Tracking sugar in the blood is crucial for both healthy individuals and diabetic patients. Currently, Continuous Glucose Monitors (GCM) are available by the NHS for hypoglycaemia. They measure glucose in interstial fluid using an invasive sensor, which sends alarms and data to a display device. In many cases, they requires calibration twice a day with invasive finger-prick blood glucose level tests.
Researchers at the University of Warwick have been improving upon existing technologies through using artificial intelligence to detect hypoglycaemic events from raw ECG signals acquired using off-the-shelf non-invasive wearable or ambient sensors. Warwick’s algorithm is based on deep-learning and automatically detects hypo events in real-me using nonfiltered data from a single ECG sensor. The algorithm is trained using individual patients’ own CGM and ECG measurements for 1 week. Aer which, the algorithm operates as described below, but without any further glucose measurements being required. The features mentioned above, can be relevant to the clinicians to understand how heartbeats change in individual subjects during a hypo event, which may beer inform therapy.