Systems and methods for an artificial intelligence system
AssigneeHealthStream, Inc.
InventorRobert Frist, Jr.
Life of this patent
Estimating the term…
Abstract
A method is disclosed, which may include generating, in a natural language processing (NLP) system, a plurality of entity data objects. The method may include generating, in the NLP system, a plurality of activity data objects. The method may include generating, on at least one server, an evaluation data object. The evaluation data object may include a problem data object, an observation data object, or an action data object. The method may include configuring each problem data object, observation data object, or action data object of the evaluation data object with a scoring rubric. Other methods, systems, and computer-readable media are also disclosed.
Claims20
A system configured to provide a computer-implemented medical training simulation, comprising:
- at least one user device comprising a computing device;
- a server;
- a natural language processing (NLP) system comprising at least one computing device;
- at least one processor, wherein the at least one user device, the at least one processor, the server and the NLP system are configured for at least one of wired or wireless communication over a data network;
- at least one computer-readable storage medium including at least one memory storing one or more instructions, wherein the at least one computer processor, in response to executing the one or more instructions, is configured to:
- receive a plurality of user responses from a user interface of the computing device, wherein the user interface includes a text-chat interface displaying a conversation between multiple parties, wherein switching the user interface between the text-chat interface and a medical scenario information user interface is achieved via at least one of swipe motion and gestures, wherein the plurality of user responses includes an action response, a problem response, and an observation response;
- send the plurality of responses to the NLP system that is configured to output a computer-readable parsing;
- receive the computer-readable parsing, wherein the computer-readable parsing includes a first activity key, and wherein the first activity key includes a first text string indicating at least one of administering, notifying, assisting, or monitoring;
- provide an evaluation data object, wherein the evaluation data object includes an action data object, wherein the action data object includes a second activity key and an action scoring rubric, the action scoring rubric comprising a first value and a second value;
- map the first activity key to the second activity key by determining whether the first activity key includes the second activity key;
- and
- calculate a first sub-score based on the action scoring rubric and whether the first activity key includes the second activity key, wherein the first sub-score indicates a clinical judgment evaluation of a user, wherein calculating the first-sub score comprises:
- responsive to a determination that the first activity key includes the second activity key, including the first value in the first-sub score;
- and
- responsive to a determination that the first activity key does not include the second activity key, including the second value in the first-sub score.
A system configured to provide a computer-implemented medical training simulation, comprising:
- at least one user device comprising a computing device;
- a server;
- a natural language processing (NLP) system comprising at least one computing device;
- at least one processor;
- at least one computer-readable storage medium including at least one memory storing one or more instructions, wherein the at least one computer processor, in response to executing the one or more instructions, is configured to:
- present, via a user interface of the computing device, a user with medical scenario information, wherein the user interface includes a text-chat interface displaying a conversation between multiple parties, wherein switching the user interface between the text-chat interface and a medical scenario information user interface is achieved via at least one of swipe motion and gestures;
- receive a plurality of user responses from the user, wherein the plurality of user responses includes an action response;
- send the plurality of responses to the NLP system that is configured to output a computer-readable parsing;
- receive the computer-readable parsing, wherein the computer-readable parsing includes a first entity key and a first activity key, the first entity key includes a first text string and the first activity key includes a second text string;
- provide an evaluation data object, wherein the evaluation data object includes an action data object, wherein the action data object includes a second entity key, a second activity key, and an action scoring rubric, wherein the action scoring rubric comprises a first value and a second value;
- map the first entity key to the second entity key by determining whether the first entity key includes the second entity key;
- map the first activity key to the second activity key by determining whether the first activity key includes the second activity key;
- calculate a first sub-score based on the action scoring rubric and whether the first entity key includes the second entity key, wherein calculating the first-sub score comprises:
- responsive to a determination that the first entity key includes the second entity key, including the first value in the first-sub score;
- and
- responsive to a determination that the first entity key does not include the second entity key, including the second value in the first-sub score;
- calculate a second sub-score based on the action scoring rubric and whether the first activity key includes the second activity key;
- and
- calculate, by combining the first sub-score and the second sub-score, a score for a user, wherein the score indicates a clinical judgment evaluation of the user.
A system configured to provide a computer-implemented medical training simulation, comprising:
- at least one user device comprising a computing device;
- a server;
- a natural language processing (NLP) system comprising at least one computing device;
- at least one processor, wherein the at least one user device, the at least one processor, the server and the NLP system are configured for at least one of wired or wireless communication over a data network;
- at least one computer-readable storage medium including at least one memory storing one or more instructions, wherein the at least one computer processor, in response to executing the one or more instructions, is configured to:
- generate, in the NLP system, a plurality of entity data objects, wherein each entity data object includes:
- a first entity key;
- and
- one or more entity values corresponding to the first entity key;
- generate, in the NLP system, a plurality of activity data objects, wherein each activity data object includes:
- a first activity key, and
- one or more activity values corresponding to the first activity key;
- generate, the server, an evaluation data object, wherein the evaluation data object includes a problem data object, including at least one second entity key, wherein the second key includes a medical condition and an observation data object, including at least one third entity key;
- an action data object, including at least one second activity key;
- and
- a rationale data object, including at least one fourth entity key and at least one third activity key;
- configure each problem data object, observation data object, action data object, and rationale data object of the evaluation data object with a corresponding scoring rubric, wherein the scoring rubric corresponding to the rationale data object includes:
- a first numeric value corresponding to the presence of the fourth entity key and the third activity key in a computer-readable parsing of a user response;
- and
- a second numeric value corresponding to the absence of at least one of the fourth entity key or the third activity key in the computer-readable parsing of the user response, wherein the second numeric value includes at least one of a zero value or a negative value.
Dependencies are inferred from the claim text.