Method and system for diagnosing dementia through story-based question-and-answer evaluation and story-based speech extraction

Abstract

The present invention relates to a dementia diagnosis system using story-based question-and-answer evaluation, which enables dementia diagnosis to be rapidly and conveniently performed without using expert personnel, enables early diagnosis of dementia and screening of mild patients to be rapidly and accurately performed, and enables evaluation to be performed in a question-and-answer manner through voice so that an elderly subject who cannot recognize characters well can easily perform a medical examination.SOLUTION: The dementia diagnosis system includes a data input device for inputting a real story, a data storage device for storing text data input through the data input device, a voice output device for outputting diagnosis voice data to the outside as a voice, a voice receiver for receiving an answer voice of a person to be examined, and a controller which is connected to all devices including the data input device to control the devices, forms an adaptation story, a question, and a correct answer based on the real story, receives the answer of the person to be examined, and compares the correct answer with the answer to determine dementia.SELECTED DRAWING: None

Description

[Technical Field]

[0001] The present invention relates to a dementia diagnosis system and method, and more specifically, to a dementia diagnosis system and method using a story-based question-and-answer assessment and story-based voice extraction that enables dementia diagnosis to be performed quickly and easily without the need for specialized personnel, enables early diagnosis of dementia and screening of mildly ill patients to be performed quickly and accurately, and enables assessment to be performed in a question-and-answer format through voice or audio presentation of a story to allow even elderly subjects who have difficulty recognizing letters to easily undergo screening. [Background technology]

[0002] Dementia generally refers to a group of symptoms resulting from brain disease. As dementia progresses, it affects thinking, behavior, and the ability to carry out daily activities. Dementia is characterized by a decline in cognitive ability that leads to a lack of ability to carry out daily activities.

[0003] Doctors typically diagnose dementia when two or more cognitive functions are significantly impaired. These cognitive functions include memory, language, information comprehension, spatial function, judgment, and attention. Patients with dementia may have difficulty solving problems and regulating emotions and may experience personality changes.

[0004] The exact symptoms a person with dementia experiences depend on which parts of the brain are damaged by the disease that caused the dementia. Dementia occurs when some of the brain's nerve cells stop functioning, lose connections with other cells, and die. This type of dementia is usually progressive, meaning it gradually spreads through the brain and the patient's symptoms worsen over time.

[0005] Dementia is a typical age-related neurodegenerative brain disease, with a prevalence of approximately 5-10% among people aged 65 and over worldwide. Most dementia patients exhibit symptoms of progressive cognitive impairment, hallucinations, delusions, and loss of ability to function.

[0006] In the case of Alzheimer's disease, the most common form of dementia, clumps of waste material made of a protein called amyloid beta are observed in nerve fiber bundles and around brain cells within nerve cells in the cerebral cortex, and it is believed that these waste materials cause necrosis of nerve cells.

[0007] Alzheimer's disease is the most common disease, accounting for approximately 70% of all dementia cases, and it has recently been estimated that approximately 600,000 people in Japan are diagnosed with Alzheimer's disease each year.

[0008] Alzheimer's disease is a typical degenerative dementia disease that causes a decline in cognitive function and daily life due to the degeneration of cerebral nerve cells, and is a dreadful disease that can lead to death, causing great suffering not only to patients but also to their families who care for them.

[0009] Diagnosis of dementia involves taking a detailed medical history and evaluating symptoms to determine whether cognitive impairment is causing problems in daily life and social activities, and then using brain imaging to check for cerebrovascular disease and cerebral atrophy to confirm dementia.

[0010] In the early stages of dementia, it is difficult to distinguish it from senile amnesia, so dementia is diagnosed through neuropsychological testing that comprehensively evaluates not only memory but also language ability, calculation ability, spatiotemporal perception ability, judgment, etc. To diagnose this type of dementia, additional detailed testing is required based on several pieces of information obtained from interviews with the patient / guardian and screening tests. Additional detailed testing includes neuropsychological testing (SNSB), blood tests, and various types of brain imaging (CT, MRI, PET).

