Assessment support system, assessment support method, and recording medium

Abstract

An assessment assistance system (Asys) comprises: an assessment prediction unit (122a) which, on the basis of patient information of a target patient for whom an assessment is created in a nursing care record, provides a prediction as a prediction assessment vector by converting the assessment of the target patient to an assessment vector; a similarity calculation unit (122b) which calculates the similarity of the assessment vector with respect to the prediction assessment vector on the basis of the relationship between the predicted prediction assessment vector and the assessment vector of a patient for whom the assessment is recorded in the nursing care record; and a search unit (122c) which searches for and outputs at least one similar patient that is similar to the target patient on the basis of the similarity.

Description

[Technical field]

[0001] The present invention relates to an assessment support system, an assessment support method, and a recording medium for supporting the creation of an assessment, which is one of the nursing processes related to a patient. [Background technology]

[0002] As a system for supporting medical examinations and treatment plans, a system is known that defines similarity using a symptom vector that quantifies a patient's symptoms, and detects past patients whose symptom vectors are similar to those of a target patient (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2003-122845 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, an assessment recorded as one item in a nursing record is required to include an analysis of the patient's nursing policy and a description of the thought process, and it is desirable that this takes into consideration not only the patient's symptoms, but also multifaceted information regarding nursing, such as the nursing system and medical facilities that can be provided to the patient. For this reason, the contents of the assessment of a second patient that is similar to the first patient for whom an assessment is being created in terms of symptom vectors are not necessarily useful as a reference for the assessment to be written for the first patient. Therefore, the above system has low search performance for searching for a second patient similar to the first patient to support the creation of an assessment for the first patient, and may not be sufficiently reliable as an assessment creation support.

[0005] An object of the present invention is to provide an assessment support system, an assessment support method, and a recording medium that are highly reliable in supporting the creation of assessments. [Means for solving the problem]

[0006] One aspect of the assessment support system of the present invention comprises an assessment prediction unit that predicts, as a predicted assessment vector, an assessment vector obtained by vectorizing the assessment of a target patient for which an assessment is to be created in a nursing record based on patient information of the target patient, a similarity calculation unit that calculates the similarity of the assessment vector to the predicted assessment vector based on the relationship between the predicted predicted assessment vector and the assessment vector of a patient whose assessment is recorded in the nursing record, and a search unit that searches for and outputs at least one similar patient that is similar to the target patient based on the similarity.

[0007] One aspect of the assessment support method of the present invention involves having at least one computer predict, based on patient information of a target patient for whom an assessment is to be created in a nursing record, an assessment vector obtained by vectorizing the assessment of the target patient as a predicted assessment vector, calculate the similarity of the assessment vector to the predicted assessment vector based on the relationship between the predicted predicted assessment vector and the assessment vector of a patient whose assessment is recorded in the nursing record, and search for and output at least one similar patient that is similar to the target patient based on the similarity.

[0008] One embodiment of the recording medium of the present invention has recorded thereon a computer program configured to cause at least one computer to function as an assessment prediction unit that predicts, as a predicted assessment vector, an assessment vector obtained by vectorizing the assessment of a target patient for which an assessment in a nursing record is to be created, based on patient information of the target patient; a similarity calculation unit that calculates the similarity of the assessment vector to the predicted assessment vector based on the relationship between the predicted predicted assessment vector and the assessment vector of a patient whose assessment is recorded in the nursing record; and a search unit that searches for and outputs at least one similar patient that is similar to the target patient based on the similarity.

[0009] According to one aspect of each of the above-mentioned assessment support system, assessment support method, and recording medium, the search performance for patients whose assessments are similar to the target patient for whom the assessment is being created is improved, making it possible to provide highly reliable support for creating assessments. [Brief description of the drawings]

[0010] [Figure 1] 1 is a diagram showing an example of an overall configuration for realizing an assessment support system according to the present invention. [Diagram 2] FIG. 4 is a diagram illustrating an example of a hardware configuration of a model generation server. [Diagram 3] FIG. 2 is a diagram showing an example of a hardware configuration of a search server. [Figure 4] FIG. 2 is a diagram showing an example of the hardware configuration of a model generation terminal and a search terminal. [Diagram 5] FIG. 1 is a diagram illustrating the concept of similarity. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] 1. Overall structure FIG. 1 shows an example of the overall configuration for realizing the assessment support system Asys according to the present invention. The assessment support system Asys is a system that supports medical professionals (typically nurses, nursing students, etc., hereinafter referred to as "users") in creating assessment information. Here, "assessment" refers to a method included in "SOAP," which is one of the analysis methods for nursing records recorded to care for patients. Notes In "SOAP," "S" stands for the patient's subjective information, "O" for objective information, "A" for "assessment," and "P" for nursing plan. "Assessment" includes judgments and evaluations arrived at by analyzing subjective information (i.e., "S") and objective information (i.e., "O"), as well as opinions and impressions. Hereinafter, the written information entered as "assessment" in nursing records will be referred to as "assessment information."

