Predicting Blood Glucose Levels in Diabetic Patients using Neural Networks
London Central Secondary School
Diabetes is one of the leading causes of death worldwide. Type 1 diabetes occurs when the pancreas produces little or no insulin. In type 2 diabetes, the patient’s body does not use insulin correctly. Insulin is a hormone that signals liver, muscle and fat cells to uptake blood glucose. In diabetes, blood sugar levels remain high, and cells are deprived of their main source of energy. To counter this issue, patients are given insulin to inject into the fat layer just below the skin. Patients usually inject insulin when their blood sugar levels are spiking or could spike. Continuous glucose monitoring (CGM) systems allow patients to monitor this. Forecasting blood glucose can assist in the management of diabetes as a tool for meal planning and timing of insulin dosages. This project applies artificial neural networks to forecast blood glucose of patients using past values, insulin usage, nutrition, and activity.
Excellence Award – Intermediate Silver Medal
Sponsor: Youth Science Canada
Special Award – Ted Rogers Innovation Award
Sponsor: Rogers Communications Inc.