Methods and apparatus for determining a clinician's intent to order an item

• US 10,032,127 B2
• Filed: 03/01/2013
• Issued: 07/24/2018
• Est. Priority Date: 02/18/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Habboush, Isam H., Zaccagnini, Davide
• Primary Examiner(s): Kanaan, Maroun P
• Application Number:
  US13/782,048
• Publication Number:
  US 20130246098A1
• Time in Patent Office: 1,971 Days
• Field of Search: 705 2, 705 3
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 5/025 Extracting rules from data
• CPC Class Codes: G06Q 10/087 Inventory or stock management, e.g. order filling, procurement or balancing against orders
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 40/20 for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

Techniques for determining a clinician'"'"'s intent to order an item may include processing a free-form narration, of an encounter with a patient, narrated by a clinician, using a natural language understanding engine implemented by one or more processors, to extract at least one clinical fact corresponding to a mention of an orderable item from the free-form narration. The processing may comprise distinguishing between whether the at least one clinical fact indicates an intent to order the orderable item or does not indicate an intent to order the orderable item. In response to determining that the at least one clinical fact indicates an intent to order the orderable item, an order may be generated for the orderable item.

Claims

1. A method comprising:

processing, using a natural language understanding engine implemented by one or more processors, a first free-form narration, narrated by a clinician, of an encounter with a patient, the processing comprising;

detecting a mention in the first free-form narration of a first orderable item of a type selected from the group consisting of a medication, a clinical procedure, and a laboratory test;

performing a first semantic analysis of at least a part of the first free-form narration, wherein performing the first semantic analysis comprises;

determining a type of the first orderable item detected in the first free-form narration, wherein the determining is performed by the natural language understanding engine;

in response to determining that the type of the first orderable item is one of one or more first types, performing the first semantic analysis by applying a trained statistical model; and

in response to determining that the type of the first orderable item is one of one or more second types, performing the first semantic analysis by applying a rules-based system;

determining, based on the first semantic analysis, that the detected mention of the first orderable item is in a statement expressing that the clinician intends to order the first orderable item;

in response to determining that the detected mention of the first orderable item in the first free-form narration is in a statement expressing that the clinician intends to order the first orderable item, generating an order for the first orderable item; and

processing, using the natural language understanding engine implemented by the one or more processors, a second free-form narration, narrated by the clinician, of an encounter with the patient, the processing comprising;

detecting a mention in the second free-form narration of a second orderable item of the type selected from the group consisting of a medication, a clinical procedure, and a laboratory test;

performing a second semantic analysis of at least a part of the second free-form narration, wherein performing the first semantic analysis comprises;

determining a type of the second orderable item detected in the second free-form narration, wherein the determining is performed by the natural language understanding engine;

in response to determining that the type of the second orderable item is one of the one or more first types, performing the second semantic analysis by applying a trained statistical model; and

in response to determining that the type of the second orderable item is one of the one or more second types, performing the second semantic analysis by applying a rules-based system;

determining, based on the second semantic analysis, that the detected mention of the second orderable item is in a statement not expressing that the clinician intends to order the second orderable item; and

in response to determining that the detected mention of the second orderable item in the second free-form narration is in a statement not expressing that the clinician intends to order the second orderable item, not generating an order for the second orderable item.

2. The method of claim 1, further comprising identifying a set of information fields needed for generating the order for the first orderable item.

3. The method of claim 1, further comprising:

identifying one or more missing information fields in a set of information fields that have not been extracted from the first free-form narration; and

prompting the clinician to supply at least one of the one or more missing information fields.

4. The method of claim 3, further comprising inferring at least one of the one or more missing information fields from documentation of the patient'"'"'s history.

5. The method of claim 1, further comprising:

identifying a target computerized system for fulfilling the generated order;

conforming the generated order to a standard appropriate for the target computerized system for fulfilling the order; and

transmitting the conformed order to the target computerized system.

6. The method of claim 1, wherein determining, based on the first semantic analysis of the at least the part of the first free-form text narrative, that the detected mention of the first orderable item is in a statement expressing that the clinician intends to order the first orderable item comprises using contextual information of the at least the part of the first free-form text narrative.

7. The method of claim 6, wherein the contextual information includes text of the first free-form narration outside of the part.