[0012] The assessment support system Asys may be composed of, for example, a system server unit 10 and a system terminal unit 20 capable of transmitting and receiving data, as shown in FIG. 1. The data may be transmitted and received via a predetermined network or directly. The system server unit 10 may include, for example, a model generation server 11 and a search server 12. The system terminal unit 20 may include, for example, a model generation terminal 21 and a search terminal 22. The model generation terminal 21 can access the model generation server 11, and the model generation terminal 21 and the model generation server 11 may constitute a model generation system Msys. The model generation system Msys is, for example, a system that generates a prediction model described later in response to a user's operation on the model generation terminal 21. The search terminal 22 can access the search server 12, and the search terminal 22 and the search server 12 may constitute a search system Dsys. The search system Dsys is, for example, a system that searches for similar patients described later in response to a user's operation on the search terminal 22 and presents them to a user.

[0013] The model generation server 11 and the search server 12 may be able to access a database 30 that can be accessed from other systems, for example. The database 30 may store various data that can be generated and used in other systems (for example, electronic medical record systems, etc.), such as patient information about multiple patients. The patient information is information recorded about each patient (i.e., known information), and may include, for example, medical records, test data, and nursing records created by doctors about each patient. The patient information may also include, for example, assessments created and recorded in the past about the corresponding patient. The patient information may be associated with, for example, patient identification information for identifying each patient. In this system, for example, each patient may be identified by the patient identification information in each process. Hereinafter, a patient for whom an assessment has been created in the past and recorded in a nursing record (i.e., assessment information exists) is referred to as a "past patient," and a patient for whom an assessment is to be created (i.e., the patient for whom the assessment is to be created) is referred to as a "target patient."

[0014] The system server unit 10 may be configured as a so-called cloud server physically composed of a plurality of server devices. Alternatively, the system server unit 10 may be physically composed of a single server device. The model generation terminal 21 and the search terminal 22 may be integrated with the model generation server 11 and the search server 12, respectively, as described later. Also, although one model generation terminal 21 and one search terminal 22 are shown in FIG. 1, there may be a plurality of each. Alternatively, the model generation terminal 21 and the search terminal 22 may be physically one terminal.

[0015] 2. Model generation server 2 shows an example of a hardware configuration of the model generation server 11. The model generation server 11 may have, for example, a storage device 111, a computing device 112, and a communication interface 114. The storage device 111, the computing device 112, and the communication interface 114 may be connected via a data bus 115 so as to be able to transmit and receive data.

[0016] The storage device 111 can store desired data. For example, the storage device 111 may temporarily store a computer program executed by the arithmetic device 112. The storage device 111 may temporarily store data that the arithmetic device 112 temporarily uses when the arithmetic device 112 executes a computer program. The storage device 111 may store data that the model generation server 11 stores long-term. For example, the storage device 111 may include a model storage unit 111a that stores a generated prediction model, and a vector storage unit 111b that stores assessment vectors (described later) related to past patients. The storage device 111 may include at least one of a random access memory (RAM), a read only memory (ROM), a hard disk device, a magneto-optical disk device, a solid state drive (SSD), and a disk array device. That is, the storage device 111 may include a volatile recording medium and a non-volatile recording medium.

[0017] The arithmetic device 112 includes, for example, a CPU (Central Processing Unit). The arithmetic device 112 reads a computer program. For example, the arithmetic device 112 may read a computer program stored in the storage device 111. For example, the arithmetic device 112 may read a computer program that is computer-readable and stored in a non-volatile recording medium, using a recording medium reading device (not shown). The arithmetic device 112 may obtain (i.e., download or read) a computer program from a device (not shown) disposed outside the model generation server 11 via the communication interface 114.

[0018] The arithmetic device 112 executes the read computer program. As a result, a logical function block for executing the operation performed by the model generation server 11 is realized in the arithmetic device 112. In other words, the arithmetic device 112 can function as a controller for realizing a logical function block for executing the operation performed by the model generation server 11. FIG. 2 shows an example of a logical function block realized in the arithmetic device 112 for realizing each process to be executed by the model generation server 11. As shown in FIG. 2, an assessment mapping unit 112a and a prediction model learning unit 112b are realized in the arithmetic device 112. The arithmetic device 112 acquires necessary data for the process related to each unit 112a-112b, for example, via the communication interface 114 and the data bus 115. The operation of each unit 112a-112b (i.e., the process executed by each unit 112a-112b) will be described in detail later.

