Apparatus and method for recommending customized training content for patients

Abstract

The present disclosure relates to a patient-customized training content recommendation device and method for identifying a patient's cognitive inadequacies, providing training content suited to the patient, calculating a cognitive score, predicting a score for a cognitive ability assessment that is different from the training method used to obtain the cognitive score, and providing the predicted score.The research and development project title for this patent is A1 Development of a customized stroke patient rehabilitation solution and development of a telehealthcare service (Research and development project number: P0025911), which was supported by the Korea Institute for Industrial Technology Advancement, Ministry of Trade, Industry and Energy, Republic of Korea.

Description

[Technical Field]

[0001] The present disclosure relates to a device and method for recommending training content, and more particularly, to a device and method for recommending personalized training content based on a label code. [Background technology]

[0002] As we enter an aging society, cognitive impairment has emerged as a major social problem. Currently, there are no specifically developed drugs for this condition, making it difficult to treat.

[0003] Therefore, it is important to diagnose and prevent cognitive impairment early and slow its progression, and there is an increasing consumer demand for the provision of training content to treat patients. Summary of the Invention [Problem to be solved by the invention]

[0004] The present disclosure has been made in consideration of the above circumstances, and its purpose is to identify a patient's deficient cognitive abilities and provide training content that is suited to the patient.

[0005] Another object of the present disclosure is to provide a predicted score by predicting a score for a cognitive ability assessment method different from the training method used to calculate a cognitive score and obtain the corresponding cognitive score.

[0006] The problems that the present disclosure aims to solve are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description that follows. [Means for solving the problem]

[0007] The patient-tailored training content recommendation device according to the present disclosure for solving the above-mentioned problems may include a profile receiving unit that receives a patient profile including basic information and cognitive information of the patient, a cognitive ability evaluation unit that evaluates the patient's cognitive ability based on the patient profile and calculates the patient's functional cognitive score, a training recommendation unit that recommends at least one training content using an artificial intelligence rehabilitation recommendation model, and a training execution unit that calculates results for the training content provided to the patient via a user terminal.

[0008] In addition, the patient-customized training content recommendation method according to the present disclosure for solving the above-mentioned problems may include the steps of receiving a patient profile including basic information and cognitive information of the patient, evaluating the patient's cognitive ability based on the patient profile and calculating the patient's functional cognitive score, recommending at least one training content using an AI rehabilitation recommendation model, and calculating a result of the training content provided to the patient via the user terminal.

[0009] In addition, a computer program stored on a computer-readable recording medium for executing the method for realizing the present disclosure can also be provided.

[0010] In addition, a computer-readable recording medium having a computer program for executing the method for realizing the present disclosure recorded thereon can also be provided. [Effects of the Invention]

[0011] According to the means for solving the above-mentioned problem of the present disclosure, it is possible to grasp the cognitive ability deficiencies of a patient and provide training content that is suited to the patient.

[0012] According to the means for solving the above-mentioned problem of the present disclosure, it is possible to predict the score for a cognitive ability assessment method that is different from the training method used to calculate the cognitive score and obtain the corresponding cognitive score, and provide the predicted score.

[0013] The effects of the present disclosure are not limited to those mentioned above, and other effects not mentioned above will be clearly understood by those skilled in the art from the description below. [Brief explanation of the drawings]

[0014] [Figure 1] FIG. 1 illustrates a patient-customized training content recommendation system according to the present disclosure. [Figure 2] FIG. 1 is a diagram illustrating the physical configuration of a patient-customized training content recommendation device according to the present disclosure. [Figure 3] FIG. 1 is a diagram illustrating the functional configuration of a patient-customized training content recommendation device according to the present disclosure. [Figure 4] FIG. 2 is a diagram illustrating the order in which the patient-customized training content recommendation method according to the present disclosure is performed. [Figure 5] 1 is a diagram illustrating a main screen of a patient-customized training content recommendation device according to the present disclosure; [Figure 6] FIG. 10 is a diagram illustrating a screen for providing a patient list by the patient-customized training content recommendation device according to the present disclosure. [Figure 7a] FIG. 1 is a diagram illustrating training content provided by a patient-customized training content recommendation device according to the present disclosure. [Figure 7b] FIG. 1 is a diagram illustrating training content provided by a patient-customized training content recommendation device according to the present disclosure. [Figure 8a] FIG. 1 is a diagram illustrating training content recommended by a patient-customized training content recommendation device according to the present disclosure. [Figure 8b] FIG. 1 is a diagram illustrating training content recommended by a patient-customized training content recommendation device according to the present disclosure. [Figure 9] 10A and 10B are diagrams illustrating the results of training content provided by the patient-customized training content recommendation device according to the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0015] The same reference numerals refer to the same components throughout this disclosure. This disclosure does not describe all elements of each embodiment, and general content in the technical field to which this disclosure pertains or overlapping content in the embodiments will be omitted. The terms "unit, module, component, block" used in this specification can be realized by software or hardware, and depending on the embodiment, multiple "units, modules, components, blocks" may be realized as a single component, or one "unit, module, component, block" may include multiple components.

