Home // AIHealth 2025, The Second International Conference on AI-Health // View article
Authors:
Avijay Sen
Dr. Sindhu Ghanta
Pallavi Bajpai
Keywords: blood glucose prediction; Type-1 Diabetes Mellitus; insulin delivery system
Abstract:
Type-1 Diabetes Mellitus (T1DM) is a chronic condition characterized by the pancreas's inability to produce insulin, requiring continuous monitoring and management of blood glucose levels. Accurate prediction of blood glucose levels can significantly improve patient outcomes by reducing hypo- and hyperglycemic events. This study develops a personalized automated blood glucose forecasting system leveraging the past blood glucose levels and insulin pump data. Utilizing the publicly available Diatrend dataset, encompassing thirty-one days of data for five subjects, we evaluated three machine learning algorithms: K-Nearest Neighbors (KNN), Random Forest (RF), and Multilayer Perceptron (MLP). After hyper-parameter tuning, the performance of each algorithm was assessed using Root Mean Squared Error (RMSE), Mean Squared Error (MSE), and the coefficient of determination (R2), with a particular emphasis on RMSE. The Random Forest model demonstrated superior performance, achieving a test RMSE range of 14.98-23.62 across all subjects. This research highlights the efficacy of supervised machine learning algorithms in predicting blood glucose levels over one-hour intervals for T1DM patients, underscoring the potential of personalized machine learning models to improve diabetes management.
Pages: 47 to 54
Copyright: Copyright (c) IARIA, 2025
Publication date: March 9, 2025
Published in: conference
ISBN: 978-1-68558-247-0
Location: Lisbon, Portugal
Dates: from March 9, 2025 to March 13, 2025