Please use this identifier to cite or link to this item:
|Title:||Prediction of blood glucose level via the use of various machine learning models||Authors:||Aiman Ibrahim||Keywords:||Engineering::Mechanical engineering||Issue Date:||2022||Publisher:||Nanyang Technological University||Source:||Aiman Ibrahim (2022). Prediction of blood glucose level via the use of various machine learning models. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/159083||Project:||C048||Abstract:||Diabetes is a disease that occurs when one’s blood glucose level is higher than the standards. It is essential for diabetic patients to monitor their blood glucose level frequently. If the blood glucose level is not monitored regularly and if it is higher than the standards, it may result to serious complications. A high blood glucose level may lead to the damage of vital organs and nerves. Hence, diabetic patients must monitor their blood glucose level regularly. Often blood glucose meters that are found in the market are invasive and painful. Diabetic patients have to endure multiple pricks each day to monitor their blood glucose level. Machine learning models are widely used in the high technology world today. Machine learning enables users to classify an image correctly, predict text messages, make important decisions and is also used in an autonomous vehicle. The integration of various machine learning models in this final year project aims to find a non-invasive method that provides continuous blood glucose monitoring for diabetic patients In this final year project 30 participants volunteered to be part of the research study. A 3-minute Electrocardiogram (ECG) and Photoplethysmogram (PPG) signals were collected from them via the use of the current prototype sensor that was designed by one of the team members. Their blood glucose level during the time of the data collection process were also recorded down. The 3-minute data were pre-processed and an algorithm to section the PPG signals into single PPG waveform was applied. Features of the single PPG waveform were then extracted, and four various machine learning models are then applied onto the dataset. The Random Forest Regression (RFR) model was found to be the best machine learning model to estimate the blood glucose level when compared with the Support Vector Regression (SVR), XG-Boost Regressor (XG-BR) and 1D-Convolutional Neural Network (1DCNN).||URI:||https://hdl.handle.net/10356/159083||Fulltext Permission:||restricted||Fulltext Availability:||With Fulltext|
|Appears in Collections:||MAE Student Reports (FYP/IA/PA/PI)|
Updated on Dec 1, 2022
Updated on Dec 1, 2022
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.