[0016] Throughout this specification, when a part is said to be "connected" to another part, this includes not only direct connection but also indirect connection, including connection via a wireless communication network.

[0017] Furthermore, when a part is described as "comprising" a certain element, this does not mean that it excludes other elements, but that it may further include other elements, unless otherwise specified.

[0018] Throughout this specification, when an element is said to be "on" another element, this includes not only when the element is in contact with the other element, but also when there is another element between the two elements.

[0019] The terms "first," "second," etc. are used to distinguish one component from another, and the components are not limited to the terms described above.

[0020] The singular expression includes the plural expression unless the context clearly indicates otherwise.

[0021] In each step, the identifying numbers are used for convenience of explanation, and the identifying numbers do not describe the order of each step, and each step may be performed in a different order than specified unless the context clearly dictates a specific order.

[0022] The working principle and embodiments of the present disclosure will be described below with reference to the accompanying drawings.

[0023] In this specification, the term "device according to the present disclosure" includes all of the various devices that can perform computations and provide results to a user. For example, the device according to the present disclosure may include all of a computer, a server device, and a portable terminal, or may take any one of these forms.

[0024] Here, the computer may include, for example, a notebook computer, a desktop computer, a laptop computer, a tablet PC, a slate PC, etc., equipped with a web browser.

[0025] The server device is a server that communicates with external devices and processes information, and may include an application server, a computing server, a database server, a file server, a game server, a mail server, a proxy server, and a web server.

[0026] The portable terminal is, for example, a wireless communication device that ensures portability and mobility, and may include any kind of handheld-based wireless communication device such as a Personal Communication System (PCS), Global System for Mobile communications (GSM), Personal Digital Cellular (PDC), Personal Handyphone System (PHS), Personal Digital Assistant (PDA), International Mobile Telecommunication (IMT)-2000, Code Division Multiple Access (CDMA)-2000, W-Code Division Multiple Access (W-CDMA), Wireless Broadband Internet (WiBro) terminal, or a smartphone, as well as wearable devices such as a watch, a ring, a bracelet, an anklet, a necklace, glasses, contact lenses, or a head-mounted device (HMD).

[0027] FIG. 1 is a diagram illustrating a patient-customized training content recommendation system 100 according to the present disclosure.

[0028] Referring to FIG. 1, a system 100 for recommending patient-customized training content may include a user terminal 110, a device 130 for recommending patient-customized training content, and a database 150.

[0029] The user terminal 110 can be realized as a smartphone or a wearable device that can check the patient profile, functional cognitive scores, and training content provided through the patient-customized training content recommendation device 130, collected through the patient-customized training content recommendation device 130, and conversely, can provide the patient profile to the patient-customized training content recommendation device 130. However, the present invention is not limited to this and can be realized as various devices such as a tablet PC. The user terminal 110 can be connected to the patient-customized training content recommendation device 130 via a network, and multiple user terminals 110 can be connected to the patient-customized training content recommendation device 130 simultaneously.

[0030] The patient-customized training content recommendation device 130 can be realized as a computer or a server corresponding to a program that sequentially performs the following operations: receive a patient profile, evaluate the patient's cognitive ability based on the patient profile, calculate the patient's functional cognitive score, recommend training content based on the acquired functional cognitive score, and receive execution information regarding the provided training content. The patient-customized training content recommendation device 130 can be wirelessly connected to the user terminal 110 via Bluetooth (registered trademark), WiFi (registered trademark), a communication network, etc., and can exchange data with the user terminal 110 via the network.

[0031] The database 150 may correspond to a storage device that stores various information generated by the operational process of receiving a patient profile, evaluating the patient's cognitive ability based on the corresponding patient profile to calculate the patient's functional cognitive score, recommending training content based on the obtained functional cognitive score, and receiving execution information regarding the provided training content.

