AUTOMATED MAPPING OF SERVICE CODES IN HEALTHCARE SYSTEMS

• US 20140095205A1
• Filed: 09/26/2013
• Published: 04/03/2014
• Est. Priority Date: 09/28/2012
• Status: Active Grant

• Current Assignee: Cerner Corporation (Oracle Corporation)
• Original Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Inventors: Farooq, Faisal, Fung, Glenn, Yu, Shipeng, Rao, Bharat R., Krishnapuram, Balaji, DeHaan, Jan, Overhage, Marc
• Granted Patent: US 10,403,391 B2
• Time in Patent Office: Days
• US Class Current: 705/3
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records

Abstract

Automatic mapping of semantics in healthcare is provided. Data sets have different semantics (e.g., Gender designated with M and F in one system and Sex designated with 1 or 2 in another system). For semantic interoperability, the semantic links between the semantic systems of different healthcare entities are created (e.g., Gender=Sex and/or 1=F and 2=M) by a processor from statistics of the data itself. The distribution of variables, values, or variables and values, with or without other information and/or logic, is used to create a map from one semantic system to another. Similar distributions of other variable and/or values are likely to be for variables and/or values with the same meaning.

Claims

1. A method for automatic mapping of semantics in healthcare, the method comprising:

accessing first transaction data of a first healthcare entity in a first database, the first transaction data corresponding to a first semantic system;

calculating, by a processor, a first distribution of information in the first transaction data;

accessing second transaction data of a second healthcare entity in a second database, the second transaction data corresponding to a second semantic system different than the first semantic system;

calculating, by the processor, a second distribution of information in the second transaction data;

comparing, by the processor, the first and second distributions with machine learning; and

outputting, from the machine learning, a map relating the first semantic system to the second semantic system, the map being a function of the comparing.

2. The method of claim 1 wherein accessing the first and second transaction data in the first and second databases comprises accessing computerized orders, results, or services of the first and second healthcare entities, the first and second databases being separate or a same database.

3. The method of claim 1 wherein accessing the first and second transaction data in the first and second semantic systems comprises accessing the first and second transaction data representing variables, values for the variables, or the variables and the values for the variables labeled differently in the first semantic system than for the second semantic system.

4. The method of claim 1 wherein calculating the first and second distributions comprises counting occurrences for a plurality of variables, values, or variables and values of the first and second transaction data associated with a selected one of the variables, values, or variables and values of the first and second transaction data, respectively.

5. The method of claim 1 wherein comparing comprises determining statistical similarity of the first distribution with the second distribution and other distributions calculated from the second transaction dataset.

6. The method of claim 1 wherein comparing comprises performing unsupervised machine learning.

7. The method of claim 1 wherein comparing comprises comparing with the first and second distributions, linguistic information, number of associated characteristics, and data field characteristics as input features for the machine learning.

8. The method of claim 1 further comprising repeating calculating the first distribution and comparing for additional transactions for the first healthcare entity received after previously comparing.

9. The method of claim 1 wherein outputting the map comprises outputting semantic links between variables, values, or variables and values for the first dataset with variables, values, or variables and values for the second dataset.

10. The method of claim 1 wherein outputting the map comprises outputting the map with a two to five possible semantic links;

further comprising soliciting user selection of one of the two to five possible links.

11. The method of claim 1 wherein outputting the map comprises outputting the map with one-to-one semantic linking of all values, variables, or values and variables of a predetermined set, the comparing and outputting performed without user selection.

12. The method of claim 1 further comprising mapping information from the first data set to information from the second data set with the map.

13. In a non-transitory computer readable storage medium having stored therein data representing instructions executable by a programmed processor for automatic mapping of service codes in healthcare, the storage medium comprising instructions for:

obtaining the service codes, the service codes having different semantics; and

creating, by the programmed processor, a map relating the different semantics, the map created as a function of statistical information of the service codes.

14. The non-transitory computer readable storage medium of claim 13 wherein obtaining comprises obtaining the service codes as medical data from different medical entities.

15. The non-transitory computer readable storage medium of claim 13 wherein creating comprises creating with unsupervised machine learning with the statistical information being an input feature to the unsupervised machine learning and comprising at least one characteristic of a distribution of values, variables, or values and variables of the service codes relative to each other.

16. The non-transitory computer readable storage medium of claim 13 wherein creating comprises creating the map as a function of the statistical information from the service codes, patient data associated with the service codes, population statistics, and institution information.

17. The non-transitory computer readable storage medium of claim 13 wherein creating comprises calculating strengths of possible semantic links, applying the strengths to a threshold, and outputting the possible semantic links with strengths above the threshold and not with strengths below the threshold.

18. A system for automatic mapping of service codes in healthcare, the system comprising:

a memory configured to store first medical data in a first representation and second medical data in a second representation different than the first representation; and

a processor configured to link values, variables, or values and variables of the first medical data with values, variables, or values and variables of the second medical data, the links being for corresponding semantic meaning in the first and second representations.

19. The system of claim 18 wherein the processor is configured to apply unsupervised machine learning, an output of the unsupervised machine learning being a map comprising links of the values, variables, or values and variables of the first and second medical data.

20. The system of claim 18 wherein the processor is configured to link as a function of similarity of distribution of the values, variables, or values and variables of the first and second medical data.

21. The system of claim 20 wherein the processor is configured to compare the distribution of third medical data to distribution of the first medical data and generate a link as a function of the comparison.

22. The system of claim 18 wherein the processor is configured to update the links as additional transactions are received.

23. The system of claim 20 wherein the processor is configured to link as a function of a previous selection of possible links.