JP2005346239A - Similar case search system, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reflect a sample data having different data items such as inspection items to a search result, in searching a case data similar to the symptom of a patient. <P>SOLUTION: A control section 21 in a management server 2 receives a data value related to the patient in regard to a data item group including a plurality of data items, and among the above data item group, searches for the data item registered in a substitution data item DB 222 having stored information related to two data items of a large degree of correlation. In regard to the searched data item, using the information of correlation relationship in the substitution data item DB 222, the data value related to the patient is converted into a data value of the substitution data item. Then a query is generated using an item name of the data item in the data item group and the data value related to the patient, and the item name and the data value of the substitution data item, of which data value has been converted. The sample data fitting the generated query is searched for in a case DB in which the sample data related to each case in the past is registered. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a system for extracting past cases similar to a patient's symptoms from a database, and the like, in particular, a similar case search that can be searched as a similar case even between patients with different examination items. It relates to systems.

  Conventionally, when searching past cases similar to the patient's symptoms from the database, the patient's measurement value or a numerical value with a wide range of the measurement value is set as the search condition for all or part of the examination items received by the patient It is common to generate a query that has been generated and to search for data that matches the generated query.

For example, similar patient classification is extracted using the basic information of the patient as a search condition, the clinical path related to the extracted case classification is extracted, and the optimal clinical path is selected from the extracted clinical path. There is a clinical path operation support information system capable of creating a medical plan (see Patent Document 1).
Japanese Patent Laid-Open No. 2003-331055 (page 8, FIG. 11)

  However, the system as described above searches for sample data having the same examination item data as the patient. In such a system, the data is actually very similar to the case data. However, since the inspection items in the sample data are different, there arises a problem that similarities cannot be determined.

  The present invention has been made in view of the above circumstances, and in the search of case data similar to the patient's symptoms, similar cases that can be reflected in the search results also for sample data with different data items such as examination items The purpose is to provide a search system and the like.

In order to achieve the above object, a similar case search system according to the first aspect of the present invention provides:
A similar case retrieval system for retrieving sample data similar to a patient's symptom from a case database in which sample data regarding each past case is registered,
A substitute data item database for storing substitute data item information including item name information of each data item and correspondence information indicating the correspondence between data values of each data item for two data items having a high degree of correlation; ,
For a data item group consisting of a plurality of data items, accepting means for accepting data values related to the patient,
A specifying means for searching and specifying a data item registered in the substitute data item database in the data item group;
Using the correspondence information stored in the substitute data item database for the identified data item, the data value related to the patient is changed to the data item of the substitute data item that is the other data item having a high degree of correlation with the data item. Conversion means for converting to data values;
Query generation means for generating a query using the item name of the data item in the data item group and the data value related to the patient, and the item name and data value of the substitute data item obtained by converting the data value by the conversion means,
Search means for searching sample data that matches the generated query from the case database;
Output means for outputting a search result by the search means;
It is characterized by providing.

For each disease, a diagnostic criteria database in which information identifying the disease and a data item that is a criterion for determining the disease are registered in association with each other,
Means for receiving information specifying a disease;
Means for searching a data item corresponding to the specified disease from a diagnostic criteria database, and specifying the searched data item as the data item group;
May be further provided.

  The query generation means includes a query in which an item name and a data value of each data item in the data item group are set as a search condition, and a part of the data item specified by the specifying means in the data item group or A query in which the item name and data value of each data item in the data item group when all are replaced with corresponding substitute data items may be generated as a search condition.

For data items constituting sample data registered in the case database, means for extracting data items two by two and calculating the degree of correlation between the extracted data items;
For two data items with the calculated correlation degree equal to or greater than a predetermined value, correspondence information indicating the correspondence between the data values of each data item is generated, the item name of each data item, the correspondence information, Means for generating the surrogate data item information comprising:
Means for registering the generated substitute data item information in the substitute data item database;
May be further provided.