[0032] FIG. 2 is a diagram illustrating the physical configuration of a patient-customized training content recommendation device 130 according to the present disclosure.

[0033] Referring to FIG. 2, the patient-customized training content recommendation device 130 can be realized by including a processor 210, a memory 230, a user input / output unit 250, and a network input / output unit 270.

[0034] The processor 210 can execute procedures for receiving a patient profile, evaluating the patient's cognitive ability based on the patient profile to calculate the patient's functional cognitive scores, recommending training content based on the obtained functional cognitive scores, and receiving performance information regarding the provided training content, and throughout this process, can manage the memory 230 to be read or written and can schedule synchronization times between the volatile memory and non-volatile memory in the memory 230. The processor 210 can control the overall operation of the patient-customized training content recommendation device 130, and is electrically connected to the memory 230, the user input / output unit 250, and the network input / output unit 270 to control the flow of data among them. The processor 210 can be realized as a CPU (Central Processing Unit) of the patient-customized training content recommendation device 130.

[0035] The memory 230 may include a secondary storage device implemented as a non-volatile memory such as an SSD (Solid State Drive) or an HDD (Hard Disk Drive) and used to store all data required for the patient-customized training content recommendation device 130, and may include a primary storage device implemented as a volatile memory such as a RAM (Random Access Memory).

[0036] The user input / output unit 250 may include an environment for receiving user input and an environment for outputting specific information to the user. For example, the user input / output unit 250 may include an input device including an adapter such as a touchpad, touch screen, screen keyboard, or pointing device, and an output device including an adapter such as a monitor or touch screen. In the present disclosure, the user input / output unit 250 corresponds to a computing device connected via a remote connection. In such a case, the patient-customized training content recommendation device 130 may be realized as a server.

[0037] The network input / output unit 270 includes an environment for connecting to external devices or systems via a network, and may include, for example, adapters for communication such as LAN (Local Area Network), MAN (Metropolitan Area Network), WAN (Wide Area Network), and VAN (Value Added Network).

[0038] 3 is a diagram illustrating the functional configuration of a patient-customized training content recommendation device 130 according to the present disclosure. Referring to FIG. 3, the patient-customized training content recommendation device 130 may include a profile receiving unit 310, a cognitive ability assessment unit 320, a training recommendation unit 330, a training execution unit 340, a result providing unit 350, a profile updating unit 360, and a patient cognitive score prediction unit 370.

[0039] The profile receiving unit 310 may receive a patient profile including basic information and cognitive information of the patient. The basic information of the patient may include the patient's age, gender, highest level of education, and current condition. The current condition may include at least one of the patient's cognitive and language severity levels. The cognitive information may include a result of the training content, and if there is no training content that the patient has already completed, may include a separate, predefined cognitive score. Here, the predefined training content may refer to training content completed through the patient-customized training content recommendation device 130. That is, if there is training content that the patient has already completed, the profile receiving unit 310 may receive the patient's cognitive score based on the result of the training content.

[0040] Here, the predefined cognitive score may refer to a separate cognitive score that is different from the result of the training content performed through the patient-customized training content recommendation device 130. For example, the predefined cognitive score may refer to a score related to measuring cognitive ability, such as the Mini-mental state examination (MMSE), the global deterioration scale (GDS), and the clinical dementia rating (CDR).

[0041] The cognitive ability assessment unit 320 can assess the patient's cognitive ability based on the patient profile and calculate a functional cognitive score for the patient. For example, the cognitive ability assessment unit 320 can calculate a cognitive score for each of the patient's cognitive functions, such as attention, memory, executive function, reading, writing, speaking, and comprehension. More specifically, the cognitive ability assessment unit 320 can calculate absolute scores and statistical scores for the patient's cognitive functions, which will be described in more detail below.

[0042] In the present disclosure, the cognitive ability assessment unit 320 may define functional cognitive scores according to the graded subfunctions and calculate statistical values ​​for each subfunction. For example, the graded subfunctions may include mental functions as a major category, which may include perceptual functions, executive functions, language functions, calculation functions, and memory. Perceptual functions may include auditory perception, visual perception, and visual-spatial perception. Executive functions may include abstraction, organization and planning, cognitive flexibility, judgment, and problem-solving. Language functions may include linguistic comprehension and linguistic expression. Calculation functions may include simple calculation and complex calculation. As such, the subfunctions are divided into major categories, intermediate categories, and minor categories, but are not limited thereto. Therefore, the cognitive ability assessment unit 320 may define subfunctions in a similar manner to defining at least one major category and placing intermediate categories under the major category, as described below.

