Assessment evaluation device, assessment evaluation method, and program

Abstract

An assessment evaluation device 1X mainly has a written word vector generation means 17X, an ideal word vector generation means 15X, and an evaluation means 18X. On the basis of a nursing assessment for a target patient for whom a nursing record is to be created, the written word vector generation means 17X generates a written word vector, which is a word vector obtained by vectorizing the presence or absence of writing of a predetermined candidate word in the nursing assessment. The ideal word vector generation means 15X generates an ideal word vector, which is an ideal word vector based on a patient condition of the target patient. The evaluation means 18X decides an evaluation of the nursing assessment on the basis of the written word vector and the ideal word vector.

Description

[Technical field]

[0001] The present disclosure relates to the technical fields of an assessment evaluation device, an assessment evaluation method, and a storage medium for evaluating an assessment (nursing assessment), which is part of a nursing record. [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] Assessment, which is recorded as one item in the nursing record, requires an analysis of the patient's nursing policy and a description of the thought process, and it is difficult for inexperienced nurses to record assessments accurately. Therefore, in order to support the creation of appropriate nursing records, a system is needed to evaluate whether assessments have been recorded correctly.

[0005] In view of the above-mentioned problems, an object of the present disclosure is to provide an assessment evaluation device, an assessment evaluation method, and a storage medium capable of evaluating an assessment. [Means for solving the problem]

[0006] One aspect of the assessment evaluation device is A description word vector generating means for generating a description word vector, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in a nursing assessment based on a nursing assessment of a target patient for which a nursing record is to be created; an ideal word vector generating means for generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; An evaluation means for determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector; The present invention is an assessment evaluation device having the above structure.

[0007] One aspect of the assessment method includes: The computer Based on the nursing assessment of a target patient for which a nursing record is to be created, a description word vector is generated, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in the nursing assessment; generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector; It is an assessment evaluation method.

[0008] program One aspect of the present invention is Based on the nursing assessment of a target patient for which a nursing record is to be created, a description word vector is generated, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in the nursing assessment; generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; A program for causing a computer to execute a process for determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector. M It is. Effect of the Invention

[0009] According to the present disclosure, a nursing assessment rating for a subject patient can be determined. [Brief description of the drawings]

[0010] [Figure 1] 1 shows a schematic configuration of an assessment support system according to a first embodiment. [Diagram 2] 1A is a diagram showing an outline of the process performed by the assessment evaluation device, and FIG. 1B is a diagram showing an outline of the process performed by a regression model. [Diagram 3] (A) shows the hardware configuration of the assessment device, and (B) shows the hardware configuration of the learning device. [Figure 4] 2 is an example of a functional block of an assessment evaluation device. [Diagram 5] 2 is an example of a functional block of a learning device. [Figure 6] FIG. 2 is a diagram showing an outline of the processing of an assessment evaluation device in a specific example. [Figure 7] 1 is a first display example regarding the evaluation of an assessment. [Figure 8] 2 is a second example of displaying the evaluation of an assessment. [Figure 9] 3 is a third example of displaying the assessment evaluation. [Figure 10] 4 is an example of a flowchart executed by the assessment evaluation device in the first embodiment. [Figure 11] 1 shows a schematic configuration of an assessment support system according to a second embodiment. [Figure 12] 2 is an example of a functional block of an assessment evaluation device. [Figure 13] 13 is an example of a flowchart executed by the assessment evaluation device in the second embodiment. [Figure 14] 13 shows a schematic configuration of an assessment support system according to a third embodiment. [Figure 15] FIG. 13 is a block diagram of an assessment evaluation device according to a fourth embodiment. [Figure 16]13 is an example of a flowchart executed by the assessment evaluation device in the fourth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, embodiments of an assessment evaluation device, an assessment support method, and a storage medium will be described with reference to the drawings.

[0012] First Embodiment (1) System Configuration FIG. 1 shows a schematic configuration of an assessment support system 100 according to the first embodiment. The assessment support system 100 is a system that supports the creation of an assessment (specifically, a nursing assessment) by a medical professional (typically a nurse or a nursing student, hereinafter also referred to as a "user"). Here, an "assessment" is a description item included in "SOAP", which is one of the analysis methods of nursing records recorded to care for patients. In "SOAP", "S" indicates the subjective information of the patient, "O" indicates objective information, "A" indicates "assessment", and "P" indicates a nursing plan. In "assessment", opinions, impressions, etc. are described in addition to judgments and evaluations derived by analyzing subjective information (i.e., "S") and objective information (i.e., "O"). Hereinafter, the text information described as "assessment" in the nursing record is also referred to as "assessment information".

[0013] The assessment support system 100 mainly comprises an assessment evaluation device 1, an input device 2, an output device 3, a storage device 4, and a learning device 5.

[0014] The assessment evaluation device 1 performs an evaluation of an assessment designated as an assessment target. The assessment evaluation device 1 performs data communication with an input device 2, an output device 3, and a storage device 4 via a communication network or by direct wireless or wired communication. The assessment evaluation device 1 identifies and evaluates the assessment of the assessment target based on an input signal “S1” supplied from the input device 2 and information stored in the storage device 4. The assessment evaluation device 1 generates an output signal “S2” related to the evaluation result of the assessment, and supplies the generated output signal S2 to the output device 3.

[0015] FIG. 2(A) is a diagram showing an outline of the processing of the assessment evaluation device 1. In summary, the assessment evaluation device 1 determines an evaluation of an assessment based on an assessment of an assessment target expressed as text and the patient condition of the patient for whom the assessment is created, and outputs an evaluation result showing the determined evaluation. In this case, the assessment is converted into a word vector that quantifies the presence or absence of a predetermined word (also called a "candidate word") that is a keyword in creating the assessment. A candidate word is a word that is a candidate for a word to be included in the assessment of the assessment target, and a plurality of words are determined as the candidate words. Specific examples of the above-mentioned candidate words and patient conditions will be described later.

[0016] Referring again to FIG. 1, the configuration other than that of the assessment evaluation device 1 will be described.

[0017] The input device 2 is an interface that accepts manual input (external input) of information by a user. The input device 2 may be various user input interfaces such as a touch panel, a button, a keyboard, a mouse, or a voice input device. The input device 2 supplies the generated input signal S1 to the assessment evaluation device 1. The output device 3 displays or outputs sound of predetermined information based on the output signal S2 supplied from the assessment evaluation device 1. The output device 3 is, for example, a display, a projector, a speaker, etc.

