Method for modeling behavior and health changes

Abstract

One method for supporting a patient through a treatment regimen includes: accessing a log of use of a native communication application executing on a mobile computing device by a patient; selecting a subgroup of a patient population based on the log of use of the native communication application and a communication behavior common to the subgroup; retrieving a regimen adherence model associated with the subgroup, the regimen adherence model defining a correlation between treatment regimen adherence and communication behavior for patients within the subgroup; predicting patient adherence to the treatment regimen based on the log of use of the native communication application and the regimen adherence model; and presenting a treatment-related notification based on the patient adherence through the mobile computing device.

We claim: 
     
         1 . A method for improving mental health risk determination for a user, the method comprising:
 receiving a dataset from a user device associated with the user;   deriving a set of features associated with the user based at least in part on the dataset;   selecting a mental health risk model for the user based on the set of features;   evaluating the mental health risk model to determine a set of mental health risk parameters for the user;   determining a mental health risk state for the user based on the set of mental health risk parameters; and   in response to and based on the set of mental health risk parameters, initiating provision of an intervention at the user device, the intervention operable to improve the mental health risk state of the user.   
     
     
         2 . The method of  claim 1 , wherein the mental health risk model is determined based on aggregated data from a subgroup of multiple users, wherein the subgroup of multiple users is associated with the set of features. 
     
     
         3 . The method of  claim 2 , wherein the mental health risk model is selected from a set of multiple health risk models. 
     
     
         4 . The method of  claim 2 , wherein the mental health risk model is selected in response to assigning the user to the subgroup of multiple users based on the set of features. 
     
     
         5 . The method of  claim 1 , wherein the set of features is further determined based on demographic information associated with the user. 
     
     
         6 . The method of  claim 1 , wherein the dataset comprises passive information, wherein the passive information is collected passively from the user device. 
     
     
         7 . The method of  claim 6 , wherein the passive information is collected passively from a communication application executing on the user device. 
     
     
         8 . The method of  claim 6 , wherein the dataset further comprises active information, wherein the active information is entered by the user at the user device. 
     
     
         9 . The method of  claim 8 , wherein the active information is collected from a set of surveys administered to the user at the user device. 
     
     
         10 . The method of  claim 1 , wherein the set of mental health risk parameters for the user is further determined based on the dataset. 
     
     
         11 . The method of  claim 1 , further comprising, receiving a second dataset from the user device in response to the intervention, and updating the mental health risk model based on the second dataset. 
     
     
         12 . The method of Claim ii, wherein the second dataset comprises a survey dataset from a set of surveys administered to the user. 
     
     
         13 . The method of  claim 1 , further comprising:
 receiving a second dataset from a second user device associated with a second user;   deriving a second set of features associated with the second user based at least in part on the second dataset;   detecting an overlap between the second set of features and the first set of features; and   selecting the mental health risk model for the second user based on the overlap.   
     
     
         14 . The method of  claim 1 , wherein initiating provision of the intervention comprises facilitating a communication between the user and a care provider. 
     
     
         15 . The method of  claim 1 , wherein the mental health risk model defines a correlation between the dataset and the mental health risk state of the user. 
     
     
         16 . The method of  claim 1 , wherein the dataset comprises at least one of: a communication dataset, a location dataset, and a motion dataset. 
     
     
         17 . The method of  claim 1 , wherein the mental health risk model is a trained machine learning model, wherein the mental health risk model is trained based on data from an aggregated set of users, the aggregated set of users having a shared subset of the set of features of the user. 
     
     
         18 . The method of  claim 1 , wherein the set of features is derived with a factor analysis approach. 
     
     
         19 . The method of  claim 1 , wherein the intervention comprises an infographic notification displayed at the user device, wherein initiating provision of the intervention comprises automatically initiating provision of the infographic notification. 
     
     
         20 . The method of  claim 1 , wherein determining the mental health risk state for the user comprises comparing the set of mental health risk parameters with a set of thresholds.