[0043] In addition, the cognitive ability assessment unit 320 may calculate statistical values ​​for each sub-function. For example, the cognitive ability assessment unit 320 may calculate statistical values ​​for each sub-function by determining whether the mental function score is in the top 10% of patients of the same age. Here, the statistical value may refer not only to percentiles but also to statistically processed data such as a statistically processed z-score or a p-value calculated by calculating a difference from a group of the same age.

[0044] In the present disclosure, the cognitive ability assessment unit 320 may define cognitive scores for each function through a first step of defining at least one or more subfunctions, a second step of defining detailed subfunctions included in at least one or more subfunctions, and a third step of labeling the subfunctions and detailed subfunctions to generate label codes for the subfunctions and detailed subfunctions. For example, the cognitive ability assessment unit 320 may define the subfunctions as mental functions, sensory functions and pain, and voice and speaking functions. Furthermore, the cognitive ability assessment unit 320 may define the sensory functions and pain functions to include detailed subfunctions of visual functions, functions of structures around the eyes, and auditory functions.

[0045] Furthermore, the cognitive ability assessment unit 320 can assign a label code of b1 to mental functions, a label code of b2 to sensory functions and pain, and a label code of b3 to voice and speaking functions. The cognitive ability assessment unit 320 can also assign label codes to detailed sub-functions, for example, a label code of b210 to visual functions.

[0046] In the present disclosure, the cognitive ability assessment unit 320 may calculate a cognitive score for each function by summing up statistical values ​​for each sub-function. For example, the cognitive ability assessment unit 320 may calculate statistical values ​​for each of the sub-functions, namely, the voice function, the articulation function, the speaking, fluency, and rhythm function, and the alternative speech function, and sum up the statistical values ​​to calculate a cognitive score for the voice and speaking function.

[0047] The training recommendation unit 330 may recommend at least one training content using an AI rehabilitation recommendation model according to the functional cognitive scores. For example, the training recommendation unit 330 may recommend training content that can train functions that are inferior to the general cognitive functions for the patient's age, among the functional cognitive scores of the patient. For example, the training recommendation unit 330 may recommend training content that can train attention to a patient with reduced attention, and a method for this will be described in detail below.

[0048] The training content may include at least one cognitive training exercise, the type of training, the level of difficulty, the number of questions, and the time required to complete the exercise. For example, the training recommendation unit 330 may set the type of training, the level of difficulty, the number of questions, and the time required to complete the exercise while providing the training content to the patient.

[0049] In the present disclosure, the training recommendation unit 330 may provide training content that can reinforce a sub-function with a low statistical value. For example, if the statistical value of the mental function of the patient is low, the training recommendation unit 330 may provide training content that can complement the mental function.

[0050] In the present disclosure, the training recommendation unit 330 can provide training content that can reinforce detailed sub-functions with low statistical values. For example, the training recommendation unit 330 can provide training content that can reinforce memory to a patient whose statistical value for memory is measured to be low.

[0051] In the present disclosure, the training recommendation unit 330 may limit the criteria for low statistical values ​​to a specific number. For example, the training recommendation unit 330 may provide training content to strengthen the four lowest-scored sub-functions among the sub-functions. As another example, the training recommendation unit 330 may provide training content to strengthen sub-functions that have a z-score of 1.5 or less.

[0052] In the present disclosure, the training recommendation unit 330 may use an AI rehabilitation recommendation model to select cognitive trainings that include the most label codes of detailed sub-functions included in sub-functions with low statistical values ​​and recommend them as at least one training content. For example, if the label code of executive function among the detailed sub-functions is b164 and the codes of word association training include b1640, b1641, b1642, and b1643, the training recommendation unit 330 may include the corresponding training in the training content and provide it to the patient because the word association training includes four training codes that belong to the lower level of b164. That is, the training recommendation unit 330 may check the label codes of the corresponding trainings and determine whether they match the label codes missing from the patient to determine whether to recommend the corresponding training to the patient.

[0053] In the present disclosure, the training recommendation unit 330 can recommend a recommended training content list as training content. Here, the AI ​​rehabilitation recommendation model can generate a recommended training content list including a specific label code by inputting a specific label code. For example, the training recommendation unit 330 can receive an input of a specific label code for training required for a patient visiting for the first time, and can provide training content corresponding to the corresponding label code using the AI ​​rehabilitation recommendation model.

