Patient data mining
First Claim
Patent Images
1. A system for producing structured clinical information from patient records, the system comprising:
- a patient record comprising at least two data sources having patient information, at least one of the data sources being an unstructured data source and at least one of the data sources being a structured data source;
a probabilistic data miner of a computer platform configured to (a) extract multiple pieces of information related to a variable for a patient from mining structured data of the at least one structured data source and mining unstructured data of the at least one unstructured data source of the patient record, the mining of the at least one unstructured data source comprising mining free text information, and (b) combine the extracted multiple pieces of information related to the variable into a value of the variable for the patient, the value being a function of the multiple pieces related to the variable, and the data miner configured to repeat (a) and (b) for a plurality of different variables of the same patient for a same time, each repetition of extracting and combining multiple pieces of information related to the variable of the different variables being handled in the repetition such that the multiple pieces of information for one variable are different than the multiple pieces of information for other ones of the different variables, the variable and the different variables comprising characteristics of the patient at the time;
wherein one or both of (a) and (b) are performed as a function of domain-specific criteria.
6 Assignments
0 Petitions
Accused Products
Abstract
The present invention provides a data mining framework for mining high-quality structured clinical information. The data mining framework includes a data miner that mines medical information from a computerized patient record (CPR) based on domain-specific knowledge contained in a knowledge base. The data miner includes components for extracting information from the CPR, combining all available evidence in a principled fashion over time, and drawing inferences from this combination process. The mined medical information is stored in a structured CPR which can be a data warehouse.
150 Citations
54 Claims
-
1. A system for producing structured clinical information from patient records, the system comprising:
-
a patient record comprising at least two data sources having patient information, at least one of the data sources being an unstructured data source and at least one of the data sources being a structured data source; a probabilistic data miner of a computer platform configured to (a) extract multiple pieces of information related to a variable for a patient from mining structured data of the at least one structured data source and mining unstructured data of the at least one unstructured data source of the patient record, the mining of the at least one unstructured data source comprising mining free text information, and (b) combine the extracted multiple pieces of information related to the variable into a value of the variable for the patient, the value being a function of the multiple pieces related to the variable, and the data miner configured to repeat (a) and (b) for a plurality of different variables of the same patient for a same time, each repetition of extracting and combining multiple pieces of information related to the variable of the different variables being handled in the repetition such that the multiple pieces of information for one variable are different than the multiple pieces of information for other ones of the different variables, the variable and the different variables comprising characteristics of the patient at the time; wherein one or both of (a) and (b) are performed as a function of domain-specific criteria. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 42)
-
-
15. A method for producing structured clinical information from patient records, comprising:
-
(a) extracting, by a machine, multiple pieces of information from at least one structured data source and at least one unstructured data source of a patient record of a patient, the extracting from the at least one unstructured data source comprising mining unstructured free text information, the multiple pieces indicating different first values for a same variable recorded for the patient; (b) combining, probabilistically by the machine, the extracted multiple pieces of information into a second value of the variable representing the patient, the second value being a function of the first values recorded for the patient; wherein one or both of (a) and (b) are performed as a function of domain-specific criteria; and repeating (a) and (b) for a different variable, the repetition searching for different pieces of information relevant to the different variable and combining the different pieces of information into a value for the different variable, the variable and the different variable representing the patient at a given time. - View Dependent Claims (16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26)
-
-
27. A method for providing structured clinical information from patient records, the method comprising:
-
(a) mining, by a processor, a patient record having at least one unstructured data source comprising unstructured patient information, the mining comprising mining unstructured free text information, the patient record being from a healthcare provider, the mining including extracting at least one of multiple pieces of information related to each of multiple variables; (b) creating, probabilistically by the processor, structured clinical data for each of the variables from the extracted multiple pieces of information, including the at least one piece of the unstructured patient information mined from the unstructured data source, the structured clinical data being stored for answering a question regarding patients; (c) providing (a) as a service to the healthcare provider; and (d) mining from the structured clinical data as a function of the question. - View Dependent Claims (28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41)
-
-
43. A system for producing structured clinical information from patient records, comprising:
-
a patient record comprising at least unstructured data and structured data; and a probabilistic data miner machine configured to (a) extract multiple pieces of information from the structured data and the unstructured data of the patient record of a patient, the extracting from the at least unstructured data comprising mining unstructured free text information, the multiple pieces indicating different first values for a same variable recorded for the patient, and configured to (b) combine the extracted multiple pieces of information into a second value of the variable representing the patient, the second value being a function of the first values recorded for the patient; wherein one or both of (a) and (b) are performed as a function of domain-specific criteria; and wherein the probabilistic data miner is configured to repeat (a) and (b) for a different variable, the repetition searching for different pieces of information relevant to the different variable and combining the different pieces of information into a value for the different variable, the variable and the different variable representing the patient at a given time. - View Dependent Claims (44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54)
-
Specification