Electronic medical record summary and presentation

• US 10,311,206 B2
• Filed: 06/19/2014
• Issued: 06/04/2019
• Est. Priority Date: 06/19/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Boloor, Keerthana, Devarakonda, Murthy V., Tsou, Ching-Huei, Zhang, Dongyang
• Primary Examiner(s): Sorey, Robert A
• Application Number:
  US14/309,058
• Publication Number:
  US 20150370979A1
• Time in Patent Office: 1,811 Days
• CPC Class Codes: G16H 10/20 for electronic clinical trials or questionnaires
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 15/00 ICT specially adapted for medical reports, e.g. generation or transmission thereof
• 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
• CPC Class Codes: G16H 70/20 relating to practices or guidelines

Abstract

Methods, devices, and systems (for outputting a case summary) receive an electronic medical record (EMR) for the medical patient, extract medical data from the EMR, provide a list of medical problems relevant to the EMR, identifying relations between the medical problems and the medical data using a question-answering (QA) system, and output the clinical summary for the EMR. The clinical summary comprises the list of medical problems, the medical data, and the relations.

Claims

1. A computer-implemented method of outputting a clinical summary for a medical patient, said method comprising:

receiving, by a computer, an electronic medical record (EMR) for said medical patient from an EMR system, said EMR comprising structured data and unstructured data;

analyzing, by said computer, said EMR using natural language processing techniques along with a medical ontology to recognize medical concepts within said EMR;

mapping data from said EMR to standardized medical concepts in said medical ontology to identify medical problems of said medical patient associated with said EMR, using said computer;

producing, by said computer, an annotated EMR by annotating contents of said unstructured data and said structured data to identify the medical concepts based on said mapping;

extracting, by said computer, medical data from said annotated EMR using named entity and relation annotators based on standardized medical concepts;

providing, by said computer, a list of medical problems contained in said annotated EMR;

generating, by said computer, a clinical data relationship question for each medical problem on said list of medical problems based on said medical concepts using a question template based on a medical problem and said medical data extracted from said EMR;

inputting, by said computer, each said clinical data relationship question into a question-answering (QA) system that uses natural language processing techniques, said QA system being separate from said EMR system and comprising a corpus of data having structured and unstructured data in a relevant medical domain, said corpus of data being maintained in at least one database separate from said EMR system;

said QA system automatically searching said corpus of data to retrieve answers to the clinical data relationship question;

said QA system obtaining an answer to each said clinical data relationship question and evidence profiles for each said answer, wherein said answer has a corresponding score indicating a degree of probability that the answer correctly answers the clinical data relationship question;

analyzing, by said computer, each said answer;

identifying, by said computer, whether a valid relation exists between said medical problem of said medical patient associated with said EMR and said medical data based on said score;

said computer filtering out medical problems that have said score below a threshold;

said computer creating a clinical summary for said medical patient associated with said EMR using a summarization template, said clinical summary comprising specific medical information for said medical patient comprising said list of medical problems, said medical data, and said relations;

said computer prioritizing said clinical summary for said EMR based on said evidence profile and said score; and

said computer outputting said clinical summary for said EMR.

2. The method according to claim 1, wherein said clinical summary further comprises sections containing medical concepts organized into categories.

3. The method according to claim 2, wherein said sections comprise laboratory test values, medications, and a timeline of clinical encounters.

4. The method according to claim 3, further comprising:

receiving, by said computer, user interaction with a clinical encounter output in said timeline of clinical encounters; and

said computer outputting information from said EMR related to said clinical encounter including content from a clinical note correlated to said clinical encounter.

5. The method according to claim 3, further comprising:

receiving, by said computer, user interaction with a laboratory test value or a medication output in said clinical summary; and

said computer outputting a timeline of values corresponding to said laboratory test values or amounts of medication.

6. The method according to claim 1, wherein said QA system comprises a probabilistic system that analyzes unstructured information and provides answers with scores indicating a degree of probability that the answers correctly answer the clinical data relationship question.

7. A system for outputting a clinical summary for a medical patient, said system comprising:

an electronic medical record (EMR) system containing electronic medical records for a plurality of patients;

a summary system connected to said EMR system; and

a question-answering (QA) system connected to said summary system, said QA system being separate from said EMR system and comprising a corpus of data having structured and unstructured data in a relevant medical domain, said corpus of data being maintained in a database separate from said electronic medical records,

said summary system comprising;

a receiving module receiving an electronic medical record (EMR) for said medical patient and a list of medical problems from said EMR system, said EMR comprising structured data and unstructured data;

an analysis module analyzing said EMR using natural language processing techniques along with a medical ontology to recognize medical concepts within said EMR, wherein said analysis module maps data from said EMR to standardized medical concepts in said medical ontology to identify medical problems of said medical patient associated with said EMR;

an extracting module extracting medical data from said EMR using named entity and relation annotators based on standardized medical concepts, according to mapped data from said EMR, and annotating contents of said unstructured data and said structured data to identify the medical concepts to produce an annotated EMR containing a list of medical problems in said annotated EMR;