[0054] In the present disclosure, the AI ​​rehabilitation recommendation model can cluster at least one patient based on their performance results on specific training content, and generate a list of recommended training content based on the cluster to which the patient belongs.

[0055] Specifically, the AI ​​rehabilitation recommendation model can collect accuracy and reaction time from the patient's execution results for the training content and calculate the similarity of the accuracy and reaction time with other patients. That is, the AI ​​rehabilitation recommendation model can determine at least one reference value, calculate n-dimensional values ​​including the corresponding reference values ​​for each patient, and then calculate the similarity of each dimension and the similarity between specific patients by comparing the values ​​of each dimension. As an example, the AI ​​rehabilitation recommendation model can calculate the cosine similarity between each value to generate patient clusters.

[0056] In the present disclosure, the AI ​​rehabilitation recommendation model can generate a recommended training content list based on training content performed by other patients in the cluster to which the patient belongs, in addition to the training content performed by the patient. For example, the AI ​​rehabilitation recommendation model can generate a recommended training content list based on training content not performed by the patient but performed by other patients in the cluster to which the patient belongs. As another example, the AI ​​rehabilitation recommendation model can generate a recommended training content list based on training content not performed by the patient but performed by other patients in the cluster to which the patient belongs, based on the training content performed most frequently by other patients.

[0057] In the present disclosure, the AI ​​rehabilitation recommendation model can generate a recommended training content list through a first step of determining whether the patient's accuracy rate for the training content is equal to or greater than a first standard; a second step of determining whether the patient's reaction time for the training content is less than a second standard if the patient's accuracy rate is equal to or greater than the first standard; and a third step of generating a recommended training content list of training content with increased difficulty levels if the patient's reaction time is less than the second standard and the training content is not of the highest difficulty. For example, if the patient's accuracy rate for the training content is 80% or greater, the AI ​​rehabilitation recommendation model can determine whether the patient's reaction time for the training content is less than a specific time. Here, the specific time can be preset and can be the value obtained by subtracting 2 × median absolute deviation from the median reaction speed of patients in the cluster to which the patient belongs. If the patient's reaction time is less than the specific time, the AI ​​rehabilitation recommendation model can increase the difficulty level of the recommended training content by one level if the training content is not of the highest difficulty level, or can terminate the recommendation if the training content is of the highest difficulty level. In addition, the AI ​​rehabilitation recommendation model can recommend training content with a consistent level of difficulty if the patient's reaction time is above a certain time.

[0058] In the present disclosure, the AI ​​rehabilitation recommendation model can generate a recommended training content list through a first step of determining whether the patient's accuracy rate for the training content performed by the patient is below a first standard; a second step of determining whether the patient's accuracy rate for the training content performed by the patient is above a third standard if the patient's accuracy rate is below the first standard; and a third step of generating a recommended training content list of training contents with a maintained difficulty level if the patient's accuracy rate is above the third standard, and generating a recommended training content list of training contents with a lowered difficulty level if the patient's accuracy rate is below the third standard. For example, the AI ​​rehabilitation recommendation model can set the second standard to a value obtained by subtracting 2 x standard deviation from the average accuracy rate of the corresponding content performed by other patients in the cluster to which the corresponding patient belongs. The AI ​​rehabilitation recommendation model can also maintain or lower the difficulty level of the recommended training content depending on the patient's accuracy rate.

[0059] In the present disclosure, the AI ​​rehabilitation recommendation model may include a prediction model that predicts a patient's results for training content and a recommendation model that generates a list of recommended training content based on a recommendation score calculated according to a set target value. For example, the prediction model may predict a patient's results for training content using a learning method such as machine learning. More specifically, the prediction model may use a known AI learning method to predict the patient's accuracy rate for training content based on the patient's results for previous training. The recommendation model may generate a list of recommended training content based on a recommendation score calculated according to a set target value. For example, the recommendation model may calculate recommendation scores for training content according to target values ​​for difficulty, performance experience, reaction time, solving time, and sensitivity. More specifically, the recommendation model may calculate recommendation scores using the following Equations 1 to 11.

[0060] (Equation 1) Recommendation points (p,t) =1+wl*Difficulty (p,t ,s1 ,s2)+w2*Execution experience(p,t ,s2)+w3*Reaction time(p,t ,s3)+w4*Solving time(p,t ,s4)+w5*Sensitivity(p,t ,s5 ,[ts])

[0061] Here, w1 to w5 are preset weights, p is a given patient, t is the type of training, s1 to s5 are set values, and [ts] is information about training that has already been completed.