The search means calculates the difference between the data value of the searched sample data and the data value related to the patient in each data item of the data item group, and uses the difference calculated for each data item, It may further comprise a similarity score calculating means for calculating a similarity score indicating the degree of similarity between the sample data and the case of the patient,
The output means may further comprise means for outputting the searched sample data based on the similarity score.

  When there is a data item for which no data value is set in the retrieved sample data among the data items of the data item group, the similarity score calculation means calculates the data value of the substitute data item corresponding to the data item. The similarity score is extracted from the retrieved sample data, the data value of the substitute data item is converted into the data value of the data item corresponding to the substitute data item using the correspondence information in the substitute data item database You may use for calculation of.

A program according to the second aspect of the present invention is
On the computer,
An accepting step for accepting a data value related to a patient for a data item group comprising a plurality of data items;
Regarding a substitute data item database that stores substitute data item information including item name information of each data item and correspondence information indicating a correspondence relationship of data values of each data item with respect to two data items having a high degree of correlation , A specific step of searching and specifying a data item registered in the substitute data item database among the data item group,
Using the correspondence information stored in the substitute data item database for the identified data item, the data value related to the patient is changed to the data item of the substitute data item that is the other data item having a high degree of correlation with the data item. A conversion step to convert to a data value,
A query generation step of generating a query using the item name of the data item in the data item group and the data value related to the patient, and the item name and data value of the substitute data item obtained by converting the data value by the conversion unit;
A search step of searching for sample data matching the generated query from a case database in which sample data relating to past cases is registered;
An output step of outputting a search result by the search means;
Is executed.

In the computer,
Receiving information specifying a disease;
The data item corresponding to the specified disease is searched from a diagnostic criteria database in which information specifying the disease for each disease and a data item that is a criterion for the disease are registered in association with each other, and the searched data item is Identifying as data items,
May be further executed.

  The query generation step includes a query in which an item name and a data value of each data item in the data item group are set as a search condition, and a part of the data item specified by the specifying unit in the data item group or A query in which the item name and data value of each data item in the data item group when all are replaced with corresponding substitute data items may be generated as a search condition.

For the data items constituting the sample data registered in the case database, extracting the data items two by two and calculating the degree of correlation between the extracted data items;
For two data items with the calculated correlation degree equal to or greater than a predetermined value, correspondence information indicating the correspondence between the data values of each data item is generated, the item name of each data item, the correspondence information, Generating the surrogate data item information comprising:
Registering the generated substitute data item information in the substitute data item database;
May be further executed.

The search step calculates the difference between the data value of the searched sample data and the data value related to the patient in each data item of the data item group, and uses the difference calculated for each data item, The method may further comprise a similarity score calculating step for calculating a similarity score indicating the degree of similarity between the sample data and the patient's case,
The output step may further include a step of outputting the searched sample data based on the similarity score.

  When there is a data item for which no data value is set in the retrieved sample data among the data items of the data item group, the similarity score calculating step calculates the data value of the substitute data item corresponding to the data item. The similarity score is extracted from the retrieved sample data, the data value of the substitute data item is converted into the data value of the data item corresponding to the substitute data item using the correspondence information in the substitute data item database You may use for calculation of.

  According to the present invention, sample data with different data items such as examination items can be reflected in the search results in the search for case data similar to the patient's symptoms.

Hereinafter, a similar case search system according to an embodiment of the present invention will be described with reference to the drawings.
An example of the configuration of a similar case search system according to an embodiment of the present invention is shown in FIG. The similar case search system includes, for example, a case DB (database) 1, a management server 2, and a user terminal 3, and these devices are configured to be connectable via a network 10. The network 10 includes various networks such as a LAN (local area network) and the Internet.

  The case DB 1 is a database in which sample data of many past cases are registered. The sample data includes, for each case, a sample ID for specifying the sample and a plurality of data items. Here, the plurality of data items include items such as patient attribute information of the case, measurement values of various test items, and the like, for example, age, sex, systolic blood pressure, blood glucose level, hematocrit, BMI, genotype 1, hemoglobin, etc. Includes items. An example of sample data is shown in FIG.

