Patient data mining with population-based analysis

• US 8,626,533 B2
• Filed: 11/04/2002
• Issued: 01/07/2014
• Est. Priority Date: 11/02/2001
• Status: Active Grant

• Current Assignee: Cerner Innovation, Inc. (Oracle Corporation)
• Original Assignee: Siemens Corp. (Siemens AG), Siemens Medical Solutions USA Incorporated (Siemens AG)
• Inventors: Rao, R. Bharat, Sandilya, Sathyakama
• Primary Examiner(s): Gilligan, Luke
• Assistant Examiner(s): RAPILLO, KRISTINE K
• Application Number:
  US10/287,329
• Publication Number:
  US 20030125988A1
• Time in Patent Office: 4,082 Days
• Field of Search: 705/3, 705/2
• US Class Current: 705/3
• CPC Class Codes: G06F 16/30 of unstructured textual data document management systems G06F16/93
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• 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 40/63 for local operation
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients
• CPC Class Codes: G16H 70/60 relating to pathologies
• CPC Class Codes: Y10S 128/92 Computer assisted medical diagnostics

Abstract

A system and method for analyzing population-based patient information is provided. The method includes the steps of data mining a plurality of patient records using a domain knowledge base relating to a disease of interest; compiling the mined data into a plurality of structured patient records; inputting at least one patient criteria relating to the disease of interest; and extracting at least one structured patient record matching the at least one patient criteria. The system includes a data miner for mining information from the plurality of patient records using a domain knowledge base relating to a disease of interest and for compiling the mined data into a plurality of structured patient records; an interface for inputting at least one patient criteria relating to the disease of interest; and a processor adapted for extracting at least one of the structured patient records matching the at least one patient criteria.

Claims

1. A method for analyzing patient records, the method comprising:

data mining, by a processor, a plurality of first patient records including unstructured data, the data mining using a domain knowledge base relating to a disease of interest and mining the unstructured data, the data mining comprising;

mining from a plurality of different data sources for each of the first patient records, each of the first patient records being for a respective patient;

for each of a plurality of variables for each patient, extracting different values from multiple of the different data sources such that multiple of the values are provided for each of the variables representing the patient at a same time;

for each of the plurality of variables for each patient, assigning a probabilistic assertion to each of the values from the different data sources such that the multiple values for each variable are each assigned probability, resulting in different values and respective probabilistic assertions for each variable of the plurality of variables representing each patient; and

for each of the plurality of variables for each patient, combining by inference the probabilities for each of the values from the different data sources into a unified probability for the respective variable, the unified probability providing a factoid representing a final value for the respective variable such that final values are provided for the respective variables for each patient;

the mining resulting in a separate factoid for each variable of the plurality of variables, the separate factoids for the plurality of variables being provided by the mining for each of the patients from the different data sources;

compiling, with the processor, the factoids into a plurality of structured patient records of a structured computerized patient record database and associated with patient variables for each of the patients, each structured patient record for respective patients comprising the factoids; and

correlating, with a system, an outcome with the factoids associated with patient variables;

wherein correlating comprises correlating the outcome from the factoids.

2. The method as in claim 1, further comprising:

inputting at least one patient criteria relating to the disease of interest; and

extracting at least one structured patient record matching the at least one patient criteria; and

determining the outcome of the at least one structured patient record.

3. The method as in claim 2, further comprising the steps of changing a value of the at least one patient criteria and repeating the extracting and determining steps.

4. The method as in claim 1, wherein the plurality of first patient records are stored in structured and unstructured sources, the data mining converting unstructured information into structured information.

5. The method as in claim 1, wherein the plurality of first patient records are collected over a course of treatment of a plurality of patients.

6. The method as in claim 2, further comprising correlating a plurality of criteria to a plurality of patient outcomes.

7. The method as in claim 6, further comprising suggesting a hypothesis for a clinical trial based on the correlation.

8. The method as in claim 7, further comprising validating the hypothesis by performing a retrospective study on the plurality of structured patient records.

13. The method of claim 1, wherein compiling the mined data into a plurality of structured patient records associated with patient variables comprises compiling into structured patient records associated with patient states, treatment, or combinations thereof.

15. The method of claim 1 wherein data mining comprises data mining free text using the domain knowledge base, the data mining identifying values of variables related to the disease of interest.

16. The method of claim 2 further comprising:

outputting an illustration of the outcome as a function of different values of a first of the possible patient variables.

17. The method of claim 1 further comprising:

determining patient records matching a guideline from a prospective clinical trail; and

performing the compiling and correlating with the patient records matching the guideline but restricted to a category of patients not included in the prospective clinical trail.

9. A system for analyzing a plurality of patient records, the plurality of patient records being stored in structured and unstructured sources, the system comprising:

a data miner for mining information from the plurality of first patient records including the unstructured sources using a domain knowledge base relating to a disease of interest, the data miner configured to;

mine from a plurality of different data sources including the unstructured sources for each of the first patient records, each of the first patient records being for a respective patient;

for each of a plurality of variables for each patient, extract different values from multiple of the different data sources such that multiple of the values are provided for each of the variables representing the patient at a same time;

for each of the plurality of variables for each patient, assign a probabilistic assertion to each of the values from the different data sources such that the multiple values for each variable are each assigned a probability resulting in different values and respective probabilistic assertions for each variable of the plurality of variables representing each patient; and

for each of the plurality of variables for each patient, combine by inference the probabilities for each of the values from the different data sources into a unified probability for the respective variable, the unified probability providing a factoid representing a final value for the respective variable such that final values are provided for the respective variables for each patient;

the mining resulting in a separate factoid for each variable of the plurality of variables, the separate factoids for the plurality of variables being provided by the mining for each of the patients from the different data sources;

the data miner compiling mined data as the factoids, the mined data being from the unstructured sources, into a plurality of structured patient records of a structured computerized patient record database and associated with patient variables for each of the patients, each structured patient record for respective patients comprising the factoids;

a processor operable to correlate an outcome with the mined data associated with the patient variables as a function of at least one of the structured patient records;

wherein the processor is operable to correlate the outcome from the factoids.

10. The system as in claim 9, further comprising an interface for inputting at least one patient criteria relating to the disease of interest, wherein the processor is operable to extract the at least one of the structured patient records as a function of the patient criteria.

11. The system as in claim 10, wherein the processor is operable to correlate a plurality of criteria to a plurality of patient outcomes.

12. The system as in claim 11, wherein the processor is operable to suggest a hypothesis for a clinical trial based on the correlation.

14. The system of claim 9, wherein the processor is operable to correlate the outcome with patient states, treatment, or combinations thereof.