University of Twente Student Theses

Login

Predicting blood glucose for type 2 diabetes patients

Meij, D.R. de (2018) Predicting blood glucose for type 2 diabetes patients.

[img] PDF
2MB
Abstract:It is important for diabetics to keep their blood glucose in a healthy range. The Delicate project aims to use data collected on blood glucose, food intake, physical activity and health records to provide type 2 diabetics with personalized coaching and with blood glucose predictions. In this research, we predict future blood glucose values based on a patient's characteristics and behavior. We also determine how such as prediction model can be deployed and how the different input features influence the prediction. We were able to significantly (p<0.1) outperform our autoregressive baseline on longer time horizons (>= 60 minutes) using a multitask long short-term memory network (LSTM). The multitask LSTM predicts blood glucose for multiple timesteps into the future at the same time. This not only improves performance but also makes it more convenient to apply in a real-world application. The trained multitask LSTM uses input features such as food intake in a consistent manner, this makes it useful in showing patients how their actions affect their predicted blood glucose. We recommend visualizing the expected error of the predicted blood glucose in such a way that patients are aware of the limitations of the model, while still benefiting from the insight it provides.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:44 medicine, 54 computer science
Programme:Interaction Technology MSc (60030)
Link to this item:http://purl.utwente.nl/essays/76661
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page