  For example, as shown in FIG. 3, the management server 2 includes a computer including a control unit 21, a storage unit 22, an input unit 23, a display unit 24, and a communication control unit 25.

  The control unit 21 includes a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and the like, and reads out and executes an operation program or the like stored in advance in the storage unit 22 to execute a query. The generation / retrieval unit 211 and the substitute data item extraction unit 212 are logically realized.

  The query generation / search unit 211 generates a query (inquiry) for searching the case DB 1 in response to a search request from the user terminal 3, and searches for the case DB 1 using the generated query. I do. Details of the similar case search process will be described later. Further, when the query generation / search unit 211 receives data such as a disease name from the user terminal 3 together with the diagnosis reference request, the query generation / search unit 211 reads out the diagnosis reference data corresponding to the disease name indicated by the received data from the diagnosis reference DB 221 and Process to send to.

  The substitute data item extraction unit 212 is a data item having a high degree of correlation among two data items in which the degree of correlation of data values is equal to or greater than a preset value among the data items constituting the sample data registered in the case DB 1 Substitute data item extraction processing for registering data (substitute data item information) relating to the two extracted data items in the database of the storage unit 22 is performed. Details of the substitute data item extraction process will be described later.

The storage unit 22 is composed of, for example, a hard disk device or the like, and stores an operation program to be executed by the control unit 21 and various data necessary for processing.
The storage unit 22 includes a diagnostic criteria DB 221 and a substitute data item DB 222.
In the diagnostic criteria DB 221, for example, as shown in FIG. 4, a disease name of each disease and diagnostic criteria data indicating a data item serving as a criterion for diagnosing the disease are registered in association with each disease.
In the substitute data item DB 222, substitute data item information related to two data items having a correlation degree of data values of a predetermined value or more among the data items constituting the sample data registered in the case DB1 is registered. For example, as shown in FIG. 5, the substitute data item information includes, for two data items having a high degree of correlation (greater than a predetermined value), the item name of one data item, the item name of the other data item, and each data item Data such as the correlation coefficient of the data values and the correspondence information indicating the correspondence between the data values of the data items. The correspondence relationship information includes, for example, information such as a conversion formula (regression formula or the like) for calculating the data value of the other data item from the data value of one data item. In the present embodiment, for the two data items having a high degree of correlation registered in the substitute data item DB 222, the other data item corresponding to one data item is referred to as a substitute data item.

The input unit 23 is configured by an input device such as a keyboard and a pointing device, for example, and transmits input data to the control unit 21.
The display unit 24 is constituted by a display device such as a display device, for example, and outputs various data according to instructions from the control unit 21.
The communication control unit 25 controls communication with other devices via the network 10 in accordance with instructions from the control unit 21.

As shown in FIG. 6, for example, the user terminal 3 includes a computer including a control unit 31, a storage unit 32, a display unit 33, an input unit 34, and a communication control unit 35.
The control unit 31 reads out and executes an operation program or the like stored in advance in the storage unit 32 to perform search request processing for requesting search for sample data similar to a patient case. In this search request process, the control unit 31 receives input of information for specifying a data item group (data item group A) used for searching for sample data according to a predetermined input operation. For example, when there is an input for designating a specific disease, information such as a disease name of the designated disease is transmitted to the management server 2, and information on the data item corresponding to this disease (diagnosis reference data) ) And the data item indicated by the received data is defined as data item group A. When a data item is designated and input, the designated data item is set as a data item group A. Then, for the specified data item group A, the input of the data value related to the patient is accepted, and the data value of the data item group A and the search request are transmitted to the management server 2. Then, the search result data is received from the management server 2 and displayed on the display unit 34.

The storage unit 32 is composed of, for example, a hard disk device or the like, and stores an operation program to be executed by the control unit 31 and various data necessary for processing.
The input unit 33 is composed of an input device such as a keyboard and a pointing device, for example, and transmits input data to the control unit 31.
The display unit 34 is constituted by a display device such as a display device, for example, and outputs various data according to instructions from the control unit 31.
The communication control unit 35 controls communication with other devices via the network 10 in accordance with instructions from the control unit 31.

