Cognitive systems for allocating medical data access permissions using historical correlations

• US 10,691,827 B2
• Filed: 12/18/2017
• Issued: 06/23/2020
• Est. Priority Date: 12/18/2017
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Briscoe, Iyanna, Harrington, Steven L., King, Tyler N., Piccone, Newton J.
• Primary Examiner(s): Homayounmehr, Farid
• Assistant Examiner(s): Salman, Raied A
• Application Number:
  US15/844,869
• Publication Number:
  US 20190188410A1
• Time in Patent Office: 918 Days
• CPC Class Codes: G06F 16/243 Natural language query formulation
• CPC Class Codes: G06F 21/604 Tools and structures for managing or administering access control systems
• CPC Class Codes: G06F 21/6245 Protecting personal data, e.g. for financial or medical purposes
• CPC Class Codes: G06F 40/20 Natural language analysis semantic analysis of natural language G06F40/30
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/063 using electronic means
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

Embodiments of the present invention provide a computer-implemented method for allocating medical data access permissions using historical correlations. The method receives a request for medical research data from a user. The method executes natural language processing to interpret the received request. The method generates a database query based on a trained model to identify a medical research data set that is responsive to the request. The method queries one or more medical databases using the generated query to identify the medical research data set that is responsive to the request. The medical research data set that is identified includes the requested medical research data as well as additional medical research data that, although not requested by the user, is found to be relevant to the request based on the trained model. The method includes transmitting the medical research data set to the user.

Claims

1. A computer-implemented method for allocating medical data access permissions using historical correlations, the computer-implemented method comprising:

receiving, by a computing system comprising one or more processors, a request for first medical research data from a first user;

executing, by the computing system, natural language processing to interpret the request;

generating, by the computing system, a database query based on a trained model to identify a first medical research data set that is responsive to the request, wherein the trained model takes into consideration (a) results of the natural language processing of the received request and (b) prior medical research data requests and responses that are stored in a historical database;

querying, by the computing system, one or more medical databases using the query to identify the first medical research data set that is responsive to the request, wherein the first medical research data set that is identified by the querying includes (a) the first medical research data and (b) second medical research data that is identified by the computing system as being relevant to the request based on the trained model, wherein the second medical research data was not requested by the first user in the request;

transmitting, by the computing system, the first medical research data set to the first user;

logging, by the computing system, the request and the first medical research data set into an audit log; and

repeating the receiving, the executing, the generating, the querying, and the transmitting in response to a second request from the first user for third medical research data to identify a second medical research data set that includes the third medical research data;

establishing a statistical correlation score between the first and second medical research data sets; and

upon the established correlation score exceeding a user-defined confidence interval, logging the established correlation score in the audit log and updating the trained model based on the established correlation score.

2. The computer-implemented method of claim 1, wherein the model is trained by at least:

retrieving the prior medical research data requests and responses that are stored in the historical database; and

training one or more neural networks based on the prior medical research data requests and responses;

wherein the second medical research data that is identified during the querying of the one or more medical databases is identified by at least scoring, during the querying, each medical file stored on the one or more medical databases with a likelihood of being relevant to the query based on the trained model.

3. The computer-implemented method of claim 2, wherein the prior medical research data requests and responses include (a) a plurality of prior requests that were sent from one or more users and (b) a plurality of prior database query results that were generated by the computing system in response to processing of the plurality of prior requests.

4. The computer-implemented method of claim 1, further comprising:

subsequent to transmitting the second medical research data set to the first user, determining, by the computing system, whether a renewal request for the second medical research data set was received at the computing system from the first user during a first period of time; and

upon determining that a renewal request for use of the second medical research data set was not received, reducing the established correlation score in the audit log and updating the trained model based on the reduced correlation score.

5. The computer-implemented method of claim 1, wherein prior to receiving the request at the computing system, the request is transmitted from the first user to a server of an internal review board (IRB) of an institution for approval, wherein the request is then received by the computer system from the server of IRB in response to the IRB approving the request via the server.

6. The computer-implemented method of claim 5, further comprising:

receiving, by the computing system, a security profile pertaining to the first user, wherein the security profiled is received by the computing system from the server of the IRB; and

executing, by the computing system, natural language processing to interpret the security profile;

wherein the database query is generated further based on results of the natural language processing of the security profile.

