Techniques for improved automatic drug delivery performance using delivery tendencies from past delivery history and use patterns

A system, comprising:

• a processor;
• a memory storing programming code, an artificial pancreas application, and operable to store data related to the artificial pancreas application, wherein the programming code and the artificial pancreas application are executable by the processor;
• and
• a transceiver operable to receive and transmit signals containing information usable by or generated by the artificial pancreas application when executed by the processor,
• wherein the processor when executing the artificial pancreas application is configured to:
• receive blood glucose measurements over a period of time;
• generate predicted glucose values using a glucose prediction model;
• compare the predicted glucose values from the glucose prediction model to measured blood glucose values of the received blood glucose measurements;
• identify deviations between the predicted blood glucose values and additional blood glucose measurements;
• evaluate the identified deviations with respect to a blood glucose deviation threshold;
• determine a quantity of the identified deviations that exceed the blood glucose deviation threshold;
• based on the result of the determination, recalculate error parameters of the glucose prediction model;
• update glucose and insulin dynamics coefficients of the glucose prediction model using the recalculated error parameters to provide new glucose and insulin dynamics coefficients;
• receive subsequent blood glucose measurements over a subsequent period of time;
• using the new glucose and insulin dynamics coefficients and the received subsequent blood glucose measurements in an updated glucose prediction model, obtain predictions of future glucose measurements;
• calculate a future insulin dosage based on the obtained predictions of future glucose measurements;
• and
• cause delivery of the calculated future insulin dosage to a user.