Granted Patent B2

US 12,451,237 · App. 17/354,309 · Granted Oct 21, 2025

Inventors: Serrah Lynn Linares (Gig Harbor, WA); Gautam Shah (San Francisco, CA)

Assignee: Change Healthcare Holdings, LLC

G16H40/20G06N20/00G16H10/60G16H40/67G06Q10/1095G16H10/20

Abstract

Claims (49)

1. A computer-implemented method comprising:

parsing, by one or more processors and via a chatbot front-end of a virtual triage system that interfaces with a user, a natural language input using a first machine learning model that is trained with transcripts from historical interactions between agents and users and with medical knowledge bases to derive a parsed natural language input;

determining, by the one or more processors and via a chatbot back-end of the virtual triage system that (1) is communicatively coupled to the chatbot front-end and (2) receives the parsed natural language input, that the user is authenticated for access to the virtual triage system based on the parsed natural language input and at least one of an Internet Protocol (IP) address, a phone number, or a media access control (MAC) address associated with a computing device used by the user to access the virtual triage system;

accessing, by the one or more processors and via the chatbot back-end and in response to determining that the user is authenticated, historical health data associated with the user, wherein the historical health data includes a medical history of the user and identifiers of medications taken by the user;

generating, by the one or more processors and via the chatbot back-end based on inputting the parsed natural language input and the historical health data into a second machine learning model that is trained with the medical knowledge bases, a recommendation for the user;

controlling, by the one or more processors and via the chatbot back-end, a type of additional information to be collected from the user by the chatbot front-end based on a type of the recommendation; and

providing, by the one or more processors and based on the recommendation, a programmatic entry directly to a scheduling service of a healthcare provider,

wherein the first machine learning model and the second machine learning model are trained by a common service of the virtual triage system.

2. The computer-implemented method of claim 1 , wherein the type of the recommendation comprises a deferred care path or an urgent care path.

3. The computer-implemented method of claim 1 , wherein the recommendation comprises advising the user to seek care via the healthcare provider.

4. The computer-implemented method of claim 1 , further comprising selecting, by the one or more processors, the healthcare provider based on the parsed natural language input and the historical health data.

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

determining, by the one or more processors, available appointment times with the healthcare provider;

presenting, by the one or more processors, one or more available appointment times, of the available appointment times via the chatbot front-end;

receiving, by the one or more processors, a selected appointment time of the one or more available appointment times presented via the chatbot front-end; and

scheduling, by the one or more processors and using the scheduling service, the selected appointment time with the healthcare provider for the user.

6. The computer-implemented method of claim 1 , further comprising selecting, by the one or more processors, the healthcare provider based on patient healthcare provider preferences of the user and the recommendation.

7. The computer-implemented method of claim 1 , further comprising training, by the one or more processors, the first machine learning model and the second machine learning model using physician reviewed training data.

8. The computer-implemented method of claim 1 , further comprising conducting, by the one or more processors, a structured health and physical assessment of the user; and generating, by the chatbot back-end based on applying the-second machine learning model to (a) the parsed natural language input, (b) the structured health and physical assessment, and (c) the historical health data, the recommendation for the user.

9. A system comprising:

one or more processors; and

one or more memories storing processor-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:

parsing, via a chatbot front-end of a virtual triage system that interfaces with a user, a natural language input using a first machine learning model that is trained with transcripts from historical interactions between agents and users and with medical knowledge bases to derive a parsed natural language input;

determining, via a chatbot back-end of the virtual triage system that (1) is communicatively coupled to the chatbot front-end and (2) receives the parsed natural language input, that the user is authenticated for access to the virtual triage system based on the parsed natural language input and at least one of an Internet Protocol (IP) address, a phone number, or a media access control (MAC) address associated with a computing device used by the user to access the virtual triage system;

accessing, via the chatbot back-end and in response to determining that the user is authenticated, historical health data associated with the user, wherein the historical health data includes a medical history of the user and identifiers of medications taken by the user;

generating, via the chatbot back-end based on inputting the parsed natural language input and the historical health data into a second machine learning model that is trained with the medical knowledge bases, a recommendation for the user;

controlling, via the chatbot back-end, a type of additional information to be collected from the user by the chatbot front-end based on a type of the recommendation; and

providing, based on the recommendation, a programmatic entry directly to a scheduling service of a healthcare provider,

wherein the first machine learning model and the second machine learning model are trained by a common service of the virtual triage system.

10. The system of claim 9 , wherein the type of the recommendation comprises a deferred care path or an urgent care path.

11. The system of claim 9 , wherein the recommendation comprises advising the user to seek care via the healthcare provider.

12. The system of claim 9 , further comprising processor-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising: selecting the healthcare provider based on the parsed natural language input and the historical health data.

13. The system of claim 9 , further comprising processor-executable instructions that when executed by the one or more processors cause the one or more processors to perform operations comprising:

determining available appointment times with the healthcare provider;

presenting one or more available appointment times, of the available appointment times via the chatbot front-end;

receiving a selected appointment time of the one or more available appointment times presented via the chatbot front-end; and

scheduling, using the scheduling service, the selected appointment time with the healthcare provider for the user.

14. The system of claim 9 , further comprising processor-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising: selecting the healthcare provider based on patient healthcare provider preferences of the user and the recommendation.

15. The system of claim 9 , further comprising processor-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising: training the first machine learning model and the second machine learning model using physician reviewed training data.

16. The system of claim 9 , further comprising processor-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:

conducting a structured health and physical assessment of the user; and generating, by the chatbot back-end based on applying the second machine learning model to (a) the parsed natural language input, (b) the structured health and physical assessment, and (c) the historical health data, the recommendation for the user.

17. One or more non-transitory computer-readable media storing processor-executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:

parsing, via a chatbot front-end of a virtual triage system that interfaces with a user, a natural language input using a first machine learning model that is trained with transcripts from historical interactions between agents and users and with medical knowledge bases to derive a parsed natural language input;

determining, via a chatbot back-end of the virtual triage system that (1) is communicatively coupled to the chatbot front-end and (2) receives the parsed natural language input, that the user is authenticated for access to the virtual triage system based on the parsed natural language input and at least one of an Internet Protocol (IP) address, a phone number, or a media access control (MAC) address associated with a computing device used by the user to access the virtual triage system;

accessing, via the chatbot back-end and in response to determining that the user is authenticated, historical health data associated with the user, wherein the historical health data includes a medical history of the user and identifiers of medications taken by the user;

generating, via the chatbot back-end based on inputting the parsed natural language input and the historical health data into a second machine learning model that is trained with the medical knowledge bases, a recommendation for the user;

controlling, via the chatbot back-end, a type of additional information to be collected from the user by the chatbot front-end based on a type of the recommendation; and

providing, based on the recommendation, a programmatic entry directly to a scheduling service of a healthcare provider,

wherein the first machine learning model and the second machine learning model are trained by a common service of the virtual triage system.

Assignments (1)

Continuity (1)

References Cited (18)