Glucose measurement predictions using stacked machine learning models

US12390131B2 · kind B2 · utility

Assignee

• Dexcom, Inc. · US

Inventors

• Mark Derdzinski · Village of La Jolla, US
• Joost Herman VAN DER LINDEN · San Diego, US
• Robert Dowd · San Diego, US
• Lauren Hruby Jepson · San Diego, US
• Giada Acciaroli · Edinburgh, GB

Key dates

• Filing date: May 28, 2021
• Grant date: Aug 19, 2025
• Expiry date: Oct 23, 2041

Classification

• Technology area (CPC G): Physics
• CPC primary: G06N3/08
• WIPO field: Medical technology
• WIPO sector: Instruments

Abstract

Glucose measurement and glucose-impacting event prediction using a stack of machine learning models is described. A CGM platform includes stacked machine learning models, such that an output generated by one of the machine learning models can be provided as input to another one of the machine learning models. The multiple machine learning models include at least one model trained to generate a glucose measurement prediction and another model trained to generate an event prediction, for an upcoming time interval. Each of the stacked machine learning models is configured to generate its respective output when provided as input at least one of glucose measurements provided by a CGM system worn by the user or additional data describing user behavior or other aspects that impact a person's glucose in the future. Predictions may then be output, such as via communication and/or display of a notification about the corresponding prediction.