[0062] (Equation 2) Difficulty level (p,t,s1,s2)=w11*stage(t,s1)+w12*timer(t,s1)+w13*correct answer rate(t,s1)

[0063] (Equation 3) Stage(t,s1)=1-abs(target(s1)-prediction(t))

[0064] Here, the target (s1) is s1 / 100, s1 is the set value of the target execution level, and the forecast (t) is 1-(stage (t) / highest stage (t)). abs() represents the absolute value.

[0065] (Equation 4) timer(t,s1)=1-abs(target(s1)-prediction(t))

[0066] Here, prediction(t) is idx(sort(timer option list(t)), t) / len(timer option list(t)).

[0067] (Equation 5) Correct answer rate (p,t ,s1 ,s2)=w131*(1-abs(target(s1)-prediction(t)))

[0068] Here, s2 is the execution experience preference setting value, w131 is a preset weight value, and prediction(t) is the prediction accuracy rate for the patient's corresponding training content.

[0069] (Equation 6) Execution experience (p,t,s2) = (1-abs(goal (s2)-prediction (p,t)))

[0070] Here, the target (s2) is s2 / 100, s2 is the execution experience preference setting value, the prediction (p, t) is 1 / rank(t), and rank(t) is the position of t in the list of all training contents (T) sorted according to the history of the patient performing the corresponding training content (t).

[0071] For example, if a patient performs a lot of the relevant training content, the rank(t) value may be large.

[0072] (Equation 7) Reaction time (p,t,s3) = w31*(1-abs(target(s3)-prediction(p,t)))

[0073] Here, the target (s3) is s3 / 100, s3 is the preference setting for short reaction times, w31 is a preset weighting, prediction (p, t) is 1 / rank(t), and rank(t) is the position of t in the list sorted in order of the patient's predicted short reaction times among the entire training content (T).

[0074] For example, the rank(t) value may be larger if the patient predicts a shorter reaction time for the training content.

[0075] (Equation 8) Solving time (p,t,s4) = s41 * (1-abs(target(s4)-prediction(p,t)))

[0076] Here, the target (s4) is s4 / 100, s4 is the solving time preference setting, w41 is a preset weighting value, the prediction (p, t) is 1 / rank(t), and rank(t) is the position of t in the list sorted in order of the overall training content (T) predicted to be solved quickly by the patient.

[0077] For example, the faster the patient is expected to solve the training content, the larger the rank(t) value may be.

[0078] (Equation 9) Sensitivity (p,t ,s5 ,[ts]) = w51*ICF+w52*Session performance

[0079] (Equation 10) JPEG2026504007000002.jpg29127

[0080] Here, w511 = s5 / 100, s5 is the sensitivity preference setting value, and similarity ([ts], t) is the number of times the training content (t) recommended for the ICF code matches the training content list ([ts]) performed by the patient.

[0081] (Equation 11) Session performance = w521*(1-abs(new target(s1,[ts])-accuracy(p,t,new target(s1,[ts]),s2)))

[0082] Here, w521 is s5 / 100, and the new target (s1, [ts]) = (s1 - trained accuracy rate ([ts])) + s1.

[0083] For example, the recommendation model can calculate the recommendation score (p, t) of training content using set values ​​of 0 to 100 s1 to s5, and generate a recommended content list including training content with high recommendation scores.

[0084] The training execution unit 340 may calculate results for the training content provided to the patient via the user terminal, including the number of questions answered by the patient, the time taken, and the number of correct answers for each area.

[0085] The result providing unit 350 may provide results for the training content. For example, the result providing unit 350 may provide the results of the training content to the patient and compare the results with previously performed training content.

[0086] The profile update unit 360 may update the patient profile according to the results of the training content. For example, if the patient's cognitive function improves after performing a training content compared to a previous training content, the profile update unit 360 may update the patient profile to reflect the improved cognitive function.

[0087] The patient cognitive score prediction unit 370 can generate a predicted cognitive score by predicting a predefined cognitive score that is different from the patient's functional cognitive score based on the statistical value for each sub-function. For example, the patient cognitive score prediction unit 370 can predict and display the MMSE, GDS, and CDR scores of the patient based on the z-score value for each sub-function. In other words, the patient cognitive score prediction unit 370 can determine the degree of cognitive function of the patient compared to the same age group and match it with other predefined cognitive scores to provide convenience to the patient.

