# US 2019/0172592 A1 · Apparatus for patient record identification

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2019/0172592 A1
- Document ID: US20190172592A1
- Status: Published application
- Kind: A1
- Application number: 16/186,688
- Filed: 2018-11-12
- Published: 2019-06-06
- Earliest claimed: 2017-12-04
- Assignees: Koninklijke Philips N.V.
- Applicants: Koninklijke Philips N.V.
- Inventors: Tak Ming Chan; Choo Chiap Chiau
- Web page: https://patentgrep.com/patent/US20190172592A1

## CPC

- G16H 50/70: for mining of medical data, e.g. analysing previous cases of other patients
- G06F 16/245: Query processing
- G16H 10/60: for patient-specific data, e.g. for electronic patient records
- G06F 2216/03: Data mining
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- G16H 50/50: for simulation or modelling of medical disorders
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

The present invention relates to an apparatus (10) for patient record identification. It is described to input (210) at least one query value relating to at least one medical condition. Patient records are received (220) from a patient database. Each patient record comprises at least one attribute value relating to the at least one medical condition, and each patient record comprises information relating to an outcome of the at least one medical condition. A first set of patient records is determined (230) based on a similarity between the at least one query value input by the user and the at least one attribute value for the patient records. The first set of patient records is separated (240) into patient records relating to a positive outcome for the at least one medical condition and into patient records relating to a negative outcome for the at least one medical condition. An implemented model calculates (250) at least one outcome probability for the at least one medical condition for each patient record of the first set of patient records based on the at least one attribute value for each patient record of the first set of patient records. A second set of patient records is determined (260) from the first set of patient records comprising at least one comparison comprising utilization of a number of patient records in a first sub-set of the patient records relating to the positive outcome and a number of patient records in a second subset of the patient records relating to the negative outcome and the at least one outcome probability for the at least one medical condition for each patient record of the first sub-set and second sub-set of the patient records.

## Claims

### Claim 1

An apparatus for patient record identification, comprising:
an input unit; and
a processing unit;
wherein, the input unit is configured to enable a user to input at least one query value relating to at least one medical condition;
wherein, the input unit is configured to receive patient records from a patient database, wherein each patient record comprises at least one attribute value relating to the at least one medical condition;
wherein, the processing unit is configured to determine a first set of patient records based on a similarity between the at least one query value input by the user and the at least one attribute value for the patient records;
wherein, the processing unit is configured to separate the first set of patient records into patient records relating to a positive outcome for the at least one medical condition and into patient records relating to a negative outcome for the at least one medical condition;
wherein, the processing unit is configured to implement a model to calculate at least one outcome probability for the at least one medical condition for each patient record of the first set of patient records based on the at least one attribute value for each patient record of the first set of patient records; and
wherein, the processing unit is configured to determine a second set of patient records from the first set of patient records comprising at least one comparison comprising utilization of a number of patient records in a first sub-set of the patient records relating to the positive outcome and a number of patient records in a second subset of the patient records relating to the negative outcome and the at least one outcome probability for the at least one medical condition for each patient record of the first sub-set and second sub-set of the patient records.

### Claim 2 (depends on 1)

Apparatus according to claim 1, wherein the at least one comparison comprises a first comparison between a first ratio of the number of patient records in the first sub-set of the patient records to the number of patient records in the second set of patient records with an average value of the at least one outcome probability for the at least one medical condition for each patient record of the first sub-set of patient records and comprises a second comparison between a second ratio of the number of patient records in the second sub-set of the patient records to the number of patient records in the second set of patient records with an average value of the at least one outcome probability for the at least one medical condition for the patient records of the second sub-set of patient records.

### Claim 3 (depends on 2)

Apparatus according to claim 2, wherein the first comparison comprises the first ratio being subtracted from an average value of the at least one outcome probability for the at least one medical condition for each patient record of the first sub-set of patient records; and wherein the second comparison comprises the second ratio being subtracted from an average value of the at least one outcome probability for the at least one medical condition for each patient record of the second sub-set of patient records.

### Claim 4 (depends on 1)

Apparatus according to claim 1, wherein determination of the second set of patient records comprises utilization of a magnitude of the at least one comparison.

### Claim 5 (depends on 4)

Apparatus according to claim 4, wherein the processing unit is configured to iteratively decrease the number of records in the second set of patient records comprising utilisation of the at least one attribute value for the patient records in the second set of patient records, wherein the records in the reduced second set of patient records are chosen such that there is an associated reduction in the magnitude of the at least one comparison.

### Claim 6 (depends on 5)

Apparatus according to claim 5, wherein the iteration proceeds until the magnitude of the at least one comparison falls below a threshold value.

### Claim 7 (depends on 5)

Apparatus according to claim 5, wherein the reduction in the number of records in the second set of patient records comprises utilisation of the at least one attribute value relating to the at least one medical condition for the second set of patient records prior to the reduction.

### Claim 8 (depends on 1)

Apparatus according to claim 1, wherein the at least one outcome probability for the at least one medical condition for each patient record of the first set of patient records comprises a probability relating to the positive outcome or negative outcome for the at least one medical condition.

### Claim 9 (depends on 1)

Apparatus according to claim 1, wherein each patient record comprises information relating to an outcome of the at least one medical condition, and wherein separation of the first set of patient records into patient records relating to a positive outcome for the at least one medical condition and into patient records relating to a negative outcome for the at least one medical condition comprises utilization of the information relating to an outcome of the at least one medical condition.

### Claim 10 (depends on 1)

Apparatus according to claim 1, wherein the at least one medical condition comprises a medical outcome and wherein the user is presented with a list of medical outcomes from which the at least one query is selected.

### Claim 11 (depends on 1)

Apparatus according to claim 1, wherein the processing unit is configured to select the model from a plurality of available models based on the at least one query input by the user.

### Claim 12 (depends on 1)

A system for patient record identification, comprising:
a patient database;
an apparatus for patient record identification according to claim 1; and
an output unit;
wherein, the patient records are provided from the patient database to the input unit; and
wherein, the output unit is configured to output the second set of patient records.

### Claim 13 (depends on 12)

System according to claim 12, wherein the output unit is configured to output a first dataset comprising the first sub-set of patient records of the second set of patient records and the associated at least one outcome probability for the at least one medical condition for each patient record, and output a second dataset comprising the second sub-set of patient records of the second set of patient records and the associated at least one outcome probability for the at least one medical condition for each patient record.

### Claim 14

A method for patient record identification, comprising:
a) inputting at least one query value relating to at least one medical condition;
b) receiving patient records from a patient database, wherein each patient record comprises at least one attribute value relating to the at least one medical condition, and wherein each patient record comprises information relating to an outcome of the at least one medical condition;
c) determining a first set of patient records based on a similarity between the at least one query value input by the user and the at least one attribute value for the patient records;
d) separating the first set of patient records into patient records relating to a positive outcome for the at least one medical condition and into patient records relating to a negative outcome for the at least one medical condition;
e) calculating by an implemented model at least one outcome probability for the at least one medical condition for each patient record of the first set of patient records based on the at least one attribute value for each patient record of the first set of patient records; and
determining a second set of patient records from the first set of patient records comprising at least one comparison comprising utilization of a number of patient records in a first sub-set of the patient records relating to the positive outcome and a number of patient records in a second subset of the patient records relating to the negative outcome and the at least one outcome probability for the at least one medical condition for each patient record of the first sub-set and second sub-set of the patient records.

### Claim 15 (depends on 1)

A computer program element for controlling an apparatus and/or a system, which when executed by a processor is configured to carry out the method of claim 1.
