US20260013800
2026-01-15
Human necessities
A61B5/7282
A novel computer-implemented method and system have been developed to predict and display glucose values for individuals with diabetes. The system utilizes continuous glucose monitoring (CGM) data, carbohydrate ingestion data, and insulin administration data to forecast glucose levels. It is particularly effective in predicting hypoglycemia events by focusing on two prediction time windows: a longer first window and a shorter second window that begins simultaneously with the first.
The system processes CGM data to determine a series of predicted glucose values for a specified prediction time window. It identifies potential hypoglycemia events within a shorter, concurrent prediction window. This approach ensures that users are alerted to immediate risks without being overwhelmed by future data, which is not displayed beyond the second window.
A data processing system, equipped with a processor, is configured to handle the input from CGM devices, carbohydrate intake, and insulin administration. It calculates predicted glucose values and assesses the likelihood of hypoglycemic events. The system displays only the relevant predicted values, focusing on the immediate risk period to aid user decision-making.
The method includes generating and displaying predicted analyte trend graphs. These graphs visually represent glucose levels over the specified prediction windows. The system automatically updates the display to show the shorter prediction window when a hypoglycemia risk is detected, enhancing the user's ability to manage their condition proactively.
An application (app) is available, facilitating the visualization of predicted glucose values on user devices. It employs algorithms to determine hypoglycemic risks based on probability thresholds, using historical and current data. The app's design ensures that users receive timely alerts, focusing on the most critical prediction window for immediate intervention.