[0018] The storage device 4 is a memory that stores various information necessary for the processes executed by the assessment evaluation device 1 and the learning device 5. The storage device 4 may be an external storage device such as a hard disk connected to or built into the assessment evaluation device 1 or the learning device 5, or may be a storage medium such as a flash memory. The storage device 4 may also be a server device that performs data communication with the assessment evaluation device 1 and the learning device 5. The storage device 4 may also be composed of multiple devices. At least a part of the information stored in the storage device 4 may be stored by the assessment evaluation device 1 or the learning device 5.

[0019] Functionally, the storage device 4 mainly includes a target patient information storage unit 41, a past patient information storage unit 42, and a regression model storage unit 43.

[0020] The target patient information storage unit 41 stores target patient information, which is information about a patient (also called a "target patient") who is the subject of the assessment. The target patient information includes at least information about the patient condition of the target patient. The information about the patient condition of the target patient includes at least one of attribute information that indicates the basic attributes of the target patient, test information that indicates the past test results of the target patient, and nursing record information that is information about the nursing records of the target patient over a certain period of time in the past. The target patient information storage unit 41 may further store nursing records related to the current medical condition of the target patient.

[0021] Examples of the attribute information include age, sex, disease name, disease stage, nursing diagnosis, medical history, etc. Examples of the test information include height, weight, vital signs, blood test results, etc. Examples of the nursing record information are word vectors generated from the nursing records (particularly assessment information) of the target patient in a certain period of the past. In this case, the word vector may be a word vector extracted by preprocessing the assessment information of the target patient created in a certain period of the past, or may be a word vector obtained by converting the assessment information using a method for vectorizing a sentence. The former word vector is, for example, a vector indicating the presence or absence of a candidate word in the past assessment information of the target patient, and the latter word vector is, for example, a word vector generated based on an arbitrary sentence vectorization method such as BoW (Bag of Words) or Word2vec.

[0022] The past patient information storage unit 42 stores past patient information, which is information about patients (also called "past patients") for whom assessments have been created in the past and recorded in nursing records. It is assumed that the assessments of past patients have been appropriately generated, for example, by experienced medical professionals creating or checking them. The past patient information storage unit 42 stores past patient information about a plurality of past patients. The past patient information includes at least past assessment information representing assessments created in the past, and past patient condition information representing the patient condition of the past patient at the time the assessments were created. The past patient information is used to learn a regression model, which will be described later.

[0023] The information stored in the target patient information storage unit 41 and the past patient information storage unit 42 may be a database in which patient information related to a plurality of patients is stored, or may be various data that can be generated and used in other systems (e.g., electronic medical record systems, etc.). In this case, the patient information is information recorded for each patient (i.e., known information), and may include, for example, medical records, test data, and nursing records created (or created in the past) by a doctor for each patient. Such patient information is linked to, for example, patient identification information (also called "patient ID") for identifying each patient. The information stored in the target patient information storage unit 41 and the past patient information storage unit 42 may be managed as one database or as three or more databases, instead of being divided into two and managed in a distributed manner.

[0024] The regression model storage unit 43 stores information about a regression model that regressively outputs information about words (also called "ideal words") that should be included (i.e., ideally included) in the assessment of a target patient to be evaluated. For example, the regression model is provided for each candidate word, and is a regression model that is trained to output the likelihood that the corresponding candidate word is an ideal word (i.e., the likelihood that the candidate word will appear in an ideal assessment) when information indicating a patient condition is input. In other words, the regression model is a model that has learned the relationship between the patient condition and the necessity of describing the candidate word in the assessment. Such a regression model may be a statistical model such as a logistic regression model, or a machine learning model such as a neural network. Then, the regression model storage unit 43 stores information about parameters required to configure the trained regression model. For example, when the above-mentioned regression model is a model based on a neural network such as a convolutional neural network, the regression model storage unit 43 stores information about various parameters such as a layer structure, a neuron structure of each layer, the number of filters and filter size in each layer, and the weight of each element of each filter.

[0025] The learning device 5 learns a regression model based on the past patient information stored in the past patient information storage unit 42, and stores parameters of the learned regression model in the regression model storage unit 43. The learning by the learning device 5 is performed in advance before the assessment evaluation device 1 executes the evaluation of the assessment.

[0026] FIG. 2(B) is a diagram showing an overview of the process performed by the regression model trained by the learning device 5. In this example, there are N candidate words (N is an integer equal to or greater than 2), and N regression models (regression model 1 to regression model N) for each candidate word (candidate word 1 to candidate word N) are trained. When information indicating the patient condition of the target patient is input to each trained regression model, each regression model outputs the likelihood (probability) that the candidate word is an ideal word. Then, the assessment evaluation device 1 integrates the outputs of the regression models to generate an N-dimensional word vector (also called an "ideal word vector") indicating whether or not the candidate vector is an ideal word for each dimension, or the likelihood. Then, as described later, the assessment evaluation device 1 determines the evaluation of the assessment based on the comparison result between the ideal word vector and a word vector (also called a "described word vector") indicating the presence or absence of the description of the ideal word in the assessment of the evaluation target. In this way, the assessment evaluation device 1 determines the evaluation of the assessment of the evaluation target based on the presence or absence of the description of the ideal word specified based on the patient condition.

[0027] The configuration of the assessment support system 100 shown in FIG. 1 is an example, and various changes may be made to the configuration. For example, the assessment evaluation device 1 and the learning device 5 may be realized by a single device, which performs assessment evaluation and regression model learning. In another example, the input device 2 and the output device 3 may be configured as one. In this case, the input device 2 and the output device 3 may be configured as a tablet terminal that is integrated with or separate from the assessment evaluation device 1. The assessment evaluation device 1 may also be configured of multiple devices. In this case, the multiple devices that configure the assessment evaluation device 1 exchange information required to execute pre-assigned processing between these multiple devices. In this case, the assessment evaluation device 1 functions as a system.

[0028] (2) Hardware Configuration 3(A) shows a hardware configuration of the assessment evaluation device 1. The assessment evaluation device 1 includes, as hardware, a processor 11, a memory 12, and an interface 13. The processor 11, the memory 12, and the interface 13 are connected via a data bus 10.

[0029] The processor 11 executes a program stored in the memory 12 to function as a controller (arithmetic unit) that performs overall control of the assessment evaluation device 1. The processor 11 is, for example, a processor such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or a TPU (Tensor Processing Unit). The processor 11 may be composed of multiple processors. The processor 11 is an example of a computer.