Next, similar case search processing by the query generation / search unit 211 of the management server 2 will be described with reference to the flowchart of FIG.
When the query generation / retrieval unit 211 receives the item name and patient data value of the data item group A and the search request from the user terminal 3, in response to this, each data item of the received data item group A The substitute data item DB 222 is searched for, the corresponding substitute data item information is read, and the data item group B corresponding to the data item group A is specified based on the read substitute data item information (step S1). Specifically, the substitute data item information in which the item name of each data item in the data item group A is set is retrieved from the substitute data item DB 222 and read. In the read substitute data item information, a substitute data item group corresponding to the data item group A is defined as a data item group B.

  Next, the query generation / search unit 211 uses the patient data value for the received data item group A and the substitute data item information read from the substitute data item DB 222 to search for cases similar to the patient's symptoms. A query for searching is generated, and query generation / search processing is performed for searching using the generated query (step S2).

Details of the query generation / retrieval process will be described with reference to the flowchart of FIG.
First, the query generation / search unit 211 obtains a range of data to be set as a search condition by setting an appropriate width for the patient data value for each data item in the data item group A (step S11). The method for setting the width is arbitrary. For example, it may be calculated according to a formula for calculating a preset width, a preset numerical value may be used, or an input from a user may be accepted. For example, the width may be obtained using the standard deviation. In this case, the query generation / search unit 211 is pre-calculated for each data item using, for example, sample data registered in the case DB 1 and stored in the storage unit. The standard deviation data of each data item belonging to data item group A is extracted from the standard deviation data of each data item stored in 22 etc., and (patient data value ± standard deviation) is calculated for each data item. Thus, the data range for each data item in the data item group A is obtained.

Next, for the data item group B, the query generation / retrieval unit 211 sets the data range corresponding to the data range of each data item of the data item group A set in step S11 to the substitute data item information of the data item group B. Is acquired using the correspondence relationship information (step S12).
Next, the query generation / retrieval unit 211 converts the data item in the data item group A into a query that uses the item name of each data item in the data item group A and the data range defined in step S11 as a search condition, and the data item. A query for obtaining all patterns of the data item group when a part or all of the group B is replaced and using the data item group of each pattern and the data range acquired in steps S11 and S12 as a search condition Are generated (step S13).

  For example, as shown in FIG. 9A, when an item name and a data value are set for each data item of the data item group A, the query generation / search unit 211 sets a data range for the data item group A. To do. Then, as shown in FIG. 9B, the data range of the data item group B is set for the substitute data item information read from the substitute data item DB 222. Then, as shown in FIG. 9C, a query (query number 1) using the item name and data range of each data item in the data item group A as a search condition, and the data item in the data item group A are represented by data items. A query (query number 2 to 8) is generated that uses the item name and data range of each data item in the data item group when part or all of group B is replaced as a search condition. In the process of generating all patterns of the data item group when the data items in the data item group A are replaced with part or all of the data item group B, for example, n (n = 1) from the data item group B 2,... (Total number of items in data item group B)), all combinations are extracted, and data items in each combination are replaced with corresponding data items in data item group A. You may perform the process to do.

After generating the query in step S13, the query generation / retrieval unit 211 includes the data item group A and the data item group when the data item of the data item group A is replaced with part or all of the data item group B, The sample data for which the data value is set is retrieved from the case DB 1 and extracted (step S14).
Next, the query generation / search unit 211 searches the sample data extracted in step S14 for sample data (similar sample data) that matches the query generated in step S13 (step S15).
Next, the query generation / search unit 211 performs a similarity score calculation process for calculating a similarity score with the patient data received from the user terminal 3 for each similar sample data searched in step S15 (step S16). . Details of the similarity score calculation processing will be described later. Then, the query generation / search unit 211 sorts the similar sample data in descending order of the similarity score calculated in step S16 (step S17), and returns to the main flow.
After the query generation / search process in step S3, the query generation / search unit 211 transmits similar sample data sorted in descending order of similarity score to the user terminal 3 that is the search request source (step S3). The process ends.

