Specification-based automation methods for medical content extraction, data aggregation and enrichment

• US 20060005139A1
• Filed: 06/02/2005
• Published: 01/05/2006
• Est. Priority Date: 06/10/2004
• Status: Active Grant

• Current Assignee: Siemens Corp. (Siemens AG)
• Original Assignee: Siemens Corp. (Siemens AG)
• Inventors: Comaniciu, Dorin, Liu, Peiya, Palanivelu, Sridharan
• Granted Patent: US 7,707,169 B2
• Time in Patent Office: Days
• US Class Current: 715/760
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records

Abstract

A method for knowledge generation from raw medical records uses XML-based specifications. The method includes content extraction, data aggregation and data enrichment. The method operates on various sources of medical data including financial data, clinical documents and medical images.

Claims

1. A method for generating a structured medical content database from a diverse set of raw medical data, the method comprising the steps of:

extracting content from the raw medical data according to mark-up language-based specifications, to produce a medical records database;

aggregating records in the medical records database according to rules in a generalized specification-based language, to produce an aggregated medical records database; and

enriching the aggregated medical records database using a generalized pipeline processing engine comprising mark-up language templates.

2. The method of claim 1, wherein the step of extracting content is performed according to XML-based specifications.

3. The method of claim 1, wherein the step of extracting content comprises extracting text-oriented content related to keywords and image-oriented content related to geometrical objects.

4. The method of claim 1, wherein the step of extracting content further comprises splitting documents in the raw medical data using XML-based specifications.

5. The method of claim 1, wherein the step of extracting content further comprises indexing documents in the raw medical data using XML-based specifications.

6. The method of claim 1, wherein the generalized specification-based language used in the aggregating step is an XML-based language.

7. The method of claim 1, wherein the generalized specification-based language used in the aggregating step further comprises:

a data extraction specification identifying at least one input data source and at least one output target table;

a constraint differential description containing at least one data normalization operation; and

at least one record merge rule for merging records to create the aggregated medical records database.

8. The method of claim 7 wherein the data extraction specification is based on a virtual database view of a transformation of the extracted content.

9. The method of claim 8 wherein the data extraction specification is further based on a target table view of a data mapping of fields of the virtual database view to fields of the at least one output target table.

10. The method of claim 1 wherein the mark-up language templates used in the step of enriching the aggregated medical records database are XML-based templates.

11. The method of claim 1, wherein the mark-up language templates perform at least one task chosen from a group consisting of structuring, transformation, hyper linking, query, discovering, mining and statistical analysis.

12. A computer program product comprising a computer readable recording medium having recorded thereon a computer program comprising code means for, when executed on a computer, instructing said computer to control steps in a method for generating a structured medical content database from a diverse set of raw medical data, the steps comprising:

extracting content from the raw medical data according to mark-up language-based specifications, to produce a medical records database;

aggregating records in the medical records database according to rules in a generalized specification-based language, to produce an aggregated medical records database; and

enriching the aggregated medical records database using a generalized pipeline processing engine comprising mark-up language templates.

13. The computer program product of claim 12, wherein the step of extracting content is performed according to XML-based specifications.

14. The computer program product of claim 12, wherein the step of extracting content comprises extracting text-oriented content related to keywords and image-oriented content related to geometrical objects.

15. The computer program product of claim 12, wherein the step of extracting content further comprises splitting documents in the raw medical data using XML-based specifications.

16. The computer program product of claim 12, wherein the step of extracting content further comprises indexing documents in the raw medical data using XML-based specifications.

17. The computer program product of claim 12, wherein the generalized specification-based language used in the aggregating step is an XML-based language.

18. The computer program product of claim 12, wherein the generalized specification-based language used in the aggregating step further comprises:

a data extraction specification identifying at least one input data source and at least one output target table;

a constraint differential description containing at least one data normalization operation; and

at least one record merge rule for merging records to create the aggregated medical records database.

19. The computer program product of claim 12 wherein the data extraction specification is based on a virtual database view of a transformation of the extracted content.

20. The computer program product of claim 19 wherein the data extraction specification is further based on a target table view of a data mapping of fields of the virtual database view to fields of the at least one output target table.

21. The computer program product of claim 12 wherein the mark-up language templates used in the step of enriching the aggregated medical records database are XML-based templates.

22. The computer program product of claim 12, wherein the mark-up language templates perform at least one task chosen from a group consisting of structuring, transformation, hyperlinking, query, discovering, mining and statistical analysis.