[0030] The memory 12 is composed of various volatile and non-volatile memories such as a RAM (Random Access Memory), a ROM (Read Only Memory), and a flash memory. The memory 12 also stores a program for executing the processes executed by the assessment evaluation device 1. Note that a part of the information stored in the memory 12 may be stored in one or more external storage devices capable of communicating with the assessment evaluation device 1, or may be stored in a storage medium that is detachable from the assessment evaluation device 1.

[0031] The interface 13 is an interface for electrically connecting the assessment evaluation device 1 to other devices. These interfaces may be wireless interfaces such as a network adapter for wirelessly transmitting and receiving data to and from other devices, or may be hardware interfaces for connecting to other devices via cables or the like.

[0032] The hardware configuration of the assessment evaluation device 1 is not limited to the configuration shown in Fig. 3(A). For example, the assessment evaluation device 1 may include at least one of the input device 2 and the output device 3.

[0033] 3(B) shows the hardware configuration of the learning device 5. The learning device 5 includes, as hardware, a processor 51, a memory 52, and an interface 53. The processor 51, the memory 52, and the interface 53 are connected via a data bus 50.

[0034] The processor 51 executes a program stored in the memory 52, thereby functioning as a controller (arithmetic device) that performs overall control of the learning device 5. The processor 51 is, for example, a processor such as a CPU, a GPU, or a TPU. The processor 51 may be composed of multiple processors.

[0035] The memory 52 is composed of various types of volatile and non-volatile memories such as RAM, ROM, and flash memory. The memory 52 also stores programs for executing the processes executed by the learning device 5. Note that some of the information stored in the memory 52 may be stored in one or more external storage devices capable of communicating with the learning device 5, or may be stored in a storage medium that is removable from the learning device 5.

[0036] The interface 53 is an interface for electrically connecting the learning device 5 to other devices. These interfaces may be wireless interfaces such as a network adapter for wirelessly transmitting and receiving data to and from other devices, or may be hardware interfaces for connecting to other devices via cables or the like. An input device, an output device, and the like may also be connected to the interface 53.

[0037] The hardware configuration of the learning device 5 is not limited to the configuration shown in Fig. 3(B) For example, the learning device 5 may include at least one of an input device and an output device.

[0038] (3) Functional Blocks Fig. 4 is an example of a functional block of the assessment evaluation device 1. The processor 11 of the assessment evaluation device 1 functionally includes a patient condition acquisition unit 14, an ideal word vector generation unit 15, an assessment acquisition unit 16, a described word vector generation unit 17, an evaluation unit 18, and an output control unit 19. Note that in Fig. 4, the blocks through which data is exchanged are connected by solid lines, but the combination of blocks through which data is exchanged is not limited to that shown in Fig. 4. The same applies to other functional block diagrams described later.

[0039] The patient condition acquisition unit 14 acquires information on the patient condition that is linked to the patient ID of the target patient and stored in the target patient information storage unit 41 via the interface 13. In this case, the patient condition acquisition unit 14 converts the information on the patient condition acquired from the target patient information storage unit 41 into a vector format so as to be consistent with the input format of the regression model, and supplies the patient condition expressed in the vector format to the ideal word vector generation unit 15. Note that the patient condition is not limited to the vector format and may be expressed by a tensor other than the vector format.

[0040] The ideal word vector generating unit 15 generates an ideal word vector based on the patient condition supplied from the patient condition acquiring unit 14. In this case, the ideal word vector generating unit 15 inputs the patient condition into a regression model configured with reference to the regression model storage unit 43, and acquires the likelihood that each of the N candidate words is an ideal word from the regression model. Then, the ideal word vector generating unit 15 converts the likelihood for the N candidate words output by the regression model to 0 or 1 using a threshold value, for example, and generates an ideal word vector having the converted N values ​​as elements of each dimension. In this case, "0" indicates that the target candidate word is not an ideal word, and "1" indicates that the target candidate word is an ideal word. In this way, the ideal word vector becomes a vector whose elements are values ​​obtained by binarizing the likelihood output by the regression model for each candidate word. Details of the above-mentioned binarization process will be described later.

[0041] Instead of converting each element of the ideal word vector to a value of 0 or 1 by performing binarization processing, the ideal word vector generating unit 15 may use the likelihood of each candidate word output by the regression model as an element of each dimension of the ideal word vector. In this case, the ideal word vector is a vector that represents the likelihood that each of the N candidate vectors is an ideal word. The ideal word vector generating unit 15 supplies the ideal word vector to the evaluation unit 18.

[0042] The assessment acquisition unit 16 acquires text information (i.e., assessment information) representing the assessment to be evaluated. In this case, for example, the assessment acquisition unit 16 acquires an input signal S1 representing the assessment to be evaluated from the input device 2 via the interface 13, and generates assessment information based on the input signal S1. In another example, the assessment acquisition unit 16 acquires assessment information included in the nursing record of the target patient from the target patient information storage unit 41. Then, the assessment acquisition unit 16 supplies the acquired assessment information to the described word vector generation unit 17.

[0043] The described word vector generating unit 17 generates a described word vector of the assessment to be evaluated. Here, the described word vector is expressed in the same space as the coordinate system of the ideal word vector, and is a word vector in which the presence or absence of a candidate word in the assessment to be evaluated is expressed by 0 or 1 for each dimension. In this embodiment, an element of the described word vector that is "0" indicates that the target candidate word does not exist, and an element of the described word vector that is "1" indicates that the target candidate word exists. The described word vector generating unit 17 supplies the described word vector to the evaluation unit 18.

[0044] The evaluation unit 18 determines the evaluation of the assessment to be evaluated based on the ideal word vector supplied from the ideal word vector generation unit 15 and the described word vector supplied from the described word vector generation unit 17. In this case, the evaluation unit 18 calculates the similarity between the ideal word vector and the described word vector, and determines the evaluation of the assessment based on the calculated similarity. Examples of the similarity in this case include the Jaccard coefficient, the Minkowski distance (including Manhattan distance, Euclidean distance, and Chebyshev distance), and the cosine similarity. In addition, the evaluation of the assessment may be a binary value representing the goodness or badness of the assessment, a level representing the goodness or badness of the assessment in three or more stages, a score normalized to a predetermined value range (for example, 0 to 100), or a combination of these. The evaluation unit 18 supplies the determined evaluation result of the assessment to the output control unit 19.

