AI-Based Prediction for Glucose Levels: A Comparative Study of Machine Learning and Deep Learning Approaches

DOI

10.2991/978-94-6463-496-9_18How to use a DOI?

Keywords

Blood glucose Prediction; Photoplethysmography (PPG); Machine learning; Deep learning; Regression; CNN

Abstract

For the sake of diabetes management, patients need to measure their blood glucose level consistently, which could be challenging and stressful, since most of the time the used method is going to be invasive method which involves pricking the skin to obtain a blood sample to use it to measure the glucose level, or a semi-invasive method that requires the insertion of a sensor beneath the skin, so it still involve level of invasiveness. For this purpose, several studies have been done to accomplish the non invasive measurement that can make people with diabetes more relaxed while checking their blood glucose level daily, using different machine learning and deep learning algorithms, different sensors and devices and different physiological factors, such as Photoplethysmography (PPG). In this paper, we compare using numerous metrics the performance of Machine Learning and Deep Learning techniques, in order to predict glucose levels non invasively using our collected PPG dataset. We assess the performance of several machine learning algorithms including Linear Regression, Lasso regression, Ridge Regression, Support Vector Regression, Gradient Boosting, Regressor, AdaBoost Regressor, XGB Regressor, and also, different deep learning algorithms CNN, RNN, CLDNN. All the experiments were conducted using our collected dataset. The results show that the Convolutional Neural Network (CNN) outperformed significantly the rest of models. Where, it provides the lowest values for most error metrics (MAE, MSE, RMSE, MAPE), a very high R2 score, indicating superior performance in minimizing prediction errors.

Copyright

© 2024 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Amani Othmane
AU  - Imane Youkana
AU  - Laid Kahloul
AU  - Samir Bourekkache
PY  - 2024
DA  - 2024/08/31
TI  - AI-Based Prediction for Glucose Levels: A Comparative Study of Machine Learning and Deep Learning Approaches
BT  - Proceedings of the International Conference on Emerging Intelligent Systems for Sustainable Development (ICEIS 2024)
PB  - Atlantis Press
SP  - 231
EP  - 245
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-496-9_18
DO  - 10.2991/978-94-6463-496-9_18
ID  - Othmane2024
ER  -