Dexcom, Inc.

University of Virginia Patent Foundation

Universitá Degli Studi Di Padova

Giovanni Sparacino

Claudio Cobelli

Andrea Facchinetti

Apurv Ullas Kamath

Hari Hampapuram

Naresh Chandra Bhavaraju

Chiara Zecchin

Anna Leigh Davis

Stefania Guerra

Michele Schiavon

Anna Leigh Rack-Gomer

Martina Vettoretti

Simone Del Favero

Boris P Kovatchev

Giuseppe De Nicolao

Naresh Bhavaraju

A mathematical model of type 1 diabetes (T1D) patient decision-making can be used to simulate, in silico, realistic glucose/insulin dynamics, for several days, in a variety of subjects who take therapeutic actions (e.g. insulin dosing) driven by either self-monitoring blood glucose (SMBG) or continuous glucose monitoring (CGM). The decision-making (DM) model can simulate real-life situations and everyday patient behaviors. Accurate submodels of SMBG and CGM measurement errors are incorporated in the comprehensive DM model. The DM model accounts for common errors the patients are used to doing in their diabetes management, such as miscalculations of meal carbohydrate content, early/delayed insulin administrations and missed insulin boluses. The DM model can be used to assess in silico if/when CGM can safely substitute SMBG in T1D management, to develop and test guidelines for CGM driven insulin dosing, to optimize and individualize off-line insulin therapies and to develop and test decision support systems.

Individualized multiple-day simulation model of type i diabetic patient decision-making for developing, testing and optimizing insulin therapies driven by glucose sensors

Continuous Glucose Monitoring (CGM) devices provide glucose concentration measurements in the subcutaneous tissue with limited accuracy and precision. Therefore, CGM readings cannot be incorporated in a straightforward manner in outcome metrics of clinical trials e.g. aimed to assess new glycemic-regulation therapies. To define those outcome metrics, frequent Blood Glucose (BG) reference measurements are still needed, with consequent relevant difficulties in outpatient settings. Here we propose a 'retrofitting' algorithm that produces a quasi continuous time BG profile by simultaneously exploiting the high accuracy of available BG references (possibly very sparsely collected) and the high temporal resolution of CGM data (usually noisy and affected by significant bias). The inputs of the algorithm are: a CGM time series; some reference BG measurements; a model of blood to interstitial glucose kinetics; and a model of the deterioration in time of sensor accuracy, together with (if available) a priori information (e.g. probabilistic distribution) on the parameters of the model. The algorithm first checks for the presence of possible artifacts or outliers on both CGM datastream and BG references, and then rescales the CGM time series by exploiting a retrospective calibration approach based on a regularized deconvolution method subject to the constraint of returning a profile laying within the confidence interval of the reference BG measurements. As output, the retrofitting algorithm produces an improved “retrofitted” quasi-continuous glucose concentration signal that is better (in terms of both accuracy and precision) than the CGM trace originally measured by the sensor. In clinical trials, the so-obtained retrofitted traces can be used to calculate solid outcome measures, avoiding the need of increasing the data collection burden at the patient level.

Retrospective retrofitting method to generate a continuous glucose concentration profile by exploiting continuous glucose monitoring sensor data and blood glucose measurements

A device for generating alerts for Hypo and Hyperglycemia Prevention from Continuous Glucose Monitoring (CGM) determines a dynamic risk based on both information of glucose level and a trend obtainable from a CGM signals. The device includes a display whose color depends on the DR (for example, red for high DR, green for low risk). When DR exceeds a certain threshold, alerts are generated to suggest the patient to pay attention to the current glucose reading and to its trend, both of which are shown on the display in numbers and symbols (e.g. an arrow with different slope or color).

Alert system for hypo and hyperglycemia prevention based on clinical risk

Individualized multiple-day simulation model of type 1 diabetic patient decision-making for developing, testing and optimizing insulin therapies driven by glucose sensors

Systems and methods for providing sensitive and specific alarms indicative of glycemic condition are provided herein. In an embodiment, a method of processing sensor data by a continuous analyte sensor includes: evaluating sensor data using a first function to determine whether a real time glucose value meets a first threshold; evaluating sensor data using a second function to determine whether a predicted glucose value meets a second threshold; activating a hypoglycemic indicator if either the first threshold is met or if the second threshold is predicted to be met; and providing an output based on the activated hypoglycemic indicator.

Systems and methods for dynamically and intelligently monitoring a host's glycemic condition after an alert is triggered

Continuous Glucose Monitoring (CGM) devices provide glucose concentration measurements in the subcutaneous tissue with limited accuracy and precision. Therefore, CGM readings cannot be incorporated in a straightforward manner in outcome metrics of clinical trials e.g. aimed to assess new glycaemic-regulation therapies. To define those outcome metrics, frequent Blood Glucose (BG) reference measurements are still needed, with consequent relevant difficulties in outpatient settings. Here we propose a 'retrofitting' algorithm that produces a quasi continuous time BG profile by simultaneously exploiting the high accuracy of available BG references (possibly very sparsely collected) and the high temporal resolution of CGM data (usually noisy and affected by significant bias). The inputs of the algorithm are: a CGM time series; some reference BG measurements; a model of blood to interstitial glucose kinetics; and a model of the deterioration in time of sensor accuracy, together with (if available) a priori information (e.g. probabilistic distribution) on the parameters of the model. The algorithm first checks for the presence of possible artifacts or outliers on both CGM datastream and BG references, and then rescales the CGM time series by exploiting a retrospective calibration approach based on a regularized deconvolution method subject to the constraint of returning a profile laying within the confidence interval of the reference BG measurements. As output, the retrofitting algorithm produces an improved 'retrofitted' quasi-continuous glucose concentration signal that is better (in terms of both accuracy and precision) than the CGM trace originally measured by the sensor. In clinical trials, the so-obtained retrofitted traces can be used to calculate solid outcome measures, avoiding the need of increasing the data collection burden at the patient level.

A mathematical model of type 1 diabetes (T1D) patient decision-making can be used to simulate, in silico, realistic glucose/insulin dynamics, for several days, in a variety of subjects who take therapeutic actions (e.g. insulin dosing) driven by either self-monitoring blood glucose (SMBG) or continuous glucose monitoring (CGM). The decision-making (DM) model can simulate real-life situations and everyday patient behaviors, Accurate submodels of SMBG and CGM measurement errors are incorporated in the comprehensive DM model. The DM model accounts for common errors the patients are used to doing in their diabetes management, such as miscalculations of meal carbohydrate content, early/delayed insulin administrations and missed insulin boluses. The DM model can be used to assess in silico if/when CGM can safely substitute SMBG in T1D management, to develop and test guidelines for CGM driven insulin dosing, to optimize and individualize off-line insulin therapies and to develop and test decision support systems.

A method, system, and device for improving the accuracy of a continuous glucose monitoring sensor by estimating a CGM signal at a time t+PH using a value of CGM at time t, using a real-time short-time glucose prediction horizon to estimate the real time denoised CGM value with a noise estimation algorithm.

Growing community of inventors

Padua, Italy

Giovanni Sparacino