[0045] The output control unit 19 outputs information related to the evaluation of the assessment determined by the evaluation unit 18. In this case, the output control unit 19 generates, for example, an output signal S2 for displaying a display screen showing the evaluation result of the assessment, and supplies the output signal S2 to the output device 3 via the interface 13, thereby causing the output device 3 to display the above-mentioned display screen. Specific display examples will be described later.

[0046] 5 is an example of functional blocks of the learning device 5. The processor 51 of the learning device 5 functionally has a word vector generation unit 54 and a learning unit 55.

[0047] The word vector generation unit 54 acquires past assessment information to be used for learning from the regression model storage unit 43 via the interface 53. Then, the word vector generation unit 54 generates a word vector based on the acquired past assessment information. In this case, the word vector generation unit 54 generates a word vector indicating whether or not each of the N candidate words is included for each past assessment of a patient. In this case, the word vector is represented in the same space (same coordinate system) as the described word vector and the ideal word vector, and is a word vector in which the presence or absence of a candidate word in a previously created assessment is represented by 0 or 1 for each dimension. Then, the word vector generation unit 54 supplies the word vector generated for each past assessment to the learning unit 55.

[0048] The learning unit 55 learns the regression model based on the word vectors supplied from the word vector generation unit 54 and the past patient condition information acquired from the regression model storage unit 43 via the interface 53, and stores the parameters of the regression model obtained by learning in the regression model storage unit 43. In this case, the learning unit 55 expresses the patient condition of each past patient in a vector format, and uses the patient condition of the past patient expressed in a vector format as an input of the regression model. The vector expressing the patient condition of the past patient is a vector expressed in the same coordinate system (same feature space) as the vector expressing the patient condition used by the ideal word vector generation unit 15. In addition, the learning unit 55 recognizes a correct answer (i.e., an ideal output) that the regression model should output based on the word vectors generated by the word vector generation unit 54. Then, the learning unit 55 determines the parameters of the regression model based on a pair of an input to the regression model and a correct answer to be output. In this case, for example, the learning unit 55 determines the parameters of the regression model so that the error (loss) between the output of the regression model according to the input and the correct answer corresponding to the input to the regression model is minimized. The algorithm for determining the above parameters so as to minimize the loss may be any learning algorithm used in machine learning, such as gradient descent or backpropagation.

[0049] Each of the components of the patient condition acquisition unit 14, the ideal word vector generation unit 15, the assessment acquisition unit 16, the written word vector generation unit 17, the evaluation unit 18, and the output control unit 19 described in FIG. 4 can be realized, for example, by the processor 11 executing a program. Also, the word vector generation unit 54 and the learning unit 55 described in FIG. 5 can be realized, for example, by the processor 51 executing a program. Also, each component may be realized by recording a necessary program in an arbitrary non-volatile storage medium and installing it as necessary. At least a part of each of these components may be realized by any combination of hardware, firmware, and software, without being limited to being realized by software by a program. Also, at least a part of each of these components may be realized by using an integrated circuit programmable by a user, such as an FPGA (Field-Programmable Gate Array) or a microcontroller. In this case, the integrated circuit may be used to realize a program composed of each of the above components. Also, at least a part of each of the components may be configured by an ASSP (Application Specific Standard Produce), an ASIC (Application Specific Integrated Circuit), or a quantum processor (quantum computer control chip). In this way, each component may be realized by various hardware. The same applies to other embodiments described below. Furthermore, each component may be realized by the cooperation of multiple computers using, for example, cloud computing technology. The same applies to other embodiments described below.

[0050] (4) Internal processing example Next, a specific example of the internal processing executed by the assessment evaluation device 1 and the learning device 5 in the first embodiment will be described.

[0051] First, when learning a regression model, N candidate words are determined. The candidate words are determined to be, for example, N words (N is an integer equal to or greater than 2) that appear most frequently in the assessments of past patients stored in the past patient information storage unit 42. In this case, words in the assessments of past patients are extracted by morphological analysis, and N candidate words are determined by counting the frequency of the extracted words. Note that the candidate words are not limited to this method, and may be determined to be important keywords in the assessments based on any method (including a personal method). In this specific example, it is assumed that three words (N=3), "breathing," "sleep," and "fatigue," are determined as candidate words.

[0052] Next, the features of the patient's condition to be input to the regression model (more specifically, the features corresponding to each axis of the feature space representing the patient's condition) are determined. Here, as an example, the patient's condition is represented by a two-dimensional coordinate system with an axis corresponding to "pulse" and an axis corresponding to "disease name". Note that "pulse" and "disease name" are quantified based on a predetermined rule. Hereinafter, the variable representing "pulse" is "X" and the variable representing "disease name" is "Y". As a result, the patient's condition is represented as a vector (X,Y) in the above-mentioned two-dimensional coordinate system.

[0053] The learning device 5 then learns a regression model based on the past patient condition information and past assessment information stored in the regression model storage unit 43. In this case, for example, the learning device 5 learns a total of three regression models that output the likelihood that a candidate word is an ideal word. In this case, the regression model for the candidate word "breathing" outputs a posterior probability "P(breathing|X,Y)" according to the patient condition (X,Y) regarding the occurrence of breathing in the target assessment. Similarly, the regression model for the candidate word "sleep" outputs a posterior probability "P(sleep|X,Y)" according to the patient condition (X,Y) regarding the occurrence of sleep in the target assessment, and the regression model for the candidate word "fatigue" outputs a posterior probability "P(fatigue|X,Y)" according to the patient condition (X,Y) regarding the occurrence of fatigue in the target assessment.

[0054] Then, in the evaluation stage, the assessment evaluation device 1 acquires the patient condition of the target patient and generates an ideal word vector by using the learned regression model. For example, if the target patient's "pulse" is "X1" and the "disease name" is "Y1", the following probability vector is calculated. P(P(breathing|X1,Y1),P(sleep|X1,Y1),P(malaise|X1,Y1))

[0055] Hereinafter, it is assumed that P(breathing|X1,Y1) is "0.8", P(sleep|X1,Y1) is "0.1", P(fatigue|X1,Y1) is "0.6", and the probability vector P(0.8,0.1,0.6) is obtained.