Next, the similarity score calculation process in the query generation / search process will be described with reference to the flowchart of FIG.
First, the query generation / retrieval unit 211 extracts one similar sample data (step S20), sets the value of the number of data items in the data item group A to the variable N, and sets 0 to the variables S and I, respectively. An initialization process is performed (step S21).

Next, the query generation / retrieval unit 211 adds 1 to I (step S22), and whether a data value is set for the data item DA (I) of the data item group A in the similar sample data to be processed. Is determined (step S23).
When a data value is set for the similar sample data to be processed (step S23: YES), for the data item DA (I), the difference between the patient data value and the data value of the similar sample data to be processed is calculated. A difference score s (I) is obtained (step S24). The calculation method of the difference score is arbitrary, and a value obtained by performing a predetermined calculation process on a difference between data values (for example, an absolute value of a numerical value obtained by subtracting the other data value from one data value) may be used. For example, the numerical value having the standard deviation as a unit may be obtained by dividing the difference between the data values by a predetermined value such as the standard deviation of the data item DA (I). Further, the difference between the data values may be used as it is as the difference score.

  In step S23, if no data value is set in the similar sample data for the data item DA (I) in the data item group A (step S23: NO), refer to the substitute data item information for the data item group B. Then, the substitute data item of the data item DA (I) is specified, and the data value of the specified substitute data item in the similar sample data is converted into the data value of the data item DA (I) using the correspondence information. Conversion (projection) is performed (step S25), and the process proceeds to step S24.

After step S24, the query generation / search unit 211 adds the obtained difference score s (I) to S (step S26). Then, it is determined whether or not “I <N” is satisfied. If satisfied (step S27: YES), the process proceeds to step S22.
If “I <N” is not satisfied (step S27: NO), the similarity score is calculated as 1 / S, for example (step S28).

  Then, it is determined whether or not the above process has been completed for all similar sample data (step S29). If completed (YES), the similarity score calculation process is terminated, and the process returns to the query generation / search process. If not completed (NO), the process proceeds to step S20.

  Next, proxy data item extraction processing by the proxy data item extraction unit 212 of the management server 2 will be described with reference to the flowchart of FIG. This proxy data item extraction process is periodically executed at predetermined time intervals using an internal timer or in response to a predetermined input instruction from the input unit.

  The surrogate data item extraction unit 212 acquires the number of data items in the sample data registered in the case DB1 and sets it as a variable N (step S31), extracts two data items from the data items in the sample data of the case DB1, A combination C of data items is generated, and a value of (N × (N−1) / 2) is set as the number of combinations M (step S32). Further, 0 is set to the variable I (step S33).

Next, the substitute data item extraction unit 212 adds 1 to I (step S34), takes out one combination C (I), and sample data in which data values are set for both data items indicated by this combination Is extracted (step S35).
Then, the substitute data item extraction unit 212 determines whether the number of samples extracted in step S35 is sufficient to obtain the correlation (step S36). The specific determination content is arbitrary, and for example, it may be determined whether the number of extracted samples is equal to or greater than a predetermined value set in advance.

When it is determined in step S36 that there is a sufficient number of samples (step S36: YES), the substitute data item extraction unit 212 calculates the degree of correlation (correlation coefficient) between the data items indicated by the combination C (I). (Step S37). And it is discriminate | determined whether the calculated correlation degree exceeds the threshold value (step S38), and when it exceeds, the correspondence information which shows the correspondence of each data value will be produced | generated (step S39). Specifically, for example, a regression equation for obtaining the data value of the other data item (substitute data item) from one data item is calculated.
Then, substitute data item information including data such as the item name, correlation degree, and correspondence information of each data item is generated and registered in the substitute data item DB 222 (step S40). Then, it is determined whether or not “I <M” is satisfied (step S41). If satisfied (YES), the process proceeds to step S34, and if not satisfied (NO), this process is terminated.