[0088] In the present disclosure, the patient recognition score prediction unit 370 may calculate the accuracy of the predicted recognition score value depending on the size of the population from which the statistical values ​​for each sub-function are calculated. For example, the patient recognition score prediction unit 370 may calculate that the accuracy of matching with the predicted recognition score value increases as the size of the population from which the statistical values ​​for each sub-function are calculated increases. That is, the patient recognition score prediction unit 370 may collect more patient data and provide other predicted recognition scores with higher accuracy as the number of patients using the patient-customized training content recommendation device 130 increases. For example, the patient recognition score prediction unit 370 may calculate the accuracy of the predicted recognition score value if it collects data from a population exceeding a certain number of patients, and may display a simple predicted value and not calculate the accuracy of the predicted value if it does not exceed the certain number of patients.

[0089] 5 to 9 are diagrams illustrating screens on which the patient-customized training content recommendation device 130 according to the present disclosure is implemented.

[0090] 5, the patient-customized training content recommendation device 130 allows the user to check the hospital name, administrator, and registration date through the main screen. Furthermore, the patient-customized training content recommendation device 130 allows the user to count the number of therapists and patients in real time, check the amount spent, check the number of recently registered patients, and check any necessary notifications through the main screen. Furthermore, the patient-customized training content recommendation device 130 allows the user to check inquiries, a list of registered therapists, and patient information through the main screen.

[0091] 6, the patient-customized training content recommendation device 130 allows users to register / modify patients, write simple notes, and check the current registration status of patients through a patient list screen. Also, patients can be sorted through the patient list screen of the patient-customized training content recommendation device 130, and detailed information of a selected patient can be checked.

[0092] Referring to Figures 7a and 7b, the patient-customized training content recommendation device 130 can check the information of the selected patient and the training history of the patient through the general training page, select which areas need training, and set the order, difficulty, etc. for each training.

[0093] Referring to Figures 8a and 8b, the patient-customized training content recommendation device 130 allows the user to check the information of the selected patient through the automatic training page, set the target time and accuracy rate, set the weighting between previously experienced training and new training, select areas that require particular treatment, and select and control whether to make detailed settings for the automatically recommended recommended content.

[0094] Referring to FIG. 9, the patient-customized training content recommendation device 130 allows the user to check the frequency of training for each detailed area through the report page, as well as the total training time, accuracy, and information on recent training. The user can also classify each detailed item to see which cognitive functions are lacking.

[0095] FIG. 4 is a diagram illustrating a flow of executing a patient-customized training content recommendation method according to the present disclosure.

[0096] Referring to FIG. 4, the patient-customized training content recommendation method may receive a patient profile including basic information and cognitive information of the patient through the profile receiving unit 310 (S410).

[0097] The patient-customized training content recommendation method may evaluate the cognitive ability of the patient based on the patient profile via the cognitive ability evaluation unit 320, and calculate the patient's functional cognitive score (S420).

[0098] The patient-customized training content recommendation method can recommend at least one training content through an AI rehabilitation recommendation model according to the functional cognitive scores via the training recommendation unit 330 (S430).

[0099] The patient-customized training content recommendation method may calculate results for the training content provided to the patient through the user terminal by the training execution unit 340 (S440).

[0100] Meanwhile, the disclosed embodiments may be realized in the form of a recording medium storing computer-executable instructions. The instructions may be stored in the form of program code, which, when executed by a processor, generates program modules to perform the operations of the disclosed embodiments. The recording medium may be realized as a computer-readable recording medium.

[0101] Computer-readable recording media include all types of recording media that store computer-readable instructions, such as ROM (Read Only Memory), RAM (Random Access Memory), magnetic tape, magnetic disk, flash memory, and optical data storage devices.

[0102] The disclosed embodiments have been described above with reference to the accompanying drawings. Those skilled in the art will understand that the present disclosure may be embodied in forms different from the disclosed embodiments without changing the technical concept or essential features of the present disclosure. The disclosed embodiments are illustrative and should not be construed as limiting.

Claims

1. a profile receiving unit for receiving a patient profile including basic information and cognitive information of the patient; a cognitive ability assessment unit that assesses the cognitive ability of the patient based on the patient profile and calculates a functional cognitive score of the patient; a training recommendation unit that recommends at least one training content according to an artificial intelligence rehabilitation recommendation model; a training execution unit that calculates a result for the training content provided to the patient via a user terminal; A patient-customized training content recommendation device including:

2. The basic information of the patient is The patient's current condition, age, sex, and highest level of education, The cognitive information is The patient-customized training content recommendation device of claim 1, further comprising a result for the training content, and if there is no training content that the patient has already completed, a separate predefined cognitive score.