[0019] The communication interface 114 enables access to other devices, such as the model generation terminal 21, the database 30, other servers in the system server unit 10 (e.g., the search server 12), and external systems, via communication lines, and enables sending and receiving various data with the other devices.

[0020] 3. Search Server 3 shows an example of the hardware configuration of the search server 12. The search server 12 may include, for example, a storage device 121, a computing device 122, and a communication interface 124. The storage device 121, the computing device 122, and the communication interface 124 may be connected via a data bus 125 so as to be able to transmit and receive data.

[0021] The storage device 121 can store desired data. For example, the storage device 121 may temporarily store a computer program executed by the arithmetic device 122. The storage device 121 may temporarily store data that the arithmetic device 122 temporarily uses when the arithmetic device 122 executes a computer program. The storage device 121 may store data that the search server 12 stores for a long period of time. For example, the storage device 121 may include a similarity storage unit 121a that stores similarities described later. The storage device 121 may include at least one of a RAM, a ROM, a hard disk device, a magneto-optical disk device, an SSD, and a disk array device. In other words, the storage device 121 may include a volatile recording medium and a non-volatile recording medium.

[0022] The arithmetic device 122 includes, for example, a CPU. The arithmetic device 122 reads a computer program. For example, the arithmetic device 122 may read a computer program stored in the storage device 121. For example, the arithmetic device 122 may read a computer program that is computer-readable and stored in a non-volatile recording medium using a recording medium reading device (not shown). The arithmetic device 122 may obtain (i.e., download or read) a computer program from a device (not shown) disposed outside the search server 12 via the communication interface 124.

[0023] The arithmetic device 122 executes the read computer program. As a result, a logical function block for executing the operation performed by the search server 12 is realized in the arithmetic device 122. In other words, the arithmetic device 122 can function as a controller for realizing the logical function block for executing the operation performed by the search server 12. FIG. 3 shows an example of a logical function block realized in the arithmetic device 122 for realizing each process to be executed by the search server 12. As shown in FIG. 3, for example, an assessment prediction unit 122a, a similarity calculation unit 122b, a search unit 122c, and an output control unit 122d are realized in the arithmetic device 122. The arithmetic device 122 acquires necessary data for the process related to each unit 122a-122d, for example, via the communication interface 124 and the data bus 125. The operation of each unit 122a-122d (i.e., the process executed by each unit 122a-122d) will be described in detail later.

[0024] The communication interface 124 enables access to other devices, such as the search terminal 22, the database 30, other servers in the system server section 10 (e.g., the model generation server 11), and external systems, via communication lines, and enables sending and receiving various data with the other devices.

[0025] 4. Model generation terminal The model generation terminal 21 functions as an interface of the model generation server 11 for the user. The model generation terminal 21 operates in response to an input operation by the user and an instruction from the model generation server 11. An example of the hardware configuration of the model generation terminal 21 is shown in FIG. 4. The model generation terminal 21 may have, for example, a storage unit 211, a calculation unit 212, an input unit 213, an output unit 214, and a communication interface 215. The storage unit 211, the calculation unit 212, the input unit 213, the output unit 214, and the communication interface 215 may be connected via a data bus 216 so as to be capable of transmitting and receiving data. The model generation terminal 21 may be provided so as to be integrated with the model generation server 11. In this case, for example, the input unit 213 and the output unit 214 may be connected to the data bus 115 of the model generation server 11. Furthermore, the functions of the storage unit 211 and the calculation unit 212 may be realized by, for example, the storage device 111 and the calculation device 112 of the model generation server 11.

[0026] The storage unit 211 can store desired data. For example, the storage unit 211 may temporarily store a computer program executed by the calculation unit 212. The storage unit 211 may temporarily store data that is temporarily used by the calculation unit 212 when the calculation unit 212 is executing a computer program. The storage unit 211 may store data that is to be stored by the model generation terminal 21 for a long period of time. The storage unit 211 may include at least one of a RAM, a ROM, a hard disk device, a magneto-optical disk device, an SSD, and a disk array device. In other words, the storage unit 211 may include a volatile recording medium and a non-volatile recording medium.