  If it is determined in step S36 that there is not a sufficient number of samples (step S36: NO), or if it is determined in step S38 that the correlation does not exceed the threshold value (step S38: NO), step The process proceeds to S34.

Next, the processing operation of the similar case search system according to the embodiment of the present invention will be described with reference to the flowchart of FIG.
First, in response to a predetermined input operation by the user, the user terminal 3 receives input of information for specifying a data item group (data item group A) used for searching for similar sample data (step S41). Specifically, for example, the user terminal 3 displays an input screen including an input field for inputting a data item name, an input field for inputting a disease name, and the like, and the item name of the data item to which the data item group A belongs, When searching for similar sample data, an input of a disease name of a disease of particular interest is received.

Next, the user terminal 3 determines whether or not there is an input for designating a specific disease (step S42). If there is an input for designating a specific disease (YES), the designation is made. Information such as the disease name of the disease that has been transmitted is transmitted to the management server 2 together with the diagnostic criteria request (step S43).
In response to this, the management server 2 reads out the diagnostic reference data corresponding to the received disease name from the diagnostic reference DB 221 and transmits it to the user terminal 3 (step S44).
The user terminal 3 receives the diagnostic reference data from the management server 2, and sets the data item indicated by it as the data item group A (step S45).

  In step S42, when there is no input for designating a specific disease (step S42: NO), the data item designated by the user is set as the data item group A (step S46).

Next, the user terminal 3 accepts input of data values such as measurement values of the patient P for the set data item group A, and manages the data item name of the data item group A and the input data value together with the search request. It transmits to the server 2 (step S47).
In response to the search request from the user terminal 3, the management server 2 executes a similar case search process. Since the details of the similar case search process are as described above, only the main process will be briefly described here.

The management server 2 searches the substitute data item DB 222 for each data item of the received data item group A, reads out the corresponding substitute data item information, and specifies the data item group B (step S48). Then, a query is generated using the patient data value for the data item group A and the substitute data item information read from the substitute data item DB 222 (step S49), and the case DB 1 is searched using the generated query. Is performed (step S50). And the management server 2 calculates a similarity score about the searched similar sample data, and transmits to the user terminal 3 of a search request origin with the information of the calculated similarity score as a search result (step S51).
The user terminal 3 displays the search result received from the management server 2, that is, sample data of cases similar to the symptoms of the patient P (step S52).

  As described above, according to the present invention, by specifying a data item group whose correlation degree is equal to or greater than a predetermined value with the specified data item group, and generating a query in which the search condition is also set, the data item Similar cases can be searched from different sample data.

The present invention can be variously modified and applied.
For example, in the query generation / retrieval process in the above embodiment, sample data having a data value for the data item group A or data item group obtained by replacing the data item of the data item group A with the data item of the data item group B Sample data having values, and only those sample data whose data values fall within a preset data range are targets for calculating the similarity score. However, the present invention is not limited to this, and the data item group A The similarity score is calculated for all of the sample data having the data value and the sample data having the data value for the data item group obtained by replacing the data item of the data item group A with the data item of the data item group B. Also good. The query generation / retrieval process in this case will be described with reference to the flowchart of FIG.

First, the query generation / retrieval unit 211 calculates a data value corresponding to the data value of the data item of the data item group A for the data item group B using the correspondence information in the substitute data item information (step S61). .
Next, the query generation / search unit 211 replaces the query using the item name of each data item in the data item group A as a search condition and the data item in the data item group A with a part or all of the data item group B. In this case, all patterns of the data item group are obtained, and a query using each item name as a search condition is generated for the data item group of each pattern (step S62).
Next, the query generation / retrieval unit 211 searches and extracts sample data that matches the generated query from the case DB 1 (step S63).
Next, the query generation / search unit 211 performs a similarity score calculation process for calculating the similarity score with the patient data received from the user terminal 3 for the extracted sample data (step S64).
Next, the query generation / search unit 211 extracts sample data whose calculated similarity score exceeds a predetermined threshold value as similar sample data (step S65).
Next, the query generation / search unit 211 sorts the similar sample data in descending order of the similarity score (step S66), and returns to the main flow.