3. The cognitive ability assessment unit The patient-customized training content recommendation device according to claim 2, wherein the functional recognition scores are defined according to stepwise divided sub-functions, and statistics for each of the sub-functions are calculated.

4. The cognitive ability assessment unit a first step of defining at least one sub-function; a second step of defining detailed sub-functions included in said at least one sub-function; and a third step of labeling the sub-functions and the detailed sub-functions to generate label codes for the sub-functions and the detailed sub-functions.

5. The cognitive ability assessment unit The patient-customized training content recommendation device according to claim 4, wherein the statistical values ​​are summed for each of the sub-functions to calculate the cognitive scores for each function.

6. The artificial intelligence rehabilitation recommendation model: By inputting a specific label code, a recommended training content list including the specific label code is generated; The training recommendation unit:

4. The patient-customized training content recommendation device according to claim 3, wherein the recommended training content list is recommended as the training content.

7. The artificial intelligence rehabilitation recommendation model: The patient-customized training content recommendation device according to claim 1, characterized in that at least one patient is clustered based on the results of execution of specific training content, and a list of recommended training content is generated based on the cluster to which the patient belongs.

8. The artificial intelligence rehabilitation recommendation model: The patient-customized training content recommendation device of claim 7, characterized in that in addition to the training content performed by the patient, training content performed by other patients in the cluster to which the patient belongs is generated as the recommended training content list.

9. The artificial intelligence rehabilitation recommendation model: a first step of determining whether the patient's accuracy rate in the training content is equal to or greater than a first standard; a second step of determining whether the patient's reaction time to the training content is less than a second standard when the patient's accuracy rate is equal to or greater than a first standard; and a third step of generating, as the recommended training content list, training content with a higher level of difficulty than the training content performed by the patient if the patient's reaction time is less than a second standard and the training content performed by the patient is not of the highest level of difficulty.

10. The artificial intelligence rehabilitation recommendation model: a first step of determining whether the patient's accuracy rate of the training content is less than a first standard; a second step of determining whether the patient's correct answer rate for the training content is equal to or greater than a third standard when the patient's correct answer rate is less than a first standard; The patient-customized training content recommendation device of claim 7, characterized in that the recommended training content list is generated by a third step in which training content with the same level of difficulty is generated as the recommended training content list if the patient's accuracy rate is equal to or higher than a third standard, and training content with a lowered level of difficulty is generated as the recommended training content list if the patient's accuracy rate is less than the third standard.

11. The artificial intelligence rehabilitation recommendation model: a predictive model that predicts the patient's outcome for the training content; a recommendation model that generates a list of recommended training contents based on a recommendation score calculated according to a set target value; The patient-customized training content recommendation device according to claim 1 , further comprising:

12. The training content includes: At least one cognitive training program with a set type of training, difficulty level, number of questions, and time to complete the program is included; The training recommendation unit: The patient-customized training content recommendation device according to claim 5, wherein the device provides training content that can reinforce the sub-functions with low statistical values.

13. The training recommendation unit: The patient-customized training content recommendation device of claim 12, characterized in that the cognitive training that includes the most label codes of the detailed sub-functions included in the sub-functions with the lowest statistical values ​​is selected by the artificial intelligence rehabilitation recommendation model and recommended as the at least one training content.

14. a result providing unit that provides results for the training content; a profile update unit that updates the patient profile according to the results of the training content; and a patient cognitive score prediction unit that predicts the predefined cognitive scores that are different from the patient's functional cognitive scores according to the statistical values ​​of the sub-functions to generate a predicted cognitive score value, The patient cognitive score prediction unit The patient-customized training content recommendation device according to claim 3, wherein the accuracy of the predicted cognitive score is calculated according to the size of a population from which statistics for each of the sub-functions are calculated.

15. A method for recommending customized training content for a patient performed on a user terminal, receiving a patient profile including demographic and cognitive information of the patient; assessing the patient's cognitive ability based on the patient profile and calculating a functional cognitive score for the patient; recommending at least one training content according to an artificial intelligence rehabilitation recommendation model; calculating a result for the training content provided to the patient via the user terminal; A method for recommending customized training content for patients.

Images