Perfoming analytics on protected health information
First Claim
1. A method for analyzing patient data, the method comprising:
- accessing, by a computer system, one or more databases comprising health information for a plurality of patients, the health information including protected health information;
accessing, by the computer system, an analytical model;
isolating, by the computer system, the protected health information from a user interface to restrict user access to the protected health information;
receiving, by the computer system via the user interface, instructions to apply the analytical model to health information for each of the plurality of patients;
applying, by the computer system, the analytical model using a machine-learning algorithm to health information for each of the plurality of patients to train the analytical model with the health information from the plurality of patients, the trained analytical model comprising trained weights of the analytical model, wherein the health information used to train the analytical model includes the protected health information isolated from the user interface; and
storing, by the computer system, a result comprising the trained weights of the analytical model trained with the health information to the one or more databases;
presenting, by the computer system via the user interface, a summary of the result of the analytical model, wherein the summary of the result of the analytical model excludes any of the protected health information.
2 Assignments
0 Petitions
Accused Products
Abstract
This disclosure includes techniques for analyzing patient data. In one example, a method includes accessing, by a computer system, one or more databases comprising health information, with protected health information, for a plurality of patients, accessing, by the computer system, an analytical model, and receiving, by the computer system via a user interface, instructions to apply the analytical model to health information for each of the plurality of patients. The protected health information is isolated from the user interface to restrict access to the protected health information. The method further includes applying, by the computer system, the analytical model to health information for each of the plurality of patients, and storing a result of the analytical model to the one or more databases.
26 Citations
21 Claims
-
1. A method for analyzing patient data, the method comprising:
-
accessing, by a computer system, one or more databases comprising health information for a plurality of patients, the health information including protected health information; accessing, by the computer system, an analytical model; isolating, by the computer system, the protected health information from a user interface to restrict user access to the protected health information; receiving, by the computer system via the user interface, instructions to apply the analytical model to health information for each of the plurality of patients; applying, by the computer system, the analytical model using a machine-learning algorithm to health information for each of the plurality of patients to train the analytical model with the health information from the plurality of patients, the trained analytical model comprising trained weights of the analytical model, wherein the health information used to train the analytical model includes the protected health information isolated from the user interface; and storing, by the computer system, a result comprising the trained weights of the analytical model trained with the health information to the one or more databases;
presenting, by the computer system via the user interface, a summary of the result of the analytical model, wherein the summary of the result of the analytical model excludes any of the protected health information. - View Dependent Claims (2, 3, 4, 5, 6, 7)
-
-
8. A computer system for storing and analyzing health information for a plurality of patients comprising:
-
one or more databases comprising health information for a plurality of patients, the health information including protected health information; a user interface; and one or more processors configured to; receive, via the user interface, instructions to apply an analytical model to health information for each of the plurality of patients; isolate the protected health information from the user interface to restrict user access to the protected health information; apply, based on instructions received via the user interface, the analytical model using a machine-learning algorithm to patient data for each of the plurality of patients to train the analytical model with the health information from the plurality of patients, the trained analytical model comprising trained weights of the analytical model, wherein the health information used to train the analytical model includes the protected health information isolated from the user interface; and storing a result comprising the trained weights of the analytical model trained with the health information to the one or more databases;
wherein the one or more processors are further configured to present, via the user interface, a summary of the updated health information, wherein the summary of the updated health information excludes any of the protected health information. - View Dependent Claims (9, 10, 11, 12, 13, 14)
-
-
15. A non-transitory computer-readable storage medium comprising instructions that, when executed, cause one or more processors to:
-
access one or more databases comprising health information for a plurality of patients, the health information including protected health information; isolate the protected health information from a user interface to restrict user access to the protected health information; receive, via the user interface, instructions to apply an analytical model to health information for each of the plurality of patients; apply the analytical model using a machine-learning algorithm to health information for each of the plurality of patients to train the analytical model with the health information from the plurality of patients, the trained analytical model comprising trained weights of the analytical model, wherein the health information used to train the analytical model includes the protected health information isolated from the user interface; and store a result comprising the trained weights of the analytical model trained with the health information to the one or more databases;
present, via the user interface, a summary of the updated health information, wherein the summary excludes any of the protected health information. - View Dependent Claims (16, 17, 18, 19, 20, 21)
-
Specification