[0011] As an example of additional workup, magnetic resonance imaging (MRI) can play an important role in differentiating between types of dementia. This test can tell you whether you have dementia that is more like Alzheimer's disease or vascular dementia, and it can also help determine whether you have dementia that is due to other diseases.

[0012] However, the above-mentioned conventional techniques have the following problems.

[0013] Dementia such as Alzheimer's disease is an irreversible disease, and no cure has yet been developed. Therefore, the best approach is to diagnose dementia early and slow its progression. However, diagnosing dementia through detailed medical history taking and evaluation by specialists has been problematic, as it means that a shortage of specialists means that many patients cannot be diagnosed quickly, and at the same time, it increases the cost of screening the entire population for dementia at the national level.

[0014] In addition, conventional brain imaging diagnostics used to diagnose dementia involve expensive precision equipment, but can only detect brain disease or cerebral atrophy when it has progressed significantly, making it difficult to use for the purpose of early diagnosis of dementia.

[0015] In addition, dementia patients are usually elderly and illiterate or have low educational backgrounds, so they often have difficulty recognizing letters, making it difficult to conduct tests using questionnaires, etc., and there is a problem that the tests require a lot of time and manpower.

[0016] Therefore, there is an urgent need to develop appropriate and innovative technology that can diagnose dementia easily, quickly, and at an early stage. Summary of the Invention [Problem to be solved by the invention]

[0017] Therefore, the present invention has been devised to solve the above-mentioned conventional problems,

[0018] The object of the present invention is to provide a dementia diagnosis system and method that uses a story-based question-and-answer format assessment that enables dementia diagnosis to be performed quickly and easily without the need for specialized personnel, enables early diagnosis of dementia and screening of mildly affected patients to be performed quickly and accurately, and enables assessment through questions and answers via voice so that even elderly subjects who have difficulty recognizing letters can easily undergo screening. [Means for solving the problem]

[0019] In order to achieve the same object as above, the "dementia diagnostic system using story-based question-and-answer assessment" according to the present invention comprises a data input device for inputting text data for an actual story, a data storage device in which the text data input via the data input device is stored and which is capable of storing dementia information for each test subject, an audio output device for outputting diagnostic audio data to the outside so that the test subject can hear it, an audio receiver for receiving answers spoken by the test subject, a story data formation unit connected to the data input device, the data storage device, the audio output device and the audio receiver and which automatically forms dramatized story data, question data and correct answer data based on the actual story input as text data, an answer data formation unit receiving the answers of the test subject and forming answer data, a dementia discrimination unit which compares and analyzes the correct answer data and the answer data to check the error rate and determine the presence or absence of dementia and / or the state of dementia depending on the level of the error rate, and which sequentially outputs dramatized story data and question data and responds to the output questions. and a controller for controlling the data storage device, the audio output device, and the audio receiver to receive answers from the test subjects and store the identified dementia information for each test subject. The actual story is in the form of a story centered on a person already known to the test subject and an event related to the person, and is made up of a plurality of sentences. The dramatized story data is data obtained by extracting some of the noun-form words included in the actual story, changing the extracted noun-form words to other noun-form words of the same or similar type to form an dramatized story, and converting the dramatized story into audio data. The question data is data obtained by setting the changed noun-form words included in the dramatized story as correct answers, forming a plurality of questions related to the dramatized story, and converting the questions into audio data. The correct answer data is data obtained by extracting correct answers corresponding to the questions and converting the extracted correct answers into audio data. The answer data is audio data obtained by receiving and storing answers from the test subjects to the questions via the audio receiver.