  As correspondence information in the substitute data item information, for example, as shown in FIG. 14, each data range set for one data item is associated with each data range set for the other data item. Information may be used. When such correspondence information is used, the query generation / search unit 211 refers to the corresponding substitute data item information when acquiring the data range for the data item group B in the query generation / search process (step S12). , For one data item (data item of data item group A), the data range in which the patient's data value is entered is identified, and the data range of the other data item corresponding to the identified data range is designated as data item group B The data range of the data item may be acquired.

In addition, the system configuration in the above embodiment is an example, and for example, the management server 2 may include the case DB 1.
In addition, the management server 2 may perform processing for normalizing the distribution of each data item of the sample data registered in the case DB 1 to adjust the data variation of each data item.

The system of the present invention can be realized using a normal computer system, not a dedicated system. For example, a program for executing the above operation is stored in a computer-readable recording medium (FD, CD-ROM, DVD, etc.) and distributed, and the program is installed in the computer to execute the above processing. The management server 2, the user terminal 3 and the like may be configured. Alternatively, it may be stored in a disk device of a server device on a network such as the Internet and downloaded to a computer, for example.
In addition, when the OS realizes the above functions by sharing the OS or jointly with the OS and the application, etc., only the part other than the OS may be stored and distributed in the medium, or may be downloaded to the computer. Good.

It is a figure which shows the structure of the similar case search system of embodiment of this invention. It is a figure which shows an example of sample data. It is a figure which shows the structural example of a management server. It is a figure which shows an example of the data registered into diagnostic criteria DB. It is a figure for demonstrating the data structure of substitute data item information. It is a figure which shows the structural example of a user terminal. It is a flowchart for demonstrating a similar case search process. It is a flowchart for demonstrating a query production | generation / search process. It is a figure for demonstrating a query production | generation / search process concretely. It is a flowchart for demonstrating a similarity score calculation process. It is a flowchart for demonstrating a proxy data item extraction process. It is a flowchart for demonstrating the processing operation of the similar case search system which concerns on embodiment of this invention. It is a flowchart for demonstrating the modification of a query production | generation / search process. It is a flowchart for demonstrating the modification of correspondence information.

Explanation of symbols

1 Case DB
2 Management Server 3 User Terminal 10 Network 21 Control Unit 211 Query Generation / Search Unit 212 Substitute Data Item Extraction Unit 22 Storage Unit 221 Diagnostic Criteria DB
222 Substitute data item DB
DESCRIPTION OF SYMBOLS 23 Input part 24 Display part 25 Communication control part 3 User terminal 31 Control part 32 Memory | storage part 33 Display part 34 Input part 35 Communication control part

Claims (12)

