Glucose prediction using machine learning and time series glucose measurements

• Main Authors: PARK ALEXANDER S, DEREZINSKI MICHAEL
• Format: Patent
• Language: Chinese; English
• Published: 30.12.2022
• Subjects: DIAGNOSIS; HUMAN NECESSITIES; HYGIENE; IDENTIFICATION; INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIRCHEMICAL OR PHYSICAL PROPERTIES; MEASURING; MEDICAL OR VETERINARY SCIENCE; PHYSICS; SURGERY; TESTING

Glucose prediction using machine learning (ML) and time series glucose measurements is described. In view of a large number of people wearing glucose monitoring devices, and since some wearable glucose monitoring devices can continuously produce measurements, the platform providing such devices may have a large amount of data. In practice, even not in practice, humans cannot process this amount of data, and this amount of data covers a huge amount of state space, which is impossible to be covered if there is no huge amount of data. In an embodiment, a glucose monitoring platform includes an ML model trained using historical time series glucose measurements of a user population. The ML model predicts an upcoming glucose measurement for a particular user by receiving a time series of glucose measurements prior to a time and determining an upcoming glucose measurement for the particular user within an interval after the time based on a pattern learned from historical time series of glucose measurements. 描述了使用机器学