a relation identification module identifying relations between said medical problems in said annotated EMR and said medical data; and

an outputting module;

said relation identification module generating a clinical data relationship question for each medical problem on said list of medical problems based on said medical concepts using a question template based on a medical problem and said medical data extracted from said EMR,

said relation identification module inputting each said clinical data relationship question into said QA system,

said QA system using natural language processing techniques and automatically searching said corpus of data to retrieve answers to the clinical data relationship question,

said QA system obtaining an answer to each said clinical data relationship question and evidence profiles for each said answer, wherein said answer has a corresponding score indicating a degree of probability that the answer correctly answers the clinical data relationship question,

said relation identification module analyzing each said answer using a summarization template to identify valid relations between the medical problems of said medical patient associated with said EMR and said medical data,

said relation identification module identifying whether a relation exists between said medical problem and said medical data based on said score by filtering out candidate medical problems that have said score below a threshold to leave medical problem concepts of said medical patient, and

said outputting module creating a clinical summary for said medical patient associated with said EMR comprising said list of medical problems, said medical data, and said relations, using said summarization template, prioritizing said clinical summary for said EMR based on said evidence profile and said score; and

outputting said clinical summary for said EMR, wherein said clinical summary comprises said list of medical problems, said medical data, and said relations.

8. The system according to claim 7, wherein said clinical summary further comprises sections containing medical concepts organized into categories.

9. The system according to claim 8, wherein said sections comprise laboratory test values, medications, and a timeline of clinical encounters.

10. The system according to claim 9, said receiving module receiving user interaction with a clinical encounter output in said timeline of clinical encounters, and

said outputting module outputting information from said EMR related to said clinical encounter including content from a clinical note correlated to said clinical encounter.

11. The system according to claim 9, said receiving module receiving user interaction with a laboratory test value or a medication output in said clinical summary;

and

said outputting module outputting a timeline of values corresponding to said laboratory test values or amounts of medication.

12. The system according to claim 7, wherein said QA system comprises a probabilistic system that analyzes unstructured information and provides answers with scores indicating a degree of probability that the answers correctly answer the clinical data relationship question.

13. A computer program product, said computer program product comprising a computer readable storage medium having program instructions embodied therewith, wherein the computer readable storage medium is not a transitory signal per se, the program instructions being executable by a computer, to perform a method of outputting a clinical summary for a medical patient comprising:

automatically, by said computer, receiving an electronic medical record (EMR) for said medical patient from an EMR system, said EMR comprising structured data and unstructured data;

automatically, by said computer, analyzing said EMR using natural language processing techniques along with a medical ontology to recognize medical concepts within said EMR;

automatically, by said computer, mapping data from said EMR to standardized medical concepts in said medical ontology to identify medical problems of said medical patient associated with said EMR;

automatically, by said computer, producing an annotated EMR by annotating contents of said unstructured data and said structured data to identify the medical concepts, based on said mapping;

automatically, by said computer, extracting medical data from said annotated EMR using named entity and relation annotators based on standardized medical concepts;

automatically, by said computer, providing a list of medical problems found in said annotated EMR;

automatically, by said computer, generating a clinical data relationship question for each medical problem on said list of medical problems based on said medical concepts using a question template based on a medical problem and said medical data extracted from said EMR;

automatically, by said computer, inputting each said clinical data relationship question into a question-answering (QA) system that uses natural language processing techniques, said QA system being separate from said EMR system and comprising a corpus of data having structured and unstructured data in a relevant medical domain, said corpus of data being maintained in at least one database separate from said EMR system;

said QA system automatically searching said corpus of data to retrieve answers to the clinical data relationship question;

said QA system obtaining an answer to each said clinical data relationship question and evidence profiles for each said answer, wherein said answer has a corresponding score indicating a degree of probability that the answer correctly answers the clinical data relationship question;

automatically, by said computer, analyzing each said answer;

automatically, by said computer, identifying whether a valid relation exists between said medical problem and said medical data based on said score;

automatically, by said computer, filtering out medical problems that have said score below a threshold;

automatically, by said computer, creating a clinical summary for said medical patient associated with said EMR using a summarization template, said clinical summary comprising specific medical information for said medical patient comprising said list of medical problems, said medical data, and said relations;

automatically, by said computer, prioritizing said clinical summary for said EMR based on said evidence profile and said score; and

automatically, by said computer, outputting said clinical summary for said EMR.

14. The computer program product according to claim 13, wherein said clinical summary further comprises sections containing medical concepts organized into categories.

15. The computer program product according to claim 14, wherein said sections comprise laboratory test values, medications, and a timeline of clinical encounters.

16. The computer program product according to claim 15, further comprising:

receiving, by said computer, user interaction with a clinical encounter output in said timeline of clinical encounters; and

said computer outputting information from said EMR related to said clinical encounter including content from a clinical note correlated to said clinical encounter.

17. The computer program product according to claim 15, further comprising:

receiving, by said computer, user interaction with a laboratory test value or a medication output in said clinical summary; and

said computer outputting a timeline of values corresponding to said laboratory test values or amounts of medication.