[0020] Furthermore, the "dementia diagnosis system using story-based voice extraction" according to the present invention includes a data storage device that stores diagnostic voice data including story data for diagnosing dementia, and that stores patient voice feature data obtained by extracting the voice features of dementia patients by comparing and analyzing the voice features of normal people and the voice features of dementia patients, and that can store dementia information for each test subject; an audio output device that outputs the diagnostic voice data stored in the data storage device to the outside so that the test subject can hear it; an audio receiver that receives the voice spoken by the test subject and converts it into audio data; and a control device that is connected to the data storage device, the audio output device, and the audio receiver and controls the data storage device, the audio output device, and the audio receiver to receive the voice of the test subject after outputting the story data, and that, after the output of the story data is completed, recalls the content of the story output to the test subject and speaks it out loud. a speech request unit that requests the test subject to speak, a voice feature analysis unit that analyzes voice features from voice data received from the test subject, a dementia determination unit that compares the voice features of the test subject analyzed by the voice feature analysis unit with the voice feature data of the patient to determine the presence or absence of dementia and / or the state of dementia, and a controller that controls the storage of dementia information for each test subject determined by the dementia determination unit in the data storage device, wherein the story data is in the form of a story centered on a person already known to the test subject and an event related to this person, forming an actual story consisting of a plurality of sentences, extracting some of the noun-form words included in the actual story, and changing the extracted noun-form words to other noun-form words of the same or similar type to form an dramatized story, and converting the dramatized story into voice data. [Effects of the Invention]

[0021] The present invention as described above enables quick and easy dementia diagnosis without the need for specialized personnel, enables early diagnosis of dementia and screening of mild cases quickly and accurately, and allows elderly subjects who have difficulty recognizing letters to easily undergo screening by conducting an assessment in a question-and-answer format through audio or by presenting an audio story. As a result, it has the effect of enabling screening of hundreds of thousands of dementia patients among the entire population at a national level quickly and easily at a relatively low cost. [Brief explanation of the drawings]

[0022] [Figure 1] FIG. 1 is a schematic diagram showing the configuration of a dementia diagnosis system using a story-based question-and-answer style evaluation according to the present invention. [Figure 2A] FIG. 2A is a schematic configuration diagram showing a controller of a dementia diagnostic system using a story-based question-and-answer style assessment according to the present invention. [Figure 2B] FIG. 2B is a schematic configuration diagram showing a controller of the dementia diagnostic system using a story-based question-and-answer style assessment according to the present invention. [Figure 3] FIG. 3 is an illustrative diagram showing an example of an actual story used in the dementia diagnosis system according to the present invention. [Figure 4] FIG. 4 is an illustrative diagram showing an example of an adapted story used in the dementia diagnosis system according to the present invention. [Figure 5] FIG. 5 is an illustrative view showing an example of a story notification used in the dementia diagnosis system according to the present invention. [Figure 6] FIG. 6 is an illustrative diagram showing an example of a question notification text used in the dementia diagnosis system according to the present invention. [Figure 7] FIG. 7 is an illustrative diagram showing an example of a question used in the dementia diagnosis system according to the present invention. [Figure 8] FIG. 8 is a schematic configuration diagram showing a dementia diagnosis system according to another embodiment of the present invention. [Figure 9]FIG. 9 is a schematic configuration diagram showing a controller of a dementia diagnosis system according to another embodiment of the present invention. [Figure 10] FIG. 10 is a flow chart showing the dementia diagnosis method using a story-based question-and-answer style assessment according to the present invention. [Figure 11] FIG. 11 is a schematic diagram showing the configuration of a dementia diagnosis system using story-based voice extraction according to the present invention. [Figure 12] FIG. 12 is a schematic diagram showing the configuration of a controller of the dementia diagnosis system using story-based voice extraction according to the present invention. [Figure 13] FIG. 13 is a flow chart illustrating the method for diagnosing dementia using story-based voice extraction according to the present invention. BEST MODE FOR CARRYING OUT THE INVENTION