[0027] The calculation unit 212 includes, for example, a CPU. The calculation unit 212 reads a computer program. For example, the calculation unit 212 may read a computer program stored in the storage unit 211. For example, the calculation unit 212 may read a computer program that is computer-readable and stored in a non-volatile recording medium using a recording medium reading device (not shown). The calculation unit 212 may acquire (that is, download or read) a computer program from a device (not shown) disposed outside the model generation terminal 21 via the communication interface 215. The calculation unit 212 executes the read computer program. As a result, a logical function block for executing the operation performed by the model generation terminal 21 is realized in the calculation unit 212. That is, the calculation unit 212 can function as a controller for realizing a logical function block for executing the operation performed by the model generation terminal 21.

[0028] The input unit 213, for example, accepts an input operation from the outside (for example, an input operation to the model generation terminal 21 by a user) and sends operation information to the calculation unit 212 so that processing according to the input operation is performed. The calculation unit 212 may, for example, transmit the operation information transmitted from the input unit 213 to the model generation server 11 via the communication interface 215. In addition, the calculation unit 212 may, for example, perform processing according to the operation information transmitted from the input unit 213 and transmit the processing result to the model generation server 11 via the communication interface 215. The input mode in the input unit 213 may be any of key input, voice input, touch input, button input, and the like. The input unit 213 may include, for example, a keyboard, a mouse, a touch panel, a microphone, and a button.

[0029] The output unit 214 appropriately outputs, for example, instructions and various information transmitted from the model generation server 11 in a manner that the user can confirm. In addition, the output unit 214 may appropriately output, for example, a processing result processed by the calculation unit 212 in a manner that the user can confirm. The output mode of the output unit 214 may be visual output, audible output, data output, or the like. The output unit 214 may include, for example, a screen, a speaker, a storage medium, or the like. The communication interface 215 accesses the model generation server 11 via, for example, a communication line, and enables transmission and reception of various data to and from the model generation server 11.

[0030] 5. Search terminal The search terminal 22 functions as an interface for the user with the search server 12. The search terminal 22 operates in response to input operations by the user and instructions from the search server 12. The hardware configuration of the search terminal 22 may be similar to that of the model generation terminal 21, except that in the above description of the model generation terminal 21, "model generation terminal 21" corresponds to "search terminal 22" and "model generation server 11" corresponds to "search server 12". Therefore, a description of the hardware configuration of the search terminal 22 will be omitted.

[0031] 6. Model Generation System Returning to FIG. 2, the processing performed in the model generation system Msys will be described, focusing on the operations of the assessment mapping unit 112a and the prediction model learning unit 112b realized in the arithmetic device 112 of the model generation server 11.

[0032] The assessment mapping unit 112a may perform a vector generation process in response to a vector generation operation received by the input unit 213 of the model generation terminal 21, for example. The vector generation operation may be performed, for example, when one or more assessments are newly created or updated. As the vector generation process, the assessment mapping unit 112a may, for example, vectorize an actual sentence (i.e., assessment information) created as an assessment in a nursing record, and map it as an assessment vector in a multi-dimensional assessment space. The assessment mapping unit 112a may acquire the assessment information, for example, from the nursing record held by the database 30. The assessment mapping unit 112a may use an existing method for vectorizing a sentence in order to obtain an assessment vector from the assessment information. For example, BoW (Bag of Words), Word2vec, etc. may be adopted as the existing method. The assessment mapping section 112a may, for example, store an assessment vector obtained by vectorizing the assessment information for each past patient in the vector storage section 111b of the storage device 111 in association with, for example, the patient identification information.

[0033] The prediction model learning unit 112b may perform a model learning process following the vector generation process. Alternatively, the prediction model learning unit 112b may perform a model learning process in response to a model learning operation received by the input unit 213 of the model generation terminal 21. The prediction model learning unit 112b may learn a prediction model for predicting an assessment vector of a target patient as a model learning process. Hereinafter, an assessment vector predicted for a target patient is referred to as a "predicted assessment vector". Note that the assessment vector obtained by the assessment mapping unit 112a and the predicted assessment vector may be vectors in the same assessment space. The prediction model may be, for example, a machine learning model for outputting a predicted assessment vector when patient information on a target patient is input. The prediction model learning unit 112b may learn a structure of a prediction model using a teacher data set in which patient information of a past patient is input as input data and the assessment vector of the past patient is correct answer data as a model learning process. The prediction model learning unit 112b may, for example, obtain the assessment vector of each past patient to be used as the correct answer data from the vector storage unit 111b, and obtain the patient information of each past patient to be used as the input data from the database 30. The prediction model learning unit 112b may, for example, adjust the parameter value of the prediction model so that the difference between the predicted assessment vector output by the prediction model based on the patient information as the input data and the assessment vector as the correct answer data becomes small, that is, so that the output predicted assessment vector becomes closer to the assessment vector of the correct answer data. The prediction model learning unit 112b may, for example, store the generated prediction model in the model storage unit 111a of the storage device 111.