A similar case retrieval system for retrieving sample data similar to a patient's symptom from a case database in which sample data regarding each past case is registered,
A substitute data item database that stores substitute data item information including information on the item names of the data items and correspondence information indicating the correspondence between the data values of the data items for two data items having a high degree of correlation; ,
For a data item group consisting of a plurality of data items, accepting means for accepting data values related to the patient,
A specifying means for searching and specifying a data item registered in the substitute data item database in the data item group;
Using the correspondence information stored in the substitute data item database for the identified data item, the data value related to the patient is changed to the data item of the substitute data item that is the other data item having a high degree of correlation with the data item. Conversion means for converting to data values;
Query generation means for generating a query using the item name of the data item in the data item group and the data value related to the patient, and the item name and data value of the substitute data item obtained by converting the data value by the conversion means,
Search means for searching the case database for sample data that matches the generated query;
Output means for outputting a search result by the search means;
A similar case retrieval system comprising: For each disease, a diagnostic criteria database in which information identifying the disease and a data item that is a criterion for determining the disease are registered in association with each other,
Means for receiving information specifying a disease;
Means for searching a data item corresponding to the specified disease from a diagnostic criteria database, and specifying the searched data item as the data item group;
The similar case search system according to claim 1, further comprising: The query generation means includes a query in which an item name and a data value of each data item in the data item group are set as a search condition, and a part of the data item specified by the specifying means in the data item group or A query in which the item name and the data value of each data item in the data item group in the case of replacing all with the corresponding substitute data items are set as search conditions,
The similar case search system according to claim 1 or 2, wherein For data items constituting sample data registered in the case database, means for extracting data items two by two and calculating the degree of correlation between the extracted data items;
For two data items with the calculated correlation degree equal to or greater than a predetermined value, correspondence information indicating the correspondence between the data values of each data item is generated, the item name of each data item, the correspondence information, Means for generating the surrogate data item information comprising:
Means for registering the generated substitute data item information in the substitute data item database;
The similar case search system according to any one of claims 1 to 3, further comprising: The search means calculates the difference between the data value of the searched sample data and the data value related to the patient in each data item of the data item group, and uses the difference calculated for each data item, Similar score calculation means for calculating a similarity score indicating the degree of similarity between sample data and the patient's case,
The output means further comprises means for outputting the searched sample data based on the similarity score.
The similar case search system according to any one of claims 1 to 4, wherein: When there is a data item for which no data value is set in the retrieved sample data among the data items of the data item group, the similarity score calculation means calculates the data value of the substitute data item corresponding to the data item. The similarity score is extracted from the retrieved sample data, the data value of the substitute data item is converted into the data value of the data item corresponding to the substitute data item using the correspondence information in the substitute data item database Used to calculate
The similar case search system according to claim 5. On the computer,
An accepting step for accepting a data value related to a patient for a data item group comprising a plurality of data items;
Regarding a substitute data item database that stores substitute data item information including item name information of each data item and correspondence information indicating a correspondence relationship of data values of each data item with respect to two data items having a high degree of correlation , A specific step of searching and specifying a data item registered in the substitute data item database among the data item group,
Using the correspondence information stored in the substitute data item database for the identified data item, the data value related to the patient is changed to the data item of the substitute data item that is the other data item having a high degree of correlation with the data item. A conversion step to convert to a data value,
A query generation step of generating a query using the item name of the data item in the data item group and the data value related to the patient, and the item name and data value of the substitute data item obtained by converting the data value by the conversion unit;
A search step of searching for sample data matching the generated query from a case database in which sample data relating to past cases is registered;
Outputting a search result by the search means;
A program for running In the computer,
Receiving information specifying a disease;
The data item corresponding to the specified disease is searched from a diagnostic criteria database in which information specifying the disease for each disease and a data item that is a criterion for the disease are registered in association with each other, and the searched data item is Identifying as data items,
The program according to claim 7 for further executing The query generation step includes a query in which an item name and a data value of each data item in the data item group are set as a search condition, and a part of the data item specified by the specifying unit in the data item group or A query in which the item name and the data value of each data item in the data item group in the case of replacing all with the corresponding substitute data items are set as search conditions,
The program according to claim 7 or 8, characterized in that. For the data items constituting the sample data registered in the case database, extracting the data items two by two and calculating the degree of correlation between the extracted data items;
For two data items with the calculated correlation degree equal to or greater than a predetermined value, correspondence information indicating the correspondence between the data values of each data item is generated, the item name of each data item, the correspondence information, Generating the surrogate data item information comprising:
Registering the generated substitute data item information in the substitute data item database;
The program according to any one of claims 7 to 9, for further executing The search step calculates the difference between the data value of the searched sample data and the data value related to the patient in each data item of the data item group, and uses the difference calculated for each data item, A similarity score calculating step for calculating a similarity score indicating the degree of similarity between the sample data and the case of the patient;
The output step further includes an output step of outputting the searched sample data based on the similarity score.
The program according to any one of claims 7 to 10, wherein: When there is a data item for which no data value is set in the retrieved sample data among the data items of the data item group, the similarity score calculating step calculates the data value of the substitute data item corresponding to the data item. The similarity score is extracted from the retrieved sample data, the data value of the substitute data item is converted into the data value of the data item corresponding to the substitute data item using the correspondence information in the substitute data item database Used to calculate
The program according to claim 11.

Images