KR102288797B1 - Cognitive test system - Google Patents

Abstract

The present invention relates to a cognitive test system which can manage a user cognition and cognitive impairment such as dementia through a non-face-to-face method. The system comprises: a fairy tale database in which at least one or more fairy tale information is stored; a question database in which question and answer information related to the fairy tale information stored in the fairy tale database is stored; a server which transmits information related to any one selected from among the fairy tales stored in the fairy tale database and the selected one among the questions stored in the question database; a user terminal which receives, from the server, and outputs the fairy tale information and the question information related to the fairy tale information, and receives an answer to the question information output from a user; and an evaluation unit for evaluating the user answer input through the user terminal by referring to the question and answer information stored in the question database.

Description

Cognitive test system

The present invention relates to a cognitive ability test system for checking the cognitive ability change of people with cognitive decline such as mild cognitive impairment.

Recently, health has been improved due to the improvement of living standards in modern society and the development of medical technology. is increasing rapidly.

In order to prevent such geriatric diseases, it is necessary to periodically care for the psychology and emotions of elderly patients and to evaluate the cognitive abilities of elderly patients. In general, the evaluation of the psychological and emotional care and cognitive ability of these elderly elderly patients is usually carried out face-to-face, but in the recent pandemic situation, face-to-face evaluation is difficult. There is a need for a technology that can take care of emotions and evaluate the cognitive abilities of elderly patients at the same time.

Korean Patent Application Laid-Open No. 10-2019-0093953 (“Service platform to support cognitive ability improvement and emotional stability of elderly and disabled people with cognitive impairment,” published on August 12, 2019)

The present invention has been devised to solve the above problems, and the purpose of the cognitive test system of the present invention is to target people of all ages such as children as well as the elderly through a non-face-to-face method for cognitive impairment and cognitive impairment such as dementia In addition to providing a cognitive test system that can manage

A cognitive test system according to various embodiments of the present invention for solving the above-described problems includes: a moving image database in which at least one moving image information is stored; a question database storing question and answer information related to the moving picture information stored in the moving picture database; a server for transmitting information related to any one selected from among the moving pictures stored in the moving picture database and information related to the selected moving picture from among the questions stored in the question database; a user terminal for receiving and outputting the moving picture information and question information related to the moving picture information from the server, and receiving an answer to the question information output from the user; and an evaluation unit that evaluates the user's answer input through the user terminal with reference to the question and answer information stored in the question database.

In addition, the user terminal receives an answer by voice from the user, or the user selects a screen of the user terminal or inputs a predetermined text to receive an answer, and the cognitive test system provides information on the language used by the subjects. a language database in which it is stored; and a voice recognition unit that, when receiving an answer by voice from the user, recognizes the meaning of the voice spoken by the user by referring to the language database, and converts it into text.

In addition, the cognitive test system further includes a personal information database for storing personal information for each user including age and region information, wherein the language database stores information on the language used by the subject by age and region. And, the voice recognition unit is characterized in that it is characterized in that the user's age and region information of the inputted user is taken into account to determine the meaning of the voice spoken by the user.

In addition, the cognitive test system further includes a synonym database in which similar input value information for a specific word is stored, wherein the voice recognition unit refers to information stored in the synonym database, and the input value recognized by the user as a voice When the similar input value is the similar input value, it is characterized in that it is corrected with a word corresponding to the similar input value to determine the meaning of the voice spoken by the user.

In addition, the user terminal outputs a message to hear the user's story, and when the user inputs a predetermined story, transmits the input information to the server, and the server grasps the context of the input information to determine the predetermined story is corrected and stored in the moving picture database.

According to the cognitive ability test system according to various embodiments of the present invention as described above, it is possible to test the subject's cognitive ability non-face-to-face through the output of assimilation and question and the corresponding answer input. It is effective in managing cognitive impairment such as there is also

In addition, in the cognitive test system according to the present invention, it is possible to construct an infinite number of fairy tales based on personal experiences by receiving a story told by a user (test subject) and making it into a fairy tale so that it can be loaded into the fairy tale database. It also has the effect of providing benefits.

1 is a block diagram of a cognitive test system according to an embodiment of the present invention.
2 is an exemplary diagram of a question output while a moving picture is output from the user terminal of the cognitive test system according to an embodiment of the present invention.
3 is a schematic diagram illustrating a process through which a voice recognition unit of the present invention converts a voice input by a user into text;

Hereinafter, a cognitive test system according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of a cognitive test system according to an embodiment of the present invention.

