. Lee, J. Fleischer, and J. Wiens, “Shortcomings in the Evaluation of Blood Glucose Forecasting,” IEEE Trans. Biomed. Eng., vol. 71, no. 12, pp. 3424–3431, Dec. 2024, doi: 10.1109/TBME.2024.3424665.
[28] Y. Shen and S. Kleinberg, “Personalized Blood Glucose Forecasting From Limited CGM Data Using Incrementally Retrained LSTM,” IEEE Trans. Biomed. Eng., vol. 72, no. 4, pp. 1266–1277, Apr. 2025, doi: 10.1109/TBME.2024.3494732.
[29] Q. Bian, A. As’arry, X. Cong, K. A. bin M. Rezali, and R. M. K. bin Raja Ahmad, “A hybrid Transformer-LSTM model apply to glucose prediction,” PLoS One, vol. 19, no. 9, p. e0310084, Sep. 2024, doi: 10.1371/journal.pone.0310084.
[30] N. Niako, J. D. Melgarejo, G. E. Maestre, and K. P. Vatcheva, “Effects of missing data imputation methods on univariate blood pressure time series data analysis and forecasting with ARIMA and LSTM,” BMC Med. Res. Methodol., vol. 24, no. 1, p. 320, Dec. 2024, doi: 10.1186/s12874-024-02448-3.
[31] H. Nemat, H. Khadem, J. Elliott, and M. Benaissa, “Data-driven blood glucose level prediction in type 1 diabetes: a comprehensive comparative analysis,” Sci. Rep., vol. 14, no. 1, p. 21863, Sep. 2024, doi: 10.1038/s41598-024-70277-x.
[32] M. A. R. Refat, M. Al Amin, C. Kaushal, M. N. Yeasmin, and M. K. Islam, “A Comparative Analysis of Early Stage Diabetes Prediction using Machine Learning and Deep Learning Approach,” in 2021 6th International Conference on Signal Processing, Computing and Control (ISPCC), Oct. 2021, pp. 654–659, doi: 10.1109/ISPCC53510.2021.9609364.
[33] A. Zarghani, “Comparative Analysis of LSTM Neural Networks and Traditional Machine Learning Models for Predicting Diabetes Patient Readmission,” arxiv Artif. Intell., pp. 1–21, 2024. [Online]. Available at: https://arxiv.org/pdf/2406.19980.
[34] A. Prabha, J. Yadav, A. Rani, and V. Singh, “Design of intelligent diabetes mellitus detection system using hybrid feature selection based XGBoost classifier,” Comput. Biol. Med., vol. 136, p. 104664, Sep. 2021, doi: 10.1016/j.compbiomed.2021.104664.
[35] T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” Proc. ACM SIGKDD Int. Conf. Knowl. Discov. Data Min., vol. 13-17-August-2016, pp. 785–794, Aug. 2016, doi: 10.1145/2939672.2939785.
[36] M. Komorowski, D. C. Marshall, J. D. Salciccioli, and Y. Crutain, “Exploratory data analysis,” Second. Anal. Electron. Heal. Rec., pp. 185–203, Jan. 2016, doi: 10.1007/978-3-319-43742-2_15.
[37] R. Joeres, D. B. Blumenthal, and O. V. Kalinina, “DataSAIL: Data Splitting Against Information Leakage.” pp. 1–18, Nov. 17, 2023, doi: 10.1101/2023.11.15.566305.
[38] T. Han et al., “In Vivo Near-Infrared Noninvasive Glucose Measurement and Detection in Humans,” Appl. Spectrosc., vol. 76, no. 9, pp. 1100–1111, Sep. 2022, doi: 10.1177/00037028221092474.
[39] M. Yang et al., “Miniaturized Optical Glucose Sensor Using 1600–1700 nm Near‐Infrared Light,” Adv. Sens. Res., vol. 4, no. 3, Mar. 2025, doi: 10.1002/adsr.202300160.
[40] Y. LeCun, L. Bottou, G. B. Orr, and K.-R. Müller, “Efficient BackProp,” 1998, pp. 9–50, doi: 10.1007/3-540-49430-8_2.
[41] S. Ioffe and C. Szegedy, “Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift,” 32nd Int. Conf. Mach. Learn. ICML 2015, vol. 1, pp. 448–456, Feb. 2015. [Online]. Available at: https://arxiv.org/abs/1502.03167v3.
[42] C. Li et al., “YOLOv6: A Single-Stage Object Detection Framework for Industrial Applications,” arxiv Artif. Intell., pp. 1–17, 2022. [Online]. Available at: https://github.com/meituan/YOLOv6.
[43] W. L. Clarke, “The Original Clarke Error Grid Analysis (EGA),” Diabetes Technol. Ther., vol. 7, no. 5, pp. 776–779, Oct. 2005, doi: 10.1089/dia.2005.7.776.
[44] M. Zeynali, K. Alipour, B. Tarvirdizadeh, and M. Ghamari, “Non-invasive blood glucose monitoring using PPG signals with various deep learning models and implementation using TinyML,” Sci. Rep., vol. 15, no. 1, p. 581, Jan. 2025, doi: 10.1038/s41598-024-84265-8.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
___________________________________________________________
International Journal of Advances in Intelligent Informatics
ISSN 2442-6571 (print) | 2548-3161 (online)
Organized by UAD and ASCEE Computer Society
Published by Universitas Ahmad Dahlan
W: http://ijain.org
E: info@ijain.org (paper handling issues)
andri.pranolo.id@ieee.org (publication issues)
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0