[0034] Here, it is desirable that the "patient information" in this system includes at least one of physical information related to the body, disease information related to the disease, and past assessment information. The "physical information" may include, for example, vital signs, age, height, and weight. The "disease information" may include, for example, the name of the disease and the stage of the disease. The "past assessment information" is information related to assessment information created in the past, and may be, for example, the assessment information itself or an assessment vector. It is known that the consistency between patients regarding the above three elements, among the information included in the patient information, is particularly related to the similarity of the assessment information. For example, it is known that the higher the degree of consistency regarding the above three elements, the more similar the assessment information is.

[0035] 7. Search System Next, the processing performed in the search system Dsys will be described, focusing on the operations of the assessment prediction unit 122a, the similarity calculation unit 122b, the search unit 122c, and the output control unit 122d realized in the arithmetic device 122 of the search server 12 shown in Fig. 3. Each of the assessment prediction unit 122a, the similarity calculation unit 122b, the search unit 122c, and the output control unit 122d may perform the following processing in this order, for example, in response to a search operation received by the input unit 213 of the search terminal 22. The search operation may be performed, for example, at a timing according to the user's needs.

[0036] The assessment prediction unit 122a predicts an assessment vector of a target patient as a predicted assessment vector from, for example, patient information on the target patient. The prediction by the assessment prediction unit 122a may be performed, for example, by a prediction model learned by the prediction model learning unit 112b. The assessment prediction unit 122a may use, for example, a prediction model stored in the model storage unit 111a for the prediction. The assessment prediction unit 122a may acquire patient information on the target patient from, for example, the database 30.

[0037] The similarity calculation unit 122b may calculate the similarity of the assessment vector of each past patient to the predicted assessment vector based on the relationship between the predicted assessment vector in the assessment space and the assessment vector of each past patient. The similarity calculation unit 122b may have, for example, a vector acquisition unit 122bb that acquires the assessment vector of the past patient. The vector acquisition unit 122bb may acquire, for example, all or some of the assessment vectors of the past patients from among the assessment vectors stored in the vector storage unit 111b. Alternatively, the vector acquisition unit 122bb may acquire the assessment vector of the past patient by, for example, performing the above-mentioned vector generation process (or having the assessment mapping unit 112a perform the process). The similarity calculation unit 122b may store, for example, the similarity calculated for each past patient in the similarity storage unit 121a of the storage device 121 in association with, for example, patient identification information. The similarity calculation section 122b may calculate the similarity based on, for example, the Minkowski distance (including the Manhattan distance, the Euclidean distance, and the Chebyshev distance), the cosine similarity, or the like.

[0038] The concept of similarity in the relationship between the predicted assessment vector of a target patient and the assessment vectors of past patients is shown diagrammatically in Figure 5. The patient information space on the left side of Figure 5 is a space in which the patient information of the target patient is vectorized into an n-dimensional (two-dimensional in Figure 5) patient information vector. As shown in Figure 5, in the case of a two-dimensional patient information space, for example, the X-axis element is vital signs, value,The Y-axis element may be set to age. Position X in the patient information space indicates a patient information vector obtained from the patient information of the target patient, i.e., a position corresponding to the patient information of the target patient in the patient information space. On the other hand, the right side of FIG. 5 shows an n-dimensional (two-dimensional in FIG. 5) assessment space. Position Yp in the assessment space indicates a predicted assessment vector value predicted from the patient information of the target patient, i.e., a position corresponding to the predicted assessment information of the target patient in the assessment space. Position Yi (i=1, 2, 3...) in the assessment space indicates an assessment vector value obtained from the patient information of a past patient, i.e., a position corresponding to the assessment information of the past patient in the assessment space. As shown in FIG. 5, the similarity may be a positional relationship (distance in FIG. 5) α between position Yp, which is a predicted assessment vector value in the assessment space, and position Yi, which is an assessment vector value of the past patient. It may be determined that the closer or smaller the positional relationship α is, the higher the similarity is. In the example shown in FIG. 5, position Y1 is closer to position Yp than position Y2. Therefore, the similarity of the assessment vector value Y1 may be determined to be higher than the similarity of the assessment vector value Y2.