As shown in FIG. 1 , the cognitive test system according to an embodiment of the present invention includes a user terminal 100 , a server 200 , a moving image database 210 , a question database 220 , and an evaluation unit 230 . may include

It may include a moving picture database 210 , a question database 220 , a server 200 , a user terminal 100 , and an evaluation unit 230 .

The moving picture database 210 stores at least one piece of moving picture information. The moving picture database 210 may be configured to be able to read stored moving picture information by accessing the server 200 . The moving picture information stored in the moving picture database 210 may be composed of text and voice or may be composed of text, voice, and illustration information.

The question database 220 stores question and answer information related to the moving picture information stored in the moving picture database 210 . The question information stored in the question database 220 is provided to the user as questions for evaluating the user's cognitive ability.

The server 200 transmits question information related to any one selected among the moving pictures stored in the moving picture database 210 and any one selected from the moving picture database 210 among the questions stored in the question database 220 to the user terminal 100 . send to

The user terminal 100 receives the moving picture information from the server 200 and outputs question information related to the received moving picture information. Since the fairy tale information is a kind of story, it may be sequentially output, and the question information may be output while the fairy tale is output, while the output of the fairy tale is paused, or after the story is completed. In the embodiment described in this specification, the output is made while the story is in progress, so that the user can focus more on the cognitive test using the present invention.

The user terminal 100 is a device used by a user who is the target of the cognitive test system of the present invention. The user terminal 100 may be a kind of electronic device, and may include communication means for communication with the server 200 to be described later. As some examples, the user terminal 100 may be an electronic device such as a personal computer, a tablet PC, or a smart phone, and the communication means may be a wireless Internet or a wired Internet. In addition, since the user terminal 100 has to output text, voice and illustration, it may include a display means and an audio output means.

When a moving picture is output on the user terminal 100 and a question is output while the moving picture is being output, the user manipulates the user terminal 100 to input text to input an answer, or to output on the user terminal 100 . You can enter an answer by selecting one of the views, or you can enter an answer to a question by voice. Answer information input through the user terminal 100 is transmitted to the outside.

The user terminal 100 may output the moving picture and the answer through a separate application installed on the user terminal 100, but in the present embodiment, the moving picture and the answer are output in a web browser. This method has an advantage that the user can use the cognitive test system according to the present invention without installing a separate application. As described above, in the method of outputting a story and a question on the user terminal 100 and receiving an answer from the user, a natural language is generated to an elderly subject who is a subject using interactive content, and the subject's recognition is analyzed by analyzing the generated natural language. to judge ability.

The evaluation unit 230 determines whether the user's answer information input through the user terminal 100 is correct or not, and finally determines the user's cognitive ability. The evaluation unit 230 may be included in the server, and may be implemented as a kind of program. The evaluator 230 may determine whether the user's answer input through the user terminal 100 is correct by referring to the question/answer information stored in the question database 220 .

The table below is an example of a recording sheet for evaluating a user's dialogue act performance ability used in the evaluation unit 230 .

category number answer Primary Secondary Other demand A-1 ④ Please buy an eraser. A-2 ② I will buy you shoes. A-3 ③ Ask the three sons to stand guard. A-4 ① Guys, be quiet. Sum question B-1 ④ Shouldn't hyung wear nice clothes? B-2 ①Who are you? Who ate eggs? B-3 ③ If not you, what the hell happened? B-4 ①Did you hear the thief? Sum aver C-1 ①Then let's collect eggs and sell them. C-2 ②It's not me! I mean, I didn't eat it. C-3 ②I'm sorry, but I guess I'll just have to wear the pants I used to wear. C-4 ③Your mother's rubber shoes are very old. so it was Sum ritualization D-1 ④ Come to Dad. D-2 ② I'm sorry. D-3 ①Congratulations. D-4 ④Thank you. Sum total

As shown in the table above, the evaluation unit 230 determines the user's ability to perform a dialogue act by using the correct answer to the answer input by the user, and the type of question can also be divided into items such as request / question / assertion / ritualization. . The evaluation history of the evaluation unit 230 may be stored in a personal information database 240 to be described later, and when the evaluation unit 230 evaluates the same subject later, whether items such as cognition are improved by referring to the evaluation history Or you can comment on whether you have regressed.

2 illustrates an example of a question output while a moving picture is output from the user terminal 100 of the cognitive test system according to an embodiment of the present invention.

As shown in FIG. 2 , in the present embodiment, four examples are output as examples for a question. The user may select one of the four views through voice or may select one of the four views by operating the user terminal 100 . In Fig. 2, the problem can be output by voice, and in order to distinguish the problem from the view on the screen, the four views (options) are highlighted, so that a relatively elderly user can more easily select a view. .

