Joint state estimation prediction that evaluates differences in predicted vs. corresponding received data

Abstract

Systems and methods are provided for reconciling untrusted data of a subject using trusted data pertaining to the subject. Systems and methods are directed to evaluating differences in predicted data with respect to corresponding received data. Systems and methods estimate metabolic states from a combination of trusted and untrusted metabolic inputs, along with optionally using a personalized mathematical model with parameter optimization. Systems and methods provide for reconciled untrusted inputs with their measured impact of the glycemic signals that is consistent with a metabolic model. Estimation of future metabolic states for decision support and automated insulin dosing is enabled. Replay of scenarios with estimated or reconciled data is also provided.

What is claimed is: 
     
         1 . A method comprising:
 predicting data over a time period for a subject;   receiving untrusted data directed to management of diabetes;   simulating a plurality of predictive data traces over the time period using a spectrum of possible variances of the untrusted data;   comparing the simulated predictive data traces to the predicted data to identify glycemic effects; and   outputting a visualization or a recommendation based on the glycemic effects.   
     
     
         2 . The method of  claim 1 , wherein the untrusted data comprises glucose data, and wherein the predictive data traces comprise predictive glucose traces. 
     
     
         3 . The method of  claim 2 , wherein predicting the glucose data is based on trusted CGM data and an individualized model of the glucose-insulin kinetics of the subject. 
     
     
         4 . The method of  claim 2 , wherein the predicting the glucose data comprises providing a best estimate glucose trace representing glucose state over time. 
     
     
         5 . The method of  claim 1 , wherein the untrusted data directed to management of diabetes comprises at least one of a timing of insulin, an amount of insulin, meal data, or activity data. 
     
     
         6 . The method of  claim 1 , wherein the glycemic effects are associated with differences in at least one of an amount of diabetes management data or a timing of diabetes management data. 
     
     
         7 . A method comprising:
 receiving alternative metabolic inputs, reconciled estimated untrusted metabolic inputs, trusted metabolic inputs, and final estimated metabolic inputs;   performing real time prediction using the alternative metabolic inputs, the reconciled estimated untrusted metabolic inputs, the trusted metabolic inputs, and the final estimated metabolic inputs; and   outputting predicted metabolic states based on the real time prediction.   
     
     
         8 . The method of  claim 7 , wherein performing the real time prediction comprises:
 extrapolating time series for the reconciled estimated untrusted metabolic inputs and the trusted metabolic inputs; and   estimating metabolic states into the future using the extrapolated time series, the alternative metabolic inputs, and the final estimated metabolic states.   
     
     
         9 . The method of  claim 8 , further comprising filtering the extrapolated time series to prevent jitter in the predicted metabolic states. 
     
     
         10 . The method of  claim 8 , wherein the estimated metabolic states are in time series form. 
     
     
         11 . The method of  claim 10 , further comprising filtering the estimated metabolic states to generate the predicted metabolic states. 
     
     
         12 . The method of  claim 8 , wherein the estimating the metabolic states uses a behavior model of a subject. 
     
     
         13 . The method of  claim 8 , wherein extrapolating the time series uses weighting of historical data based on at least one of a time of day, features of a current estimated state, or a database of past metabolic inputs. 
     
     
         14 . A system comprising:
 at least one processor; and   a non-transitory computer readable medium comprising instructions that, when executed by the at least one processor, cause the system to:
 predict data over a time period for a subject; 
 receive untrusted data directed to management of diabetes; 
 simulate a plurality of predictive data traces over the time period using a spectrum of possible variances of the untrusted data; 
 compare the simulated predictive data traces to the predicted data to identify glycemic effects; and 
 output a visualization or a recommendation based on the glycemic effects. 
   
     
     
         15 . The system of  claim 14 , wherein the untrusted data comprises glucose data, and wherein the predictive data traces comprise predictive glucose traces. 
     
     
         16 . The system of  claim 15 , wherein predicting the glucose data is based on trusted CGM data and an individualized model of the glucose-insulin kinetics of the subject. 
     
     
         17 . The system of  claim 15 , wherein the predicting the glucose data comprises providing a best estimate glucose trace representing glucose state over time. 
     
     
         18 . The system of  claim 14 , wherein the untrusted data directed to management of diabetes comprises at least one of a timing of insulin, an amount of insulin, meal data, or activity data. 
     
     
         19 . The system of  claim 14 , wherein the glycemic effects are associated with differences in at least one of an amount of diabetes management data or a timing of diabetes management data. 
     
     
         20 . A system comprising:
 at least one processor; and   a non-transitory computer readable medium comprising instructions that, when executed by the at least one processor, cause the system to:
 receive alternative metabolic inputs, reconciled estimated untrusted metabolic inputs, trusted metabolic inputs, and final estimated metabolic inputs; 
 perform real time prediction using the alternative metabolic inputs, the reconciled estimated untrusted metabolic inputs, the trusted metabolic inputs, and the final estimated metabolic inputs; and 
 output predicted metabolic states based on the real time prediction. 
   
     
     
         21 . The system of  claim 20 , wherein performing the real time prediction comprises:
 extrapolating time series for the reconciled estimated untrusted metabolic inputs and the trusted metabolic inputs; and   estimating metabolic states into the future using the extrapolated time series, the alternative metabolic inputs, and the final estimated metabolic states.   
     
     
         22 . The system of  claim 21 , further comprising instructions that, when executed by the at least one processor, cause the system to filter the extrapolated time series to prevent jitter in the predicted metabolic states. 
     
     
         23 . The system of  claim 21 , wherein the estimated metabolic states are in time series form. 
     
     
         24 . The system of  claim 23 , further comprising instructions that, when executed by the at least one processor, cause the system to filter the estimated metabolic states to generate the predicted metabolic states.