[0039] The similarity in the present system may be a composite similarity calculated by taking into consideration not only the predicted assessment vector of the target patient and the assessment vector of the past patient, but also the patient information of the target patient and the patient information of the past patient. For example, the composite similarity may be calculated as the similarity in the assessment space + the similarity in the patient information space. The method used for the similarity in the patient information space may be the same as that for the similarity in the assessment space, or may be a different method.

[0040] The similarity calculation unit 122b may calculate the composite similarity so that the similarity in the assessment space is adjusted by the patient information. The similarity calculation unit 122b may use, for example, assessment-related information (for example, at least one of physical information, disease information, and past assessment information) that is known to be significantly related to the similarity of the assessment information among the patient information. The similarity calculation unit 122b may calculate the composite similarity so that the higher the degree of agreement of the assessment-related information, the higher the similarity in the assessment space. For example, the similarity calculation unit 122b may set the similarity to the lowest level (for example, infinity when the similarity is distance) when at least a part of the assessment-related information does not match. The similarity calculation unit 122b may calculate the composite similarity by setting a weight according to the information included in the assessment-related information.

[0041] Returning to FIG. 3, the search unit 122c searches for similar patients similar to the target patient from among past patients, for example, based on the similarity calculated by the similarity calculation unit 122b. For example, the search unit 122c may sort the similarity calculated for each past patient in descending order and output them as similar patients. The search unit 122c may search for a predetermined number of past patients (for example, 10 patients) in descending order of similarity, and identify them as similar patients. Alternatively, the search unit 122c may identify all past patients with a similarity equal to or higher than a threshold value (i.e., past patients with a similarity equal to or higher than a predetermined value) as similar patients. The search unit 122c performs the search, for example, by referring to the similarity storage unit 121a. The search unit 122c may identify similar patients based on, for example, the above-mentioned judgment of the level of similarity. The number of patients to be identified as similar patients and the threshold similarity may be appropriately set by a user (for example, a user operating the search terminal 22). The search unit 122c may, for example, select a plurality of provisionally similar patients based only on the similarity in the assessment space, and further narrow down the selected provisionally similar patients to those who satisfy a predetermined condition based on patient information (for example, assessment-related information that is significantly related to the similarity of the assessment information (for example, at least one of physical information, disease information, and past assessment information)), and identify the patient who satisfies the predetermined condition as a similar patient. The "predetermined condition" may, for example, be set so that provisionally similar patients with a relatively high degree of agreement in assessment-related information are relatively more likely to be identified as similar patients.

[0042] The output control unit 122d presents the similar patients identified by the search unit 122c to the user of the search terminal 22. For example, the output control unit 122d presents similar patients (for example, 10 patients with the highest similarity) to the user of the search terminal 22 as a similar patient list. For example, the output control unit 122d may cause the similar patient list to be displayed on the output unit 214 (for example, a screen) of the search terminal 22. For example, the output control unit 122d may cause the user to select at least one similar patient from the displayed similar patient list via the input unit 213 of the search terminal 22. When at least one similar patient is selected, the output control unit 122d may, for example, obtain assessment information of each selected similar patient from the database 30, and display it on the screen of the search terminal 22 and / or download it to the search terminal 22. This allows the user to select assessment information similar to the assessment information to be entered for the target patient. reference It becomes possible to do so.

[0043] As described above, in this system, one physical server device may function as the model generation server 11 and the search server 12. Furthermore, each of the model generation server 11 and the search server 12 may be realized by a plurality of server devices. A common storage device common to the model generation server 11 and the search server 12 may be provided, and data generated by the model generation server 11 and used by the search server 12 (e.g., assessment vectors of past patients, etc.) may be stored in this common storage device.