Since the illustration shown in FIG. 2 is a part of the fairy tale, the illustration is switched while the fairy tale is continuously output. In the present invention, whenever a screen is switched on the user terminal 100, a cushioner is added to prevent a decrease in concentration due to screen switching.

In the above-described method, moving information is output on the user terminal 100, and the subject inputs an answer to this, a rather passive method was performed.

The cognitive ability test system according to the present invention is not limited thereto, and by outputting a message asking to hear the user's story on the user terminal 100, the user can be induced to tell his or her own story orally.

Although the user can input his/her own story on the user terminal 100 through text input even if it is not an oral form, the interactive conversation method is more useful in cognitive testing or caring for the user's emotions/psychology, so in this embodiment, the user Explains an example in which a person speaks his/her own story through voice.

The server 200 may analyze the text of the user's story input to the user terminal 100 to generate a predetermined recall fairy tale (or story), and store the generated fairy tale (or story) in the fairy tale database 210 . . As such, the recall fairy tale based on the subject's experience or story may be provided to other people for a fee if necessary, or a predetermined advertisement may be output together, and the generated revenue may be shared with the original author (subject) of the corresponding fairy tale.

As such, the present invention induces users to tell their own stories through their experiences or imaginations, and in particular, by telling various episodes of childhood while recalling the past, it can give emotional stability and induce memory to be restored. to prevent cognitive decline. In addition, it is possible to further receive the user's story, analyze the text, generate a predetermined recall, and store it in the children's database, so that many people can receive a cognitive test through the newly created recall.

That is, according to the present invention, since each user's personal stories can be assimilated, it is possible to build a children's story database composed of inexhaustible contents, and economic benefits based on copyright can be given to users.

In the present invention, in order to recognize the user's voice response, a type of voice recognition means may be required, and in consideration of the subject's characteristics, a database capable of assisting the voice recognition means is required.

1 , the cognitive test system according to an embodiment of the present invention may further include a personal information database 240 , a language database 250 , a voice recognition unit 260 , and a synonym database 270 . can

The personal information database 240 stores personal information for each user. Here, the user's personal information may include age and local information in addition to basic personal information. Personal information for each user stored in the personal information database 240 may be input by the user after the user registers in the cognitive test system according to the present invention.

Among the above-mentioned personal information, local information may refer to the place of origin, the place of origin and the area in which you currently reside, or both the place of origin and the area you currently live in, and the area where you have lived in the past. This is because, in the present invention, users, that is, subjects who are the subjects of the cognitive test system according to the present invention, are mostly elderly people, and thus can speak local dialects. In general, the dialect is most affected by the area where you spent your childhood, but when you relocate your residence, it is also affected by the area you moved to. Therefore, in the present invention, when a user registers in the system or after registration, information on the place of birth / place of birth, current residence / place of origin, current residence and past residence can be selectively requested as local information.

The language database 250 stores information on languages used by subjects. The language database 250 may be built in advance, and as subjects perform a cognitive test using the cognitive test system according to the present invention, it is continuously updated to continuously expand the information stored in the language database 250 . can

When the user responds with voice, the voice recognition unit 260 determines what the user's voice means by referring to the language database 250 , and converts it into text. The voice recognition unit 260 may be implemented on the user terminal 100, but in this case, since the amount of calculation of the user terminal 100 itself increases and overload may be applied to the user terminal 100, in the present embodiment, the voice recognition unit 260 may be implemented. The unit 260 may be included in the server 200 .

The voice recognition unit 260 may determine the meaning of the voice spoken by the user in consideration of the age and region information of the user who has input the voice.

The synonym database 270 stores similar input value information for a specific word. This is to correct a user's pronunciation when he/she knows that the pronunciation is incorrect or an incorrect word, so that a more accurate meaning can be grasped. For example, as shown in the table below, the word work may be pronounced as work, work, workk, workk, etc., and the meaning may be clear only when all other words except for work are corrected as work. Substantial correction is performed by the voice recognition unit 260, and more specifically, the voice recognition unit 260 refers to information stored in the synonym database 270. By correcting the word corresponding to the input value, it is possible to clearly understand the meaning of the voice spoken by the user.

synonym input Correction work
work walk Y work Y walk N walk Y

FIG. 3 schematically shows how the voice recognition unit 260 of the present invention converts the voice input by the user into text through what process.

For the realization of the voice recognition unit 260, basically, each phoneme is actually 'what kind of sound.' so that the electronic device can learn the sound. A language model using the N-gram probabilistic language model methodology that enables speech recognition is used. A voice recognition system based on the DNN/HMM model, which is an advanced voice recognition technology, is applied to the previously described acoustic model. Since the N-gram algorithm and the DNN/HMM model are well-known techniques, their description is omitted.