[0023] The present invention relates to a dementia diagnostic system that uses a story-based question-and-answer format assessment, which enables dementia diagnosis to be performed quickly and easily without the need for specialized personnel, enables early diagnosis of dementia and rapid and accurate screening of mildly affected patients, and conducts an assessment in a question-and-answer format via voice, making it easy for even elderly subjects who have difficulty recognizing characters to undergo screening. The dementia diagnostic system includes a data input device 10 that can input an actual story, a data storage device 20 that stores the text data input via the data input device 10, an audio output device 30 that can output diagnostic audio data to the outside as audio, an audio receiver 40 that receives the audio responses of the subject, and a controller 50 that is connected to various devices including the data input device 10 and controls them, creates an adapted story, questions, and correct answers based on the actual story, receives the answers of the subject, and compares the answers with the correct answers to determine dementia. DETAILED DESCRIPTION OF THE INVENTION

[0024] DETAILED DESCRIPTION OF THE INVENTION The present invention will now be described more fully hereinafter with reference to the accompanying drawings, in which preferred embodiments of the present invention are shown. It should be understood, however, that the present invention may be embodied in many different forms and is not limited to the described embodiments.

[0025] 1 to 7 are diagrams showing a dementia diagnostic system according to the present invention. As shown in the figures, the dementia diagnostic system using a story-based question-and-answer format assessment according to the present invention includes a data input device 10 for inputting an actual story, a data storage device 20 for storing the character data input via the data input device 10, an audio output device 30 for outputting diagnostic audio data as audio to the outside, an audio receiver 40 for receiving the test subject's audio responses, and a controller 50 connected to and controlling various devices including the data input device 10, forming an adapted story, questions, and correct answers based on the actual story, receiving the test subject's answers, and comparing the answers with the correct answers to determine dementia.

[0026] The data input device 10 is an input device that can input normal data into a storage device, and allows a user to input text data for an actual story into the data storage device 20 via the controller 50.

[0027] Here, the actual story is composed of multiple sentences in the form of a story centered on a person already known to the test subject and an event related to this person, and is created and prepared by a producer or operator.

[0028] The actual story is preferably a fairy tale or fable that the test subject is already familiar with. For example, the actual story may be a fairy tale such as Snow White shown in FIG. 3 or a fable such as Aesop's Fables, and a summary of the story may be used within 15 sentences.

[0029] In this case, "the subject already knows" means that the story is a classic, a traditional, or a widely known fairy tale or fable, and the subject has known the plot for a long time and can easily recall it.

[0030] It is preferable that a plurality of such real stories be prepared and various fairy tales and fables be used. This is to enable the accuracy to be improved by using different stories when repeatedly testing the same person, and also to enable fairy tales and fables that the test subject knows to be appropriately selected and used for diagnosis.

[0031] The data storage device 20 is a storage device capable of storing digital data, like a memory or hard disk, and stores various data including character data entered through the data input device 10, thereby enabling the storage of dementia information for each test subject.

[0032] The audio output device 30 is capable of outputting digital data as audio, like a speaker, and outputs diagnostic audio data to the outside so that the subject can hear it. Here, the diagnostic audio data refers to audio data of an adapted story and questions that are output to determine dementia.

[0033] The voice receiver 40 receives voice like a microphone and serves to receive answers spoken by the test subject.

[0034] The controller 50 is connected to the data input device 10, data storage device 20, audio output device 30 and audio receiver 40, and has a story data formation unit 51 that automatically forms dramatized story data, question data and correct answer data based on an actual story input as text data, an answer data formation unit 52 that receives the answers of the test subject and forms answer data, and a dementia discrimination unit 53 that compares and analyzes the correct answer data and answer data to confirm the error rate and discriminates the presence or absence of dementia and / or the state of dementia depending on the level of the error rate.

[0035] The controller 50 controls the data storage device 20, the audio output device 30, and the audio receiver 40 to sequentially output dramatized story data and question data, receive answers from the test subjects to the output questions, and store the dementia information for each of the test subjects that has been identified.

[0036] The story data forming unit 51 plays a role of automatically forming dramatized story data, question data and correct answer data based on the character data of the input actual story.

