US 12,412,664 B2
Addiction predictor and relapse detection support tool
Richard Matthew Balian, Birdsboro, PA (US)
Assigned to Cerner Innovation, Inc., Kansas City, MO (US)
Filed by CERNER INNOVATION, INC., Kansas City, KS (US)
Filed on Jun. 25, 2019, as Appl. No. 16/452,220.
Prior Publication US 2020/0411191 A1, Dec. 31, 2020
Int. Cl. G06F 16/00 (2019.01); G06F 16/245 (2019.01); G06N 5/02 (2023.01); G16H 10/60 (2018.01); G16H 50/30 (2018.01)
CPC G16H 50/30 (2018.01) [G06F 16/245 (2019.01); G06N 5/02 (2013.01); G16H 10/60 (2018.01)] 25 Claims
OG exemplary drawing
 
1. A method for automatically predicting a likelihood of addiction for an individual based on a predictive model, the method comprising:
accessing a predictive model having a plurality of input variables;
retrieving a plurality of sets of structured data for the individual via at least one application programing interface, wherein each set of structured data of the plurality of sets of structured data is retrieved from a remote disparate database, wherein at least one set of structured data of the plurality of sets of structured data is received from an academic record server, an employment record server, or a financial record server, wherein the academic record server maintains and provides access to one or more academic databases containing school records for the individual;
extracting from each retrieved set of structured data a set of input variables;
automatically standardizing the set of input variables, using a pre-model transformation to generate standardized input variables;
utilizing the predictive model and the standardized input variables to determine a prediction for the likelihood of addiction for the individual, wherein the prediction for the likelihood of addiction for the individual is based on the at least one set of structured data of the plurality of sets of structured data received from the academic record server, the employment record server, or the financial record server;
detecting an access request for an electronic health record (EHR) of the individual by a caregiver via a computing device networked with a remote health record server maintaining the individual's EHR; and
in response to detecting the access request for the individual's electronic health record (EHR), providing the determined likelihood of addiction for presentation to the caregiver via the computing device, wherein in response to the likelihood of addiction for the individual exceeding a risk threshold an automatically generated alert is presented coexistent with presentation of the determined likelihood of addiction, the automatically generated alert preventing execution of the access request until an input is received via the computing device that corresponds to acknowledgment of the caregiver of the automatically generated alert.