[0044] This system employs the concept of assessment vectors in which assessments in nursing records are vectorized in a multi-dimensional assessment space, and detects similar patients based on the similarity obtained based on the relationship between assessment vectors obtained from actual assessments created in the past and predicted assessment vectors predicted for the target patient. Therefore, the contents of the assessment of a similar patient with a high degree of similarity in this system are more likely to be similar to the contents that should be written in the assessment of the target patient than, for example, the contents of the assessment of a patient who only has similar symptoms. In other words, with regard to the contents of the assessment, the detected similar patients have a higher similarity to the target patient, and this system can provide a highly reliable support for creating assessments. According to this system, it is possible to predict the predicted assessment vector of the target patient using a prediction model. The prediction model is trained using actual patient information and assessment vectors obtained from assessments actually made in the past. Therefore, according to this system, it is possible to predict the predicted assessment vector with high accuracy. According to this system, the similarity obtained only from the relationship between the assessment vector and the predicted assessment vector can be adjusted by the relationship between the patient information on the target patient and the patient information on the patient corresponding to the assessment vector. For example, if the consistency of information related to the assessment similarity among the patient information is taken into consideration, it is possible to further improve the search ability for similar patients. For example, it is preferable that the patient information taken into consideration when calculating the similarity includes at least one of physical information, disease information, and past assessment information. These three elements are known to be elements that are closely related to the similarity of assessments. Therefore, by calculating the similarity while taking these elements into consideration, it is possible to further improve the ability to search for similar patients. The present system can output similar patients in descending order of similarity. This allows the user to more quickly recognize similar patients with high similarity in terms of assessment, improving the efficiency of the search for similar patients. Additionally / alternatively, the present system can output similar patients with a similarity equal to or greater than a predetermined value. This makes it possible to omit output processing for patients whose similarity is not the target, thereby reducing the output processing burden. Furthermore, this system, which employs the concept of assessment vectors in which assessments in nursing records are vectorized in a multi-dimensional assessment space (i.e., mapped to a multi-dimensional assessment space), may be configured to generate assessment vectors from assessments actually created in the past (assessment mapping unit), and to learn a prediction model for predicting the predicted assessment vector of a target patient for whom an assessment is to be created, using the generated assessment vectors (prediction model learning unit). This makes it possible to make predictions based on the assessments that have actually been created. This makes it possible to further improve the accuracy of predictions and provide a highly reliable assessment support system. Each process in the present system may be provided as a method to be executed by at least one computer. This makes it possible, for example, to distribute the processing load. Furthermore, a computer program for realizing each process in the present system may be provided as a recording medium on which the computer program is recorded. This makes it easy, for example, to sell or update the computer program related to the present system.

[0045] The present invention can be modified as appropriate within the scope that does not contradict the gist or concept of the invention that can be read from the claims and the entire specification, and the assessment support system, assessment support method, and recording medium that include such modifications are also included in the technical concept of the present invention.

[0046] <Additional Notes> The following supplementary notes are further disclosed regarding the above-described embodiment.

[0047] (Appendix 1) The assessment support system described in Appendix 1 is an assessment support system comprising: an assessment prediction unit that predicts, as a predicted assessment vector, an assessment vector obtained by vectorizing the assessment of a target patient for which an assessment is to be created in a nursing record, based on patient information of the target patient; a similarity calculation unit that calculates the similarity of the assessment vector to the predicted assessment vector based on the relationship between the predicted predicted assessment vector and the assessment vector of a patient whose assessment is recorded in the nursing record; and a search unit that searches for and outputs at least one similar patient that is similar to the target patient based on the similarity. (Appendix 2) The assessment support system described in Appendix 2 is the assessment support system described in Appendix 1, wherein the assessment prediction unit predicts the predicted assessment vector from the patient information of the target patient using a prediction model that outputs the predicted assessment vector based on the input patient information, and the prediction model is a model trained using the patient information and the assessment vector of the patient. (Appendix 3) The assessment support system described in Appendix 3 is the assessment support system described in Appendix 1 or 2, wherein the similarity calculation unit calculates the similarity using patient information of the target patient and the patient information of the patient. (Appendix 4) The assessment support system described in Appendix 4 is the assessment support system described in any one of Appendixes 1 to 3, wherein the patient information includes at least one of physical information regarding the body of each patient, disease information regarding the disease, and past assessment information regarding the assessment created in the past. (Appendix 5) The assessment support system described in Appendix 5 is the assessment support system described in Appendix 3 or 4, wherein the similarity calculation unit calculates the similarity so that the higher the degree of match between the patient information of the target patient and the patient information of the patient, the higher the similarity. (Appendix 6) The assessment support system described in Supplementary Note 6 is the assessment support system described in any one of Supplementary Notes 1 to 5, wherein the search unit outputs, as the similar patient, the patients from among the plurality of patients in descending order of the degree of similarity. (Appendix 7) The assessment support system described in Appendix 7 is the assessment support system described in any one of Appendixes 1 to 6, wherein the search unit outputs, among the multiple patients, a patient whose similarity is a predetermined value or greater as the similar patient. (Appendix 8) The assessment support system described in Appendix 8 is an assessment support system that includes an assessment mapping unit that maps an assessment recorded in a patient's nursing record into an assessment space as an assessment vector, and a prediction model learning unit that, when patient information of a target patient for which the assessment is to be created is input, learns a prediction model that predicts the assessment vector of the target patient as a predicted assessment vector, using the assessment vector of the patient and the patient information of the patient as a teacher dataset. (Appendix 9) The assessment support method described in Appendix 9 is an assessment support method which causes at least one computer to predict, as a predicted assessment vector, an assessment vector obtained by vectorizing the assessment of a target patient for which an assessment is to be created in a nursing record, based on patient information of the target patient, calculate a similarity of the assessment vector to the predicted assessment vector based on the relationship between the predicted predicted assessment vector and the assessment vector of a patient whose assessment is recorded in the nursing record, and search for and output at least one similar patient that is similar to the target patient based on the similarity. (Appendix 10) The recording medium described in Appendix 10 is a recording medium having recorded thereon a computer program configured to cause at least one computer to function as an assessment prediction unit that predicts, as a predicted assessment vector, an assessment vector obtained by vectorizing the assessment of a target patient for which an assessment in a nursing record is to be created, based on patient information of the target patient; a similarity calculation unit that calculates the similarity of the assessment vector to the predicted assessment vector based on the relationship between the predicted predicted assessment vector and the assessment vector of a patient whose assessment is recorded in the nursing record; and a search unit that searches for and outputs at least one similar patient similar to the target patient based on the similarity. [Explanation of symbols]