[0037] Here, the dramatized story data is data obtained by extracting some of the noun-form words contained in the actual story, changing the extracted noun-form words to other noun-form words of the same or similar type to form an dramatized story, and converting the dramatized story into audio data.

[0038] That is, the dramatized story is formed by extracting some of the noun-form words contained in the actual story and replacing the extracted noun-form words with other words of similar types.

[0039] As an example, as shown in Figure 3, the noun words extracted from the actual story are displayed in red squares. The extracted noun words shown in red squares in Figure 3 are "Snow White," "Stepmother," "Seven," "Dwarfs," "Apple," "Horse," and "Prince."

[0040] In this way, the noun-form words extracted from the actual story are changed to other words of similar types, such as “Kuroyukihime,” “aunt,” “five people,” “miner,” “peach,” “donkey,” and “miner,” as shown in Figure 4 .

[0041] That is, "Snow White" in the actual story is changed to "Snow Black," "stepmother" in the actual story is changed to "aunt," "seven" in the actual story is changed to "five," "dwarfs" in the actual story are changed to "miners," "apples" in the actual story are changed to "peaches," "horses" in the actual story are changed to "donkeys," and "princes" in the actual story are changed to "miners."

[0042] In this way, by changing some of the noun-form words in the actual story presented in Figure 3, the dramatized story shown in Figure 4 is completed.

[0043] It is preferable that the dramatized story consists of 5 to 15 sentences and contains 20 to 50 different noun-shaped words. This is because if the number of sentences is less than 5 or if the number of noun-shaped words is less than 20, the overall length of the dramatized story is too short to be easily memorized, and therefore the short-term memory of the test subject cannot be effectively tested or dementia patients cannot be appropriately screened.

[0044] Furthermore, if the dramatized story contains more than 15 sentences or more than 50 noun-like words, the dramatized story itself becomes too long and difficult to remember, which could lead to misdiagnosis of ordinary people being classified as dementia patients.

[0045] The dramatized story preferably includes at least one artificially created additional sentence that is different from the content of the actual story and is related to and linked to the content of the dramatized story.

[0046] An example of such an additional sentence is shown in part A of Figure 4, which serves to confuse and disorient the test subject, thereby hindering their memory. In other words, the test subject, especially a dementia patient, will be unable to remember the entire content of the dramatized story due to the memory interference caused by the additional sentence, and accordingly, screening for dementia patients will be more accurate.

[0047] Here, the additional text being different from the content of the actual story means that it is not in the content of the actual story, and the additional text being related to and connected to the content of the dramatized story means that it is connected to the development of the story of the dramatized story. Such additional text may be created directly by the creator and input through the data input device 10, or may be content that adds excitement by including the main character, and may be automatically generated by the controller 50.

[0048] Furthermore, the additional sentence is preferably placed at the end of the dramatized story as shown in FIG. 4, in order to effectively interfere with the test subject's memory of the dramatized story.

[0049] The question data is data in which a plurality of questions related to the dramatized story are formed, with the changed noun form words contained in the dramatized story being the correct answers, and the questions are converted into voice data.

[0050] Here, it is preferable that the number of questions consists of 3 to 12 questions. This is because if the number of questions is less than 3, the discrimination power is low and it is not possible to properly test the memory of the test subject, and if the number of questions is more than 12, it may exceed the short-term memory of the test subject, making it possible to perform the memory test inaccurately.

[0051] Furthermore, the questions are preferably questions about whether the events, actions, situations and contents presented in the dramatized story are factual, so that short-term memory can be adequately tested through simple, straightforward, fact-centered questions.

[0052] Furthermore, the questions are preferably multiple-choice questions in which the incorrect answers include noun words contained in the actual story that are the same as or similar to the correct answers, as shown in Figure 7. For example, in the case of a multiple-choice question as shown in Figure 7, the noun words contained in the actual story, namely "Snow White," "Stepmother," "Seven," "Dwarfs," "Apple," "Horse," and "Prince," are included as incorrect answers among the four options, causing confusion in the test subject's memory, and causing the test subject to select the content of a fairy tale that the test subject is already familiar with as an incorrect answer, thereby enabling the test subject to be properly identified as a dementia patient.