7. A system for allocating medical data access permissions using historical correlations, the system comprising one or more hardware processors configured to perform a method comprising:

receiving, by the system, a request for first medical research data from a first user;

executing, by the system, natural language processing to interpret the request;

generating, by the system, a database query based on a trained model to identify a first medical research data set that is responsive to the request, wherein the trained model takes into consideration (a) results of the natural language processing of the received request and (b) prior medical research data requests and responses that are stored in a historical database;

querying, by the system, one or more medical databases using the query to identify the first medical research data set that is responsive to the request, wherein the first medical research data set that is identified by the querying of the one or more medical databases includes (a) the first medical research data and (b) second medical research data that is identified by the system as being relevant to the request based on the trained model, wherein the second medical research data was not requested by the first user in the request;

transmitting, by the system, the first medical research data set to the first user;

logging, by the system, the request and the first medical research data set into an audit log; and

repeating the receiving, the executing, the generating, the querying, and the transmitting in response to a second request from the first user for third medical research data to identify a second medical research data set that includes the third medical research data;

establishing a statistical correlation score between the first and second medical research data sets; and

upon the established correlation score exceeding a user-defined confidence interval, logging the established correlation score in the audit log and updating the trained model based on the established correlation score.

8. The system of claim 7, wherein the model is trained by at least:

retrieving the prior medical research data requests and responses that are stored in the historical database; and

training one or more neural networks based on the prior medical research data requests and responses;

wherein the second medical research data that is identified during the querying of the one or more medical databases is identified by at least scoring, during the querying, each medical file stored on the one or more medical databases with a likelihood of being relevant to the query based on the trained model.

9. The system of claim 8, wherein the prior medical research data requests and responses include (a) a plurality of prior requests that were sent from one or more users and (b) a plurality of prior database query results that were generated by the system in response to processing of the plurality of prior requests.

10. The system of claim 7, wherein the method further comprises:

subsequent to transmitting the second medical research data set to the first user, determining, by the system, whether a renewal request for the second medical research data set was received by the system from the first user during a first period of time; and

upon determining that a renewal request for use of the second medical research data set was not received, reducing the established correlation score in the audit log and updating the trained model based on the reduced correlation score.

11. The system of claim 7, wherein prior to receiving the request at the system, the request is transmitted from the first user to a server of an internal review board (IRB) of an institution for approval, wherein the request is then received by the system from the server of IRB in response to the IRB approving the request via the server.

12. The system of claim 11, wherein the method further comprises:

receiving, by the system, a security profile pertaining to the first user, wherein the security profiled is received by the system from the server of the IRB; and

executing, by the system, natural language processing to interpret the security profile;

wherein the database query is generated further based on results of the natural language processing of the security profile.

13. A computer program product for allocating medical data access permissions using historical correlations, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a system comprising one or more processors to cause the system to perform a method comprising:

receiving, by the system, a request for first medical research data from a first user;

executing, by the system, natural language processing to interpret the received request;

generating, by the system, a database query based on a trained model to identify a first medical research data set that is responsive to the request, wherein the trained model takes into consideration (a) results of the natural language processing of the received request and (b) prior medical research data requests and responses that are stored in a historical database;

querying, by the system, one or more medical databases using the generated query to identify the first medical research data set that is responsive to the request, wherein the first medical research data set that is identified by the querying of the one or more medical databases includes (a) the first medical research data and (b) second medical research data that is identified by the system as being relevant to the request based on the trained model, wherein the second medical research data was not requested by the first user in the request;

transmitting, by the system, the first medical research data set to the first user;

logging, by the system, the request and the first medical research data set into an audit log; and

repeating the receiving, the executing, the generating, the querying, and the transmitting in response to a second request from the first user for third medical research data to identify a second medical research data set that includes the third medical research data;

establishing a statistical correlation score between the first and second medical research data sets; and

upon the established correlation score exceeding a user-defined confidence interval, logging the established correlation score in the audit log and updating the trained model based on the established correlation score.

14. The computer program product of claim 13, wherein the model is trained by at least:

retrieving the prior medical research data requests and responses that are stored in the historical database; and

training one or more neural networks based on the prior medical research data requests and responses;

wherein the second medical research data that is identified during the querying of the one or more medical databases is identified by at least scoring, during the querying, each medical file stored on the one or more medical databases with a likelihood of being relevant to the query based on the trained model.

15. The computer program product of claim 13, wherein the method further comprises:

subsequent to transmitting the second medical research data set to the first user, determining, by the system, whether a renewal request for the second medical research data set was received from the first user during a first period of time; and

upon determining that a renewal request for use of the second medical research data set was not received, reducing the established correlation score in the audit log and updating the trained model based on the reduced correlation score.

16. The computer program product of claim 13, wherein prior to receiving the request at the system, the request is transmitted from the first user to a server of an internal review board (IRB) of an institution for approval, wherein the request is then received by the computer from the server of IRB in response to the IRB approving the request via the server.

17. The computer program product of claim 16, wherein the method further comprises:

receiving, by the system, a security profile pertaining to the first user, wherein the security profiled is received by the system from the server of the IRB; and

executing, by the system, natural language processing to interpret the security profile;

wherein the database query is generated further based on results of the natural language processing of the security profile.