[0056] Next, the assessment evaluation device 1 generates an ideal word vector based on the probability vector P. In this case, the assessment evaluation device 1 sets a threshold to be compared with each element of the probability vector P, and sets elements equal to or greater than the threshold to "1" and elements less than the threshold to "0." For example, when the threshold is set to "0.5," an ideal word vector (1,0,1) is obtained from the probability vector P (0.8,0.1,0.6). This ideal word vector indicates that "breathing" and "fatigue" correspond to ideal words (i.e., words that should appear in the assessment), and "sleep" is not an ideal word.

[0057] Here, the above-mentioned threshold value may be set to a common predetermined value stored in advance in the memory 12 or the storage device 4, or may be set to a value according to the candidate word. In the latter case, for example, the assessment evaluation device 1 may determine a threshold value to be compared with a vector element corresponding to each candidate word according to the frequency with which each candidate word appears in the assessment of a past patient stored in the past patient information storage unit 42. In this case, for example, the assessment evaluation device 1 sets the corresponding threshold value to a lower value for a candidate word with a higher frequency, based on a predetermined formula or table information prepared in advance. As a result, a word with a higher frequency of use is more likely to be adopted as an ideal word.

[0058] The assessment evaluation device 1 may regard the probability vector P as a vector obtained by generalizing the elements of the ideal word vector into continuous quantities, and may use the probability vector P as the ideal word vector. In this case, the assessment evaluation device 1 does not execute the binarization process using the above-mentioned threshold value.

[0059] Next, the assessment evaluation device 1 generates a description word vector based on the assessment to be evaluated. Here, as an example, the assessment to be evaluated is "The cause of fatigue is thought to be lack of sleep." In this case, the assessment evaluation device 1 extracts the words ("sleep", "fatigue") described in the assessment to be evaluated by morphological analysis, and generates a described word vector by determining whether the extracted words correspond to each candidate word ("breathing", "sleep", "fatigue"). Here, the words corresponding to the candidate words "sleep" and "fatigue" are extracted, and no word corresponding to the candidate word "breathing" is extracted, so the described word vector is (0,1,1).

[0060] Then, the assessment evaluation device 1 calculates the similarity between the ideal word vector (1,0,1) and the described word vector (0,1,1), and determines an evaluation for the assessment to be evaluated based on the calculated similarity.

[0061] FIG. 6 is a diagram showing an outline of the processing of the assessment evaluation device 1 in a specific example. Here, the left diagram is a coordinate system showing the feature space of the patient condition, and the right diagram is a coordinate system showing the feature space of the ideal word vector and the described word vector. The assessment evaluation device 1 converts (X1, Y1), which corresponds to a combination of pulse and disease name showing the patient condition of the target patient, into an ideal word vector using a regression model. In addition, the assessment evaluation device 1 generates a described word vector by morphological analysis of the assessment of the evaluation target. Then, the assessment evaluation device 1 calculates the similarity between the ideal word vector and the described word vector, and determines the evaluation based on the calculated similarity.

[0062] (5) Display example Next, display examples (first to third display examples) that the output control unit 19 of the assessment evaluation device 1 causes the output device 3 to display will be described.

[0063] Fig. 7 is a first display example regarding the evaluation of the assessment. In the first display example, the output control unit 19 presents the evaluation result of the assessment input by specifying a patient ID to the user. The output control unit 19 generates an output signal S2 based on the input signal S1 generated by the input device 2, etc., and transmits the output signal S2 to the output device 3, thereby causing the output device 3 to display the display screen shown in Fig. 7.

[0064] An input related display area 60 and an evaluation result display area 61 are provided on the display screen according to the first display example. The output control unit 19 causes a patient ID input field 62 for the user to input a patient ID, an assessment input field 63 for the user to input an assessment corresponding to the patient ID specified in the patient ID input field 62, and a confirmation button 64 for instructing the user to evaluate the assessment input in the assessment input field 63 to be displayed on the input related display area 60. When the output control unit 19 detects that the confirmation button 64 has been selected, the output control unit 19 displays information indicating the evaluation result of the assessment input in the assessment input field 63. In the first display example, the output control unit 19 judges whether the assessment of the evaluation target is good or bad (problem or no problem), and displays an evaluation result indicating that the assessment is problem-free in the evaluation result display area 61.

[0065] A supplementary explanation will be given on the calculation of the evaluation result in the first display example. When the output control unit 19 detects that the confirmation button 64 has been selected, the patient condition acquisition unit 14 acquires the patient condition of the target patient linked to the patient ID "123456" specified in the patient ID input field 62 from the target patient information storage unit 41, and the ideal word vector generation unit 15 generates an ideal word vector based on the patient condition acquired by the patient condition acquisition unit 14. In addition, the assessment acquisition unit 16 acquires the text information entered in the assessment input field 63 as assessment information of the evaluation target, and the described word vector generation unit 17 generates a described word vector by morphologically analyzing the assessment information. Then, the evaluation unit 18 calculates the similarity between the ideal word vector and the described word vector, and generates an evaluation result indicating that there is no problem with the assessment since the similarity is equal to or greater than a predetermined threshold, and supplies the evaluation result to the output control unit 19.

[0066] Thus, according to the first display example, the assessment evaluation device 1 can suitably present to the user the evaluation result of the assessment input by specifying the patient ID on the display screen. Furthermore, for example, when the user is in a training stage such as a training period, the assessment evaluation device 1 can present the evaluation result of the assessment entered by the user as feedback. Thus, according to the first display example, it is possible to support the education of the user. Note that, as a modified example of the first display example, the assessment evaluation device 1 may further provide an input field for inputting the patient condition in the input related display area 60, and recognize the patient condition based on the contents of the input in the input field.

[0067] 8 is a second display example regarding the evaluation of the assessment. In the second display example, the output control unit 19 displays a score and its level indicating the quality of the assessment as the evaluation of the assessment, and also presents ideal words that were lacking in the assessment to be evaluated.

[0068] The display screen according to the second display example has an input relation display area 60 and an evaluation result display area 61. The output control unit 19 then displays the above-mentioned patient ID input field 62, assessment input field 63, and confirmation button 64 on the input relation display area 60. After detecting that the confirmation button 64 has been selected, the output control unit 19 also displays an evaluation result display field 65 and an unwritten word display field 66 on the evaluation result display area 61.

