Self-calibrating glucose monitor

A medical system comprising:

• a plurality of individual glucose sensors, wherein each glucose sensor is configured to determine a glucose level in a patient;
• sensing circuitry configured to sense a cardiac characteristic of a heart of the patient;
• processing circuitry operably coupled to one or more of the glucose sensors and the sensing circuitry, the processing circuitry configured to:
• receive a current cardiac signal indicative of the cardiac characteristic from the sensing circuitry, wherein the current cardiac signal is indicative of a current cardiac characteristic of the patient, and
• determine a representative glucose level using a machine learning algorithm and the current cardiac signal,
• wherein, previous to determining the representative glucose level, the machine learning algorithm was trained using one or more training data sets including a training input vector and a training output vector, wherein the training input vector is representative of a cardiac signal and the training output vector is representative of a glucose level associated with the cardiac signal;
• and
• a portable device configured to be worn by the patient, wherein the portable device includes a housing configured to support at least one of the plurality of glucose sensors, the sensing circuitry, or the processing circuitry.

A method for controlling processing circuitry of a medical system, the method comprising:

• receiving, using the processing circuitry, a current cardiac signal indicative of a cardiac characteristic of a heart of a patient from sensing circuitry of the medical system, the sensing circuitry configured to sense the cardiac characteristic;
• and
• determining, using the processing circuitry, a representative glucose level using a machine learning algorithm and the current cardiac signal,
• wherein, previous to determining the representative glucose level, the machine learning algorithm was trained using one or more training data sets including a training input vector and a training output vector, wherein the training input vector is representative of a cardiac signal and the training output vector is representative of a glucose level associated with the cardiac signal.