[0053] In other words, some test subjects do not fully remember the content of the fictionalized story that contains the correct answer to the question, but instead remember the content of the actual story that they have remembered for a long time and choose their answer based on that, which leads to incorrect answers.Those test subjects who are prone to making such incorrect answers are diagnosed and screened as having dementia.

[0054] Furthermore, it is preferable that the final question be one that checks the subject's accuracy. Such a question for checking the subject's accuracy is presented in the section shown in Figure 7, D, and allows the operator to evaluate the subject's metacognitive ability, which is the ability to recognize for oneself what the subject knows and what the subject does not know. Such an evaluation of metacognitive ability can be used as a reference for further checking the accuracy of dementia diagnosis by this system.

[0055] The correct answer data is data obtained by extracting the correct answers corresponding to the questions and converting the extracted correct answers into voice data. The correct answer data is used as a reference for comparing the answer data collected from the test subjects and checking the error rate.

[0056] Here, the correct answers refer to noun-form words extracted from the actual story and changed from the dramatized story. Taking the dramatized story in Figure 4 as an example, the correct answers are "Kuroyukihime," "Aunt," "Five people," "Miner," "Peach," "Donkey," and "Miner," which are displayed in red boxes.

[0057] The response data generating unit 52 receives the test subject's responses to the questions via voice and automatically generates response data. Here, the response data is voice data received and stored from the test subject's responses to the questions via the voice receiver 40.

[0058] The dementia determination unit 53 plays a role of comparing the generated correct answer data with the answer data and determining the presence or absence of dementia and / or the state of dementia based on the rate of incorrect answers to questions.

[0059] In other words, the test subjects will confuse the questions about the dramatized story of "Snow White" with the actual story of "Snow White" that they already know, which will induce them to give incorrect answers.If they receive many incorrect answers in this way and the error rate is high, they will be diagnosed as having dementia.

[0060] The dementia detection unit 53 compares and analyzes the voice of the correct answer data with the voice of the answer data to confirm whether the answer is correct or incorrect, in order to more accurately determine whether the answer is incorrect through voice recognition and voice comparison.

[0061] Here, the correct answer to the question is displayed in a red box in Figure 7, and any options other than the correct answer are naturally treated as incorrect answers. In Figure 7, the correct answer to "Question 8" is confirmed by evaluating "Questions 1" to "Questions 7," and the incorrect answer rate is the percentage of incorrect answers given to all questions.

[0062] It is preferable that the dementia discrimination unit 53 discriminates dementia if the error rate is at least 25% or higher, mild dementia if the error rate is 25-40%, and severe dementia if the error rate is at least 40% or higher; this was determined through comparative analysis and confirmation of the error rates of test subjects belonging to the normal group and the dementia group through clinical experiments using a large number of test subjects.

[0063] In other words, if the error rate was higher than 25%, all patients were confirmed as having dementia, and if the error rate was between 25 and 40%, they were classified as having mild dementia, and if it was higher than 40%, they were classified as having severe dementia.

[0064] The controller 50 notifies the output of the adapted story, and receives input of a story notification message notifying that a question regarding the adapted story will be presented after the adapted story is output, converts it into audio data, and controls the audio output device 30 to output the story notification message first before outputting the adapted story.

[0065] The story notification text is created and input as character data by a creator or an operator, and is converted into voice data and output.

[0066] An example of such a story notification is shown in FIG. 5, and the output of the story notification serves to notify the test subject in advance that an adapted story will be output aloud and questions will be presented that must be answered.

[0067] It is preferable that the story notification sentence begins with a sentence including the title of the fairy tale or fable presented in the dramatized story. The title of such a fable or fairy tale is presented in the portion ("Snow White") shown in B of Fig. 5, and by announcing the title of the fairy tale to be output aloud in the story notification sentence, for example, "Snow White," the test subject can recall the related fairy tale "Snow White" in advance and easily memorize the entire plot of "Snow White."

