(757a) Development of a Model-Based Noninvasive Glucose Monitoring Device for Non-Insulin Dependent People

Continuous-time glucose monitoring (CGM) effectively improves glucose control by providing frequently sampled information that allows the user to associate changes in their glucose levels with changes in their behavior. Currently, the most widely used and effective CGM devices rely on a sensor that is inserted invasively under the skin. The primary users of current CGM devices are insulin dependent people (type 1 and some type 2 diabetics) but there is a need for non-invasive (and less expensive) CGM devices to help other diabetic and pre-diabetic individuals monitor blood sugar levels. Thus, our motivation is the development of an accurate CGM device that will be widely used by non-insulin dependent people. We have developed prediction models that infer glucose concentrations using sensor data from a commercially available non-invasive physical activity monitor along with self-reported diet information. Using 22 test subjects with 4 weeks of data collection each, results have been obtained to support the viability and utility of these models. Key accomplishments in this research include the ability to develop subject-specific models under several modeling challenges and restrictions. Results suggest an accurate model can be developed after an initial calibration period of three days with accuracy improving over time and no need for lancet measurements after 3 to 4 weeks. Moreover, since inference comes from a stationary mathematical methodology, it does not have sensor drift and thus, will not require periodic recalibration to compensate for chemical changes as necessary for a hard-sensor. Thus, an application to a specific user has the potential to last indefinitely without the need for recalibration, which means updating the model in this context. However, lancet measurements (3-4 times per day) will be needed to improve the fit of the model or to adapt to changing relationships between the blood glucose concentrations and the measured inputs.