[0069] The output control unit 19 displays the score (here, 50) and the corresponding score level (here, warning level 1) as the evaluation result on the evaluation result display field 65. The processing of the evaluation unit 18 in this case will be explained in more detail. In the second display example, the evaluation unit 18 calculates the score by normalizing the similarity between the ideal word vector and the described word vector to a value of 0 to 100 as the evaluation result, and further determines the score level by setting multiple thresholds for the calculated score. Note that the score has a highest evaluation value of 100 and a lowest evaluation value of 0. In addition, in determining the score level, the evaluation unit 18 sets a first threshold (here, 60) for determining whether or not the level is at a level where a warning is required, and a second threshold (here, 40) for dividing the level where a warning is required into two. Then, the evaluation unit 18 determines that the calculated score corresponds to warning level 1 because the calculated score is higher than the second threshold and is equal to or lower than the first threshold, and the output control unit 19 displays the score and the level determination result on the evaluation result display field 65.

[0070] Furthermore, the output control unit 19 displays ideal words (keywords in the figure) that were not included in the assessment to be evaluated in an unlisted word display field 66. Specifically, the output control unit 19 displays words that are not included in the assessment to be evaluated among ideal words (i.e., candidate words corresponding to elements of "1" in the ideal word vector) in the unlisted word display field 66. This allows the output control unit 19 to assist the user in understanding keywords that are estimated to be lacking in the assessment.

[0071] Instead of or in addition to displaying ideal words that were not written, the output control unit 19 may display words (which may be limited to candidate words) included in the assessment to be evaluated that are not ideal words as unnecessarily written words. This allows the output control unit 19 to notify the user that there is an unnecessary entry in the assessment to be evaluated.

[0072] As described above, even in the second display example, the assessment evaluation device 1 can present to the user the evaluation result of the assessment input by designating the patient ID on the display screen. Furthermore, when the user is in a training stage such as a training period, the assessment evaluation device 1 can support the user's education by presenting the detailed evaluation result of the assessment entered by the user and the words that are missing in the assessment as feedback.

[0073] 9 is a third display example related to the evaluation of the assessment. In the third display example, the output control unit 19 collectively performs evaluation on the assessments of multiple target patients stored in the target patient information storage unit 41, and displays the evaluation results.

[0074] On the display screen of the third display example, an input related display area 60A and an evaluation result display area 61A are provided. The output control unit 19 displays a period designation input field 70, a display selection field 71, and a confirmation button 72 on the input related display area 60A. Here, the period designation input field 70 is an input field in which the user designates the period in which the assessment to be evaluated was created. The display selection field 71 is a field in which the assessment whose evaluation result is to be displayed in the evaluation result display area 61A is designated by the score level, and here, as an example, is an input field in the form of a radio button. Note that the score and the score level are calculated by the same calculation method as in the second display example. The confirmation button 72 is a button in which the user instructs the execution of the assessment evaluation.

[0075] Then, when it is detected that the confirmation button 72 has been selected, the assessment evaluation device 1 performs evaluation on the assessments created during the period specified in the period specification input field 70. Then, the output control unit 19 displays, in the evaluation result display area 61A, a list of target patients for which the evaluation result specified in the display selection field 71 is generated from among the generated evaluation results. In the third display example, the output control unit 19 displays, in the evaluation result display area 61A, a set of nursing records and evaluation results of target patients corresponding to evaluation results with a warning (i.e., evaluation results of warning level 1 or warning level 2). The nursing records include the patient ID, patient name, age, sex, disease name, disease stage, nurse in charge, and assessment of the evaluation target, and the evaluation results include a score and a score level.

[0076] Thus, according to the third display example, the assessment evaluation device 1 performs a collective evaluation of the assessments stored in a database or the like, and can present the user with the presence of assessments with warnings. As described above, according to each display example in the first embodiment, it is possible to display the evaluation result as to whether or not the assessment is written correctly, and to support the creation of nursing records. Here, support for the creation of nursing records includes educational support for the user who creates the nursing records.

[0077] (6) Processing flow FIG. 10 is an example of a flowchart executed by the assessment evaluation device 1 in the first embodiment.

[0078] First, the assessment evaluation device 1 acquires the patient condition of the target patient and the assessment of the evaluation target (step S11). For example, the assessment evaluation device 1 acquires these pieces of information from the target patient information storage unit 41 of the storage device 4 or the input device 2. Then, the assessment evaluation device 1 generates a written word vector based on the assessment acquired in step S11 (step S12). In addition, the assessment evaluation device 1 generates an ideal word vector based on the patient condition expressed in a vector format or the like and the trained regression model (step S13). Note that the processing order of steps S12 and S13 may be reversed.

[0079] Next, the assessment evaluation device 1 calculates the similarity between the described word vector calculated in step S12 and the ideal word vector calculated in step S13 (step S14).Then, the assessment evaluation device 1 determines an evaluation for the assessment to be evaluated based on the similarity calculated in step S14, and outputs the determined evaluation result by the output device 3 (step S15).

[0080] (7) Variations The evaluation unit 18 may determine the evaluation of the assessment based on the ideal word vector and the described word vector without calculating the similarity.

[0081] For example, the evaluation unit 18 may determine the above-mentioned assessment evaluation using an evaluation model that outputs an assessment evaluation result (classification result or score of good or bad) when an ideal word vector and a described word vector are input. In this case, the evaluation model is a machine learning model such as a neural network, and is trained in advance to output an assessment evaluation result when an ideal word vector and a described word vector are input, and the trained parameters are stored in the storage device 4 or memory 12, etc. The evaluation unit 18 can determine the assessment evaluation of the evaluation target even when such an evaluation model is used.

[0082] <Second embodiment> 11 shows a schematic configuration of an assessment support system 100A according to the second embodiment. An assessment evaluation device 1A of the assessment support system 100A differs from the assessment evaluation device 1 of the first embodiment in that, instead of generating ideal word vectors based on a regression model, the assessment evaluation device 1A generates ideal word vectors using information indicating the correspondence between the patient state and ideal words to be included in the assessment (i.e., the rules of correspondence). Hereinafter, the same components as those of the first embodiment are appropriately designated by the same reference numerals, and the description thereof will be omitted.

[0083] The storage device 4A functionally has a target patient information storage unit 41 and a correspondence information storage unit 43A. The correspondence information storage unit 43A stores correspondence information indicating the correspondence between the patient condition and the ideal word. For example, the correspondence information is table information (lookup table) indicating the ideal word to be set for each candidate of the expected patient condition. The assessment evaluation device 1A recognizes the ideal word based on the patient condition and the correspondence information, and converts the recognized ideal word into an ideal word vector. Note that the correspondence information may be information in which the ideal word vector is directly associated with each candidate of the patient condition. The correspondence information is generated in advance based on, for example, medical and nursing knowledge.