8. The method of claim 1, wherein the rules-based system comprises a plurality of rules that each specify one or more words or one or more combination of words and one or more facts to extract when the one or more words or the one or more combination of words are detected.

9. The method of claim 1, wherein:

the processing further comprises;

segmenting the first free-form narration into one or more sections; and

identifying, from the one or more sections of the first free-form narration, at least one first section that is likely to include a mention of an orderable item; and

performing the first semantic analysis on the at least the part comprises performing the first semantic analysis on the at least one first section.

10. Apparatus comprising:

at least one processor; and

at least one processor-readable storage medium storing processor-executable instructions that, when executed by the at least one processor, cause the at least one processor to perform a method comprising;

processing, using natural language understanding, a first free-form narration,

narrated by a clinician, of an encounter with a patient, the processing comprising;

detecting a mention in the first free-form narration of an orderable item of a type selected from the group consisting of a medication, a clinical procedure, and a laboratory test;

performing a first semantic analysis of at least a part of the first free-form narration, wherein performing the first semantic analysis comprises;

determining a type of the first orderable item detected in the first free-form narration, wherein the determining is performed by the natural language understanding engine;

in response to determining that the type of the first orderable item is one of one or more first types, performing the first semantic analysis by applying a trained statistical model; and

in response to determining that the type of the first orderable item is one of one or more second types, performing the first semantic analysis by applying a rules-based system;

determining, based on the first semantic analysis, whether the detected mention of the orderable item is in a statement expressing that the clinician intends to order the orderable item; and

in response to determining that the detected mention of the orderable item in the first free-form narration is in a statement expressing that the clinician intends to order the orderable item, generating an order for the orderable item.

11. The apparatus of claim 10, wherein the method further comprises:

identifying one or more missing information fields in a set of information fields that have not been extracted from the first free-form narration; and

prompting the clinician to supply at least one of the one or more missing information fields.

12. The apparatus of claim 11, wherein the method further comprises inferring at least one of the one or more missing information fields from documentation of the patient'"'"'s history.

13. The apparatus of claim 10, wherein the method further comprises:

identifying a target computerized system for fulfilling the generated order;

conforming the generated order to a standard appropriate for the target computerized system for fulfilling the order; and

transmitting the conformed order to the target computerized system.

14. At least one non-transitory computer-readable storage medium encoded with computer-executable instructions that, when executed, perform a method comprising:

processing, using natural language understanding, a first free-form narration, narrated by a clinician, of an encounter with a patient, the processing comprising;

detecting a mention in the first free-form narration of an orderable item of a type selected from the group consisting of a medication, a clinical procedure, and a laboratory test;

performing a first semantic analysis of at least a part of the first free-form narration, wherein performing the first semantic analysis comprises;

determining a type of the first orderable item detected in the first free-form narration, wherein the determining is performed by the natural language understanding engine;

in response to determining that the type of the first orderable item is one of one or more first types, performing the first semantic analysis by applying a trained statistical model; and

in response to determining that the type of the first orderable item is one of one or more second types, performing the first semantic analysis by applying a rules-based system;

determining, based on the first semantic analysis, whether the detected mention of the orderable item is in a statement expressing that the clinician intends to order the orderable item;

in response to determining that the detected mention of the orderable item in the first free-form narration is in a statement expressing that the clinician intends to order the orderable item, generating an order for the orderable item; and

in response to determining that the detected mention of the orderable item in the first free-form narration is in a statement not expressing that the clinician intends to order the orderable item, not generating an order for the orderable item.

15. The at least one computer-readable storage medium of claim 14, wherein the method further comprises identifying one or more missing information fields in a set of information fields that have not been extracted from the first free-form narration.

16. The at least one computer-readable storage medium of claim 15, wherein the method further comprises prompting the clinician to supply at least one of the one or more missing information fields.

17. The at least one computer-readable storage medium of claim 16, wherein the method further comprises inferring at least one of the one or more missing information fields from documentation of the patient'"'"'s history.

18. The at least one computer-readable storage medium of claim 14, wherein the method further comprises:

identifying a target computerized system for fulfilling the generated order;

conforming the generated order to a standard appropriate for the target computerized system for fulfilling the order; and

transmitting the conformed order to the target computerized system.