[0068] The controller 50 receives an input of a question notification message notifying that a predetermined number of questions will be presented, converts it into voice data, and controls the voice output device 30 to output the question notification message after the output of the dramatized story is completed.

[0069] The question notification text is created by a creator or an operator, input as character data, converted into voice data, and output.

[0070] The output of the question notification message serves to inform the test subject that questions will be output immediately after the dramatized story is output and that the test subject must answer them, and at the same time, it notifies the test subject in advance of the number of questions, as shown in part C of Figure 6 (7 questions), so that the test subject can respond appropriately to the questions that are output.

[0071] In this dementia diagnosis system, the actual stories are prepared in multiple pieces, each of which is input and formed into multiple dramatized stories, and questions and answers related to each dramatized story can be formed, respectively. This is to enable various question-and-answer evaluations to be performed through the multiple actual stories, thereby enabling more accurate dementia diagnosis and discrimination.

[0072] With a plurality of real stories input in this manner, the controller 50 further includes a recognition story confirmation unit 54 that outputs a story confirmation message by voice to the test subject to confirm which of the plurality of dramatized stories the test subject is already familiar with before the dramatized story is output, and a story selection unit 55 that selects an adaptation story that the test subject has confirmed to be already familiar with through the recognition story confirmation unit 54 so that the adaptation story confirmed through the recognition story confirmation unit 54 is output.

[0073] The cognitive story confirmation unit 54 outputs a story confirmation sentence containing, for example, the titles of multiple dramatized stories (e.g., "Snow White," "Cinderella," "The Little Mermaid," "Pinocchio," and "The Prince and the Pauper") in voice, so that the person being tested can confirm what fairy tales or fables are included as dramatized stories stored in the diagnostic device.

[0074] The story selection unit 55 allows the subject to listen to the story confirmation message and select an adapted story that the subject already knows. For example, adapted stories such as "Snow White," "Cinderella," "The Little Mermaid," "Pinocchio," and "The Prince and the Pauper" are stored, and the titles of these stories are output. If the subject already knows "Cinderella," the subject can select it by answering "Cinderella" by voice or by operating a button.

[0075] In this way, when "Cinderella" is selected, an adapted story and questions related to "Cinderella" are output to the test subject for the purpose of diagnosing dementia, and the dementia diagnosis is made based on fairy tales and fables that the test subject is already familiar with, while at the same time appropriately preventing a dementia diagnosis based on fairy tales and fables that the test subject is not familiar with, thereby improving the accuracy of the dementia diagnosis.

[0076] The controller 50 further includes a used story confirmation unit 56 that outputs the dramatized story to the test subject by voice before the dramatized story is output and confirms which of the multiple dramatized stories has been previously used in a diagnostic test, and a story exclusion unit 57 that excludes the dramatized story confirmed to have been previously used through the used story confirmation unit 56 from the output.

[0077] The used story confirmation unit 56 plays a role in confirming which of the multiple dramatized stories has been previously used in the diagnostic test of the test subject, and the story exclusion unit 57 plays a role in excluding and filtering out the dramatized story that has been confirmed to have been used so that it is not output to the test subject.

[0078] By eliminating dramatized stories in which the test subject has been used in diagnostic tests in this way, a decrease in diagnostic accuracy due to the repeated use of the same content and the same questions is appropriately prevented.

[0079] The dementia diagnosis system configured as described above may be configured as an integrated dementia diagnosis machine in which the data input device 10, data storage device 20, audio output device 30, audio receiver 40 and controller 50 are modularized and integrated, or may be configured as an automatic dementia diagnosis machine in which all functions are programmed or automated using artificial intelligence or the like.