[0084] Fig. 12 is an example of a functional block of the assessment evaluation device 1A. The assessment evaluation device 1A has the same hardware configuration as the hardware configuration of the assessment evaluation device 1 shown in Fig. 3(A). The processor 11 of the assessment evaluation device 1A functionally has a patient condition acquisition unit 14, an ideal word vector generation unit 15A, an assessment acquisition unit 16, a described word vector generation unit 17, an evaluation unit 18, and an output control unit 19. The patient condition acquisition unit 14, the assessment acquisition unit 16, the described word vector generation unit 17, the evaluation unit 18, and the output control unit 19 execute the same processes as those in the first embodiment.

[0085] The ideal word vector generating unit 15A generates an ideal word vector based on the patient condition information supplied from the patient condition acquiring unit 14. In this case, the ideal word vector generating unit 15A generates an ideal word vector corresponding to the patient condition information expressed in vector format based on the corresponding information stored in the regression model storage unit 43. Then, the ideal word vector generating unit 15A supplies the specified ideal word vector to the evaluation unit 18. After that, the evaluation unit 18 generates an evaluation result based on the similarity between the ideal word vector and the described word vector generated by the described word vector generating unit 17, and the output control unit 19 outputs the evaluation result by the output device 3.

[0086] FIG. 13 is an example of a flowchart executed by the assessment evaluation device 1A in the second embodiment.

[0087] First, the assessment evaluation device 1A acquires the patient condition of the target patient and the assessment of the evaluation target (step S21). Then, the assessment evaluation device 1 generates a described word vector based on the assessment acquired in step S21 (step S22). Also, the assessment evaluation device 1A generates an ideal word vector based on the patient condition and the correspondence information stored in the regression model storage unit 43 (step S23). Note that the processing order of steps S22 and S23 may be reversed. Next, the assessment evaluation device 1A calculates the similarity between the described word vector calculated in step S22 and the ideal word vector calculated in step S23 (step S24). Then, the assessment evaluation device 1A determines the evaluation of the assessment of the evaluation target based on the similarity calculated in step S24, and outputs the determined evaluation result by the output device 3 (step S25).

[0088] In this manner, the assessment evaluation device 1A according to the second embodiment can determine the evaluation of the assessment of the evaluation target, similarly to the assessment evaluation device 1 according to the first embodiment.

[0089] <Third embodiment> 14 shows a schematic configuration of an assessment support system 100B in the third embodiment. The assessment support system 100B in the third embodiment is a server-client model system, and an assessment evaluation device 1B functioning as a server device performs the processing of the assessment evaluation device 1 in the first embodiment or the assessment evaluation device 1A in the second embodiment. Hereinafter, the same components as those in the first or second embodiment are appropriately designated by the same reference numerals, and the description thereof will be omitted.

[0090] The assessment support system 100B mainly includes an assessment evaluation device 1B functioning as a server, a storage device 4B that stores data similar to the storage device 4 according to the first embodiment or the storage device 4A according to the second embodiment, and a terminal device 8 functioning as a client. The assessment evaluation device 1B and the terminal device 8 communicate data with each other via a network 7.

[0091] The terminal device 8 is a terminal having an input function, a display function, and a communication function, and functions as the input device 2 and the output device 3 shown in Fig. 1. The terminal device 8 may be, for example, a personal computer, a tablet terminal, a PDA (Personal Digital Assistant), etc. The terminal device 8 transmits an input signal based on a user input, etc., to the assessment evaluation device 1B.

[0092] The assessment evaluation device 1B has the same hardware configuration and functional block configuration as the assessment evaluation device 1 according to the first embodiment or the assessment evaluation device 1A according to the second embodiment. The assessment evaluation device 1A receives an input signal or the like from the terminal device 8 via the network 7, and evaluates the assessment of the target patient based on the received information. The assessment evaluation device 1B also transmits an output signal indicating information related to the above-mentioned evaluation result to the terminal device 8 via the network 7 based on a request from the terminal device 8. That is, in this case, the terminal device 8 functions as the output device 3 in the first embodiment or the second embodiment. Thereby, the assessment evaluation device 1B suitably presents information related to the assessment evaluation result to the user of the terminal device 8.

[0093] <Fourth embodiment> 15 is a block diagram of an assessment evaluation device 1X in the fourth embodiment. The assessment evaluation device 1X mainly includes a description word vector generating means 17X, an ideal word vector generating means 15X, and an evaluation means 18X. The assessment evaluation device 1X may be composed of a plurality of devices.

[0094] The described word vector generating means 17X generates a described word vector, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in a nursing assessment based on the nursing assessment of a target patient for which a nursing record is to be created. The described word vector generating means 17X can be, for example, the described word vector generating unit 17 in the first to third embodiments.

[0095] The ideal word vector generating means 15X generates an ideal word vector, which is an ideal word vector based on the patient condition of the target patient. The ideal word vector generating means 15X can be, for example, the ideal word vector generating unit 15 in the first embodiment or the third embodiment, or the ideal word vector generating unit 15A in the second embodiment or the third embodiment.

[0096] The evaluation means 18X determines an evaluation of the nursing assessment based on the described word vector and the ideal word vector. The evaluation means 18X can be, for example, the evaluation unit 18 in the first to third embodiments.

[0097] Fig. 16 is an example of a flowchart executed by the assessment evaluation device 1X in the fourth embodiment. The described word vector generating means 17X generates a described word vector, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in a nursing assessment based on the nursing assessment of a target patient for which a nursing record is to be created (step S31). The ideal word vector generating means 15X generates an ideal word vector, which is an ideal word vector based on the patient condition of the target patient (step S32). The evaluation means 18X determines an evaluation of the nursing assessment based on the described word vector and the ideal word vector (step S33).

[0098] According to the fourth embodiment, the assessment evaluation device 1X can determine the evaluation of the nursing assessment of a target patient.

[0099] In each of the above-described embodiments, the program can be stored using various types of non-transitory computer readable media and can be supplied to a computer, such as a processor. The non-transitory computer readable medium includes various types of tangible storage media. Examples of the non-transitory computer readable medium include magnetic storage media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical storage media (e.g., magneto-optical disks), CD-ROM (Read Only Memory), CD-R, CD-R / W, and semiconductor memory (e.g., mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, and RAM (Random Access Memory)). The program may also be supplied to the computer by various types of transitory computer readable media. Examples of the transitory computer readable medium include electric signals, optical signals, and electromagnetic waves. The transitory computer readable medium can supply the program to the computer via a wired communication path such as an electric wire and an optical fiber, or a wireless communication path.