[0048] 11 Model Generation Server 112a Assessment Mapping Section 112b Prediction model learning section 12 Search Server 122a Assessment and Prediction Section 122b Similarity calculation part 122c Search Department 122d Output control section Asys assessment support system

Claims

1. an assessment prediction unit that predicts, as a predicted assessment vector, an assessment vector obtained by vectorizing the assessment of a target patient based on patient information of the target patient for which an assessment in a nursing record is to be created; a similarity calculation unit that calculates a similarity of the assessment vector to the predicted assessment vector based on a relationship between the predicted assessment vector and the assessment vector of the patient whose assessment is recorded in the nursing record; a search unit that searches for and outputs at least one similar patient similar to the target patient based on the similarity; An assessment support system equipped with the above.

2. The assessment prediction unit predicts the predicted assessment vector from the patient information of the target patient using a prediction model that outputs the predicted assessment vector based on the input patient information; The prediction model is a model trained using the patient information and the assessment vector of the patient. The assessment support system according to claim 1 .

3. The assessment support system according to claim 1 , wherein the similarity calculation unit calculates the similarity using the patient information of the target patient and the patient information of the patient.

4. An assessment support system as described in any one of claims 1 to 3, wherein the patient information includes at least one of physical information regarding the body of each patient, disease information regarding the disease, and past assessment information regarding the assessment created in the past.

5. The assessment support system according to claim 3 , wherein the similarity calculation unit calculates the similarity such that the higher the degree of agreement between the patient information of the target patient and the patient information of the patient, the higher the similarity becomes.

6. The assessment support system according to claim 1 , wherein the search unit outputs, as the similar patient, the patients from among the plurality of patients in descending order of the degree of similarity.

7. The assessment support system according to claim 1 , wherein the search unit outputs, from among the plurality of patients, a patient whose similarity is equal to or greater than a predetermined value as the similar patient.

8. an assessment mapping unit that maps an assessment recorded in a patient's nursing record into an assessment space as an assessment vector; a prediction model learning unit that, when patient information of a target patient for which the assessment is to be created is input, learns a prediction model that predicts the assessment vector of the target patient as a predicted assessment vector, using the assessment vector of the patient and the patient information of the patient as a teacher data set; An assessment support system equipped with the above.

9. At least one computer Based on patient information of a target patient for which an assessment in a nursing record is to be created, an assessment vector obtained by vectorizing the assessment of the target patient is predicted as a predicted assessment vector; Calculating a similarity of the assessment vector to the predicted assessment vector based on a relationship between the predicted predicted assessment vector and the assessment vector of the patient whose assessment is recorded in the nursing record; The assessment support method further comprises searching for and outputting at least one similar patient who is similar to the target patient based on the degree of similarity.

10. At least one computer, an assessment prediction unit that predicts, as a predicted assessment vector, an assessment vector obtained by vectorizing the assessment of a target patient for which an assessment in a nursing record is to be created, based on patient information of the target patient; a similarity calculation unit that calculates a similarity of the assessment vector to the predicted assessment vector based on a relationship between the predicted assessment vector and the assessment vector of the patient whose assessment is recorded in the nursing record; and A computer program configured to function as a search unit that searches for and outputs at least one similar patient who is similar to the target patient based on the similarity.

Images