When a user inputs a voice, the voice input unit primarily processes noise and analyzes the characteristics of the voice. Through this, speech feature coefficients that can express phonetic features well for each short-time are extracted. The extracted speech feature coefficients are passed to the pattern recognition process and compared with a pre-stored model of words or phonemes. When the pattern recognition process is completed, it is transmitted to the language processing process in the form of candidate words or candidate phonemes, and in the language processing process, based on the information on the candidate words or candidate phonemes, the target vocabulary and grammatical structure to be recognized, and whether it conforms to a specific topic to output the final recognized sentence.

As shown in FIG. 1 , the cognitive test system according to an embodiment of the present invention may further include a dialogue unit 280 . The conversation unit 280 is for further inducing the user's story by responding to it in the middle when the user talks about his or her own story orally.

The dialogue unit 280 is a process of generating the semantic structure of the system response using an appropriate dialogue model based on the semantic structure of the user's utterance, and a purpose-based dialogue model is often used. In order to accomplish the purpose, it is the key to which dialog model to use, and the dialog model is divided into a rule-based dialog model and a statistics-based dialog model according to the reasoning method used. Reference numeral 280 denotes a statistical-based model, which may be an example-based conversational model.

The Example Based Dialog Model is practical and easy to develop, and it uses language analysis from a large number of dialog examples to construct a dialog example index (INDEX) and finds the most similar example to the user's utterance. Use this in your response. This is a POMDP (Partially Observable Markov Decision Process)-based model that uses reinforcement learning such as Google's AlphaGo, and the POM-based dialogue model uses a dialogue model that is reinforcement-learned through Markov Decision Process (MDP). How to generate an optimal response.

In the present invention, in order to assist the operation of the chat unit 280, a dialogue database 290 including a dialogue example may be further included. Updated, more and more dialogue examples can be accumulated. The response of the chat unit 280 may be output as a voice through the user terminal 100 .

The cognitive test system according to the present invention should be performed periodically. If a specific user's cognitive ability test is not performed for a predetermined period (for example, 3 or 6 months), the server 200 transmits an alarm to the user terminal 100 of the user to induce the user's cognitive ability test. can

According to the cognitive test system according to the present invention as described above, it is possible to support a high-quality life by more easily managing the physical, emotional, and cognitive abilities of a subject, particularly an elderly subject, in a non-face-to-face manner, and cognitive such as dementia It can detect and manage disability early, can induce the overall vitality of the elderly population, and can protect the elderly who are at high risk for infectious diseases in the current pandemic situation.

The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the gist of the present invention as claimed in the claims.

The present invention is a technology that can be usefully used as a diagnostic or educational tool for a cognitive test for the elderly or children.

100: user terminal
200 : server
210: fairy tale database
220: Question database
230: evaluation unit
240: personal information database
250: language database
260: voice recognition unit
270: synonym database
280 : Dialogue
290: Conversation Database

Claims (5)

a moving picture database in which at least one piece of moving picture information is stored;
a question database storing question and answer information related to the moving picture information stored in the moving picture database;
a server for transmitting information related to any one selected from among the moving pictures stored in the moving picture database and information related to the selected moving picture from among the questions stored in the question database;
a user terminal for receiving and outputting the moving picture information and question information related to the moving picture information from the server, and receiving an answer to the question information output from the user; and
With reference to the question and answer information stored in the question database, the evaluation unit for evaluating the user's answer input through the user terminal; a cognitive test system comprising a,
The user terminal receives an answer by voice from the user,
The cognitive test system,
a language database in which information about the language spoken by the subjects is stored; and
and a voice recognition unit that, when receiving an answer by voice from the user, recognizes the meaning of the voice spoken by the user by referring to the language database and converts it into text; and
Further comprising; a personal information database in which personal information for each user including age and local information is stored;
The language database stores information about the language used by the subjects by age and region,
The voice recognition unit recognizes the meaning of the voice spoken by the user in consideration of the age and region information of the user who has input the voice,
It further includes; a synonym database in which similar input value information for a specific word is stored;
The voice recognition unit refers to the information stored in the synonym database, and when the input value recognized as a voice by the user is the similar input value, the voice recognition unit corrects it with a word corresponding to the similar input value to determine the meaning of the voice spoken by the user to figure out,
The user terminal outputs a message to hear the user's story, and when the user inputs a predetermined story, transmits the input information to the server,
The server grasps the context of the input information, corrects it into a predetermined story, stores it in the moving picture database, and provides the user with the revenue generated by the story generated through the user's experience. cognitive test system.

delete delete delete delete

Images