[0100] In addition, some or all of the above embodiments may be described as, but are not limited to, the following supplementary notes.

[0101] [Appendix 1] A description word vector generating means for generating a description word vector, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in a nursing assessment based on a nursing assessment of a target patient for which a nursing record is to be created; an ideal word vector generating means for generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; An evaluation means for determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector; An assessment evaluation device having the above structure. [Appendix 2] The assessment evaluation device according to claim 1, wherein the evaluation means calculates a similarity between the written word vector and the ideal word vector, and determines the evaluation based on the similarity. [Appendix 3] The ideal word vector generating means generates the ideal word vector based on a regression model and the patient condition; 3. The assessment evaluation device according to claim 1 or 2, wherein the regression model is a model that learns the relationship between the patient condition and the necessity of including the candidate word in the nursing assessment. [Appendix 4] The regression model outputs a likelihood of the need for description of the candidate word, The assessment evaluation device according to claim 3, wherein the ideal word vector generation means generates a probability vector having the likelihood as an element as the ideal word vector. [Appendix 5] The regression model outputs a likelihood of the need for description of the candidate word, The assessment evaluation device according to claim 3, wherein the ideal word vector generation means generates, as the ideal word vector, a probability vector whose elements are values ​​binarized based on the likelihood and a threshold value. [Appendix 6] The assessment evaluation device described in Appendix 1 or 2, wherein the ideal word vector generation means generates the ideal word vector corresponding to the patient condition based on correspondence information indicating the correspondence between each of the candidate patient conditions and ideal words that would ideally be included in the nursing assessment. [Appendix 7] The assessment evaluation device according to any one of appendices 1 to 6, wherein the ideal word vector generation means acquires, as the patient condition, at least one of attribute information of the target patient, information regarding the test results of the target patient, or information regarding the past nursing records of the target patient. [Appendix 8] 8. The assessment evaluation device according to any one of claims 1 to 7, further comprising an output control means for causing an output device to output a result of the assessment. [Appendix 9] An assessment evaluation device as described in Appendix 8, wherein the result of the evaluation includes at least one of a judgment result of whether the nursing assessment is good or bad, a score expressing the good or bad of the nursing assessment in a predetermined value range, or a level of the score. [Appendix 10] The assessment evaluation device described in Appendix 8 or 9, wherein the output control means causes the output device to output at least one of ideal words that would ideally be included in the nursing assessment but are not included in the nursing assessment, or words that are included in the nursing assessment but are not ideal words. [Appendix 11] Based on the nursing assessment of a target patient for which a nursing record is to be created, a description word vector is generated, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in the nursing assessment; generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector; Assessment evaluation methods. [Appendix 12] Based on the nursing assessment of a target patient for which a nursing record is to be created, a description word vector is generated, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in the nursing assessment; generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; A storage medium storing a program for causing a computer to execute a process for determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector.

[0102] Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above-mentioned embodiments. Various modifications that can be understood by a person skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention. In other words, the present invention naturally includes various modifications and corrections that a person skilled in the art could make in accordance with the entire disclosure, including the claims, and the technical ideas. In addition, the disclosures of the above cited patent documents and non-patent documents are incorporated herein by reference. [Industrial Applicability]

[0103] For example, it can be used for services related to nursing record support and education on nursing record creation. [Explanation of symbols]

[0104] 1, 1A, 1B, 1X Assessment Evaluation Device 2 Input Devices 3 Output Device 4 Storage device 5 Learning Device 8 Terminal Equipment 100, 100A, 100B Assessment Support System

Claims

1. A description word vector generating means for generating a description word vector, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in a nursing assessment based on a nursing assessment of a target patient for which a nursing record is to be created; an ideal word vector generating means for generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; An evaluation means for determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector; An assessment evaluation device having the above structure.

2. 2. The assessment evaluation device according to claim 1, wherein the evaluation means calculates a similarity between the described word vector and the ideal word vector, and determines the evaluation based on the similarity.

3. The ideal word vector generating means generates the ideal word vector based on a regression model and the patient condition; 3. The assessment evaluation device according to claim 1, wherein the regression model is a model that learns a relationship between the patient condition and the necessity of including the candidate word in the nursing assessment.

4. The regression model outputs a likelihood of the need for the candidate word to be described; 4. The assessment evaluation device according to claim 3, wherein said ideal word vector generating means generates a probability vector having said likelihood as an element as said ideal word vector.

5. The regression model outputs a likelihood of the need for the candidate word to be described; 4. The assessment evaluation device according to claim 3, wherein said ideal word vector generating means generates, as said ideal word vector, a vector having as elements thereof values ​​obtained by binarizing said likelihoods.

6. 3. The assessment evaluation device according to claim 1, wherein the ideal word vector generation means generates the ideal word vector corresponding to the patient condition based on correspondence information indicating the correspondence between the patient condition and ideal words that should ideally be recorded in the nursing assessment.

7. The assessment evaluation device according to any one of claims 1 to 6, wherein the ideal word vector generation means obtains, as the patient condition, at least one of attribute information of the target patient, information regarding the test results of the target patient, or information regarding the target patient's past nursing records.

8. 8. The assessment evaluation device according to claim 1, further comprising an output control means for causing an output device to output a result of the assessment.

9. 9. The assessment evaluation device according to claim 8, wherein the result of the evaluation includes at least one of a judgment result of whether the nursing assessment is good or bad, a score expressing the quality of the nursing assessment in a predetermined value range, or a level of the score.

10. The assessment evaluation device described in claim 8 or 9, wherein the output control means causes the output device to output at least one of ideal words that would ideally be included in the nursing assessment but are not included in the nursing assessment, or words that are included in the nursing assessment but are not ideal words.

11. The computer Based on the nursing assessment of a target patient for which a nursing record is to be created, a description word vector is generated, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in the nursing assessment; generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector; Assessment evaluation methods.

12. Based on the nursing assessment of a target patient for which a nursing record is to be created, a description word vector is generated, which is a word vector obtained by vectorizing the presence or absence of a description of a predetermined candidate word in the nursing assessment; generating an ideal word vector, which is an ideal word vector based on a patient condition of the subject patient; A program for causing a computer to execute a process for determining an evaluation of the nursing assessment based on the described word vector and the ideal word vector.

Images