Method and system for generating a unified database from data sets

Patent metadata

Field	Value
Publication number	US-10025828-B2
Application number	US-201414543414-A
Country	US
Kind code	B2
Filing date	Nov 17, 2014
Priority date	Aug 23, 2013
Publication date	Jul 17, 2018
Grant date	Jul 17, 2018

Abstract

A method for generating a unified database includes receiving a structured set of data, where each set is made up of records having fields, aggregating values within a first field of the records, automatically applying a set of rules to the first field values to determine correlations among the first field values, calculating a confidence level regarding a label for the first field, providing the label to the first field, storing the first field values in the first field in the unified database, and receiving more information to increase the confidence level. A system for generating a clinical database and a method for using the database are also described.

First claim

The invention claimed is:

1. A computer-implemented method for generating a unified clinical database, comprising:
receiving a structured set of clinical data, the set comprising one or more records, each record having two or more fields;
aggregating values taken from a first field of the records;
aggregating values taken from a second field of the records;
automatically applying, by a processor, a set of rules to the aggregated first and second field values to calculate statistical correlations between the aggregated first and second field values;
calculating a confidence level regarding a label for the first field based on the rules;
if the confidence level meets or exceeds a pre-determined threshold,
applying the label to the first field;
storing the first field values in a first unified field in the unified database; and

if the confidence level does not meet or exceed the pre-determined threshold, receiving information regarding the correlations between the aggregated first and second field values and recalculating the confidence level.

2. The method of claim 1 , further comprising:
calculating a second confidence level regarding a label for the second field;
if the second confidence level meets or exceeds a pre-determined threshold,
applying the label to the second field; and
storing the second field values in a second unified field in the unified database; and

if the second confidence level does not meet or exceed the pre-determined threshold, receiving information regarding the correlations between the aggregated first and second field values and recalculating the second confidence level.

3. The method of claim 1 , further comprising:
calculating statistical distributions of the aggregated first and second field values; and
comparing the statistical distributions of the aggregated first and second field values with statistical distributions of stored data measures.

4. The method of claim 3 , wherein determining labels for the aggregated first and second field values comprises determining closest data measures having statistical distributions substantially the same as the statistical distributions of the aggregated first and second field values.

5. The method of claim 1 , wherein applying the set of rules to the aggregated first and second field values comprises determining a structure of the records.

6. The method of claim 5 , wherein the structure of the records of a CHEM-7 test comprises seven fields.

7. A computer-implemented method for generating a unified clinical database, comprising:
receiving a structured set of clinical data, the set comprising one or more records, each record having one or more fields;
aggregating values taken from a first field of the records;
automatically applying, by a processor, one or more rules to the aggregated first field values,
wherein applying said rules includes comparing the aggregated first field values to:
stored statistics of known clinical measures, if the aggregated first field values comprise numerical values;
known text entries from a clinical dictionary, if the aggregated first field values comprise alphabetical and/or alphanumeric values;
stored calendar information, if the aggregated first field values comprise date values; and
stored information concerning a clinical trial, including trial design information, if the aggregated first field values comprise alphabetical and/or alphanumeric values;
calculating a confidence level regarding a label for the first field;
if the confidence level meets or exceeds a pre-determined threshold,
applying the label to the first field; and
storing the first field values in a first unified field in the unified database; and

if the confidence level does not meet or exceed the pre-determined threshold, receiving more information and recalculating the confidence level.

8. The method of claim 7 , wherein the rules include eligibility criteria comprising inclusion criteria, exclusion criteria, or both.

9. The method of claim 7 , wherein the rules are refined based on the received data.

10. The method of claim 7 , wherein the stored statistics comprise statistical distributions of known clinical measures.

11. The method of claim 7 , wherein if the aggregated first field values comprise date values, and the date values are after the beginning of a clinical trial, then the first field label refers to testing during the clinical trial.

12. The method of claim 7 , wherein if the aggregated first field values comprise date values, and the date values are before the beginning of a clinical trial, then the first field label refers to historic data.

13. The method of claim 7 , further comprising:
aggregating values taken from a second field of the records;
automatically applying the one or more rules to the aggregated second field values to determine correlations between the aggregated first and second field values; and
determining a label for the second field values.

14. The method of claim 13 , wherein said correlations increase the confidence level regarding the label for the first field values.

15. A computer-implemented method for generating a unified database, comprising:
receiving a structured set of clinical data, the set comprising one or more records, each record having one or more fields;
aggregating values within a first field of the records;
automatically applying, by a processor, a set of rules to the first field values to calculate statistical correlations among the first field values, wherein applying the set of rules includes comparing the aggregated first field values to stored statistics of known data measures, if the aggregated first field values comprise numerical values;
calculating a confidence level regarding a label for the first field;
if the confidence level meets or exceeds a pre-determined threshold,
applying the label to the first field;
storing the first field values in the first field in the unified database; and
receiving more information to increase the confidence level; and

if the confidence level does not meet or exceed the pre-determined threshold, receiving more information and recalculating the confidence level.

16. The method of claim 15 , wherein applying the set of rules to the first field values comprises using inclusion and exclusion criteria.

17. The method of claim 15 , further comprising:
calculating a statistical distribution of the first field values; and
comparing the statistical distribution of the first field values with statistical distributions of stored data measures.

18. The method of claim 17 , further comprising:
determining a closest data measure having a statistical distribution substantially the same as the statistical distribution of the first field values.

19. The method of claim 17 , further comprising calculating the statistical distributions of the stored data measures.

20. The method of claim 15 , wherein applying the set of rules includes comparing the first field values to known text entries from a clinical dictionary, if the aggregated first field values comprise alphabetical and/or alphanumeric values.

21. The method of claim 15 , further comprising:
aggregating values within a second field of the records;
automatically applying the set of rules to the second field values to determine correlations among the second field values and first field values; and
determining a label for the second field values.

22. The method of claim 21 , wherein said determine

Assignees

• Medidata Solutions Inc

Inventors

• De Vries Glen
• Marlborough Michelle

Classifications

• G16H10/20

  for electronic clinical trials or questionnaires · CPC title

• G16H10/60

  for patient-specific data, e.g. for electronic patient records · CPC title

• G06F16/258 Primary

  Data format conversion from or to a database · CPC title

• G16H10/40

  for data related to laboratory analysis, e.g. patient specimen analysis · CPC title

• G06F16/13

  File access structures, e.g. distributed indices (arrangements of input from, or output to, record carriers G06F3/06) · CPC title