[0080] As mentioned above, this dementia diagnostic system is a groundbreaking invention that allows for rapid and accurate screening of dementia by replacing the main words in a classic true story that the test subject already knows and is familiar with with new words and presenting it as an dramatized story to harden the memory, while presenting the original words in the classic true story as incorrect answers to maximize the memory interference effect.

[0081] 8 and 9 are schematic diagrams showing a dementia diagnosis system according to another embodiment of the present invention. As shown in the figures, the dementia diagnosis system according to another embodiment of the present invention further includes a display 60 connected to the controller 50 and capable of displaying character data on a screen, and a character input device 70 connected to the controller 50 and capable of inputting characters to correct the character data.

[0082] The display 60 is capable of displaying characters and images on a screen like a monitor, and plays a role in displaying the dramatization data and question data as character data.

[0083] The character input device 70 is a device that allows characters to be input and corrected like a keyboard, and allows the character data of the dramatization data and question data to be corrected.

[0084] The controller 50 further includes a story checking and correcting unit 58 that displays the adapted story as text data on the display 60 so that the producer or operator can check and manually correct it via the text input device 70, and a question checking and correcting unit 59 that displays the question as text data on the display 60 so that the producer or operator can check and manually correct it via the text input device 70.

[0085] The story confirmation and correction unit 58 allows a producer, operator, or other person to directly check and correct the adapted story automatically adapted by the controller 50, thereby creating a high-quality adapted story in which events unfold more naturally and smoothly.

[0086] The question confirmation and correction unit 59 allows a person such as a creator or an operator to directly check and correct the questions automatically generated by the controller 50, thereby generating natural, smooth questions with adjusted difficulty levels.

[0087] The dementia diagnosis system further includes a data transmitter 80 connected to the data storage device 20 and controlled by the controller 50. The data transmitter 80 is, for example, a wired or wireless transceiver that is controlled by the controller 50 and enables the dementia information for each test subject stored in the data storage device 20 to be transmitted to an external device via a transmission line or a network.

[0088] FIG. 10 is a flow chart showing the dementia diagnosis method according to the present invention.

[0089] 10, the dementia diagnosis method using a story-based question-and-answer format assessment according to the present invention includes an actual story preparation step S1 of preparing an actual story, an actual story input step S2 of converting the actual story into text data and inputting the text data into a diagnostic device, an dramatized story formation step S3 of dramatizing the input actual story to form an dramatized story, a question formation step S4 of forming a plurality of questions whose correct answers are modified noun-form words included in the dramatized story, a voice data conversion step S5 of converting the dramatized story and the questions into voice data, an dramatized story output step S6 of outputting the voice data of the dramatized story to an external device, a question output step S7 of outputting the voice data of the questions to an external device in sequence, an answer receiving step S8 of receiving the test subject's answers to the questions, and a dementia discrimination step S9 of discriminating the presence or absence of dementia based on the error rate of the questions and answers. These steps are sequentially performed to diagnose dementia.

[0090] In the question formation step S4, the questions consist of 3 to 12 questions, which are questions about whether the events, actions, situations, and contents presented in the dramatized story are factual or not, and are multiple-choice questions in which the incorrect answers include noun-form words contained in the actual story of the same or similar type as the correct answers, and a question to check the accuracy rate of the test subject himself / herself may be placed as the final question.

[0091] In the real story preparation step S1, the real story can be a fairy tale or fable that the test subject is already familiar with.

[0092] In the adaptation story formation step S3, the adaptation story may consist of 5 to 15 sentences and include 20 to 50 different noun-form words, and the adaptation story formation step S3 may further include a sentence addition process of adding at least one artificially created additional sentence that is different from the content of the actual story but is related to and connected to the content of the adaptation story, and the additional sentence may be placed at the end of the adaptation story.

[0093] The adapted story output step S6 may further include a story notification output process for outputting a story notification message by voice before outputting the adapted story, and for notifying the test subject that a question related to the adapted story will be presented after the adapted story is output. The adapted story may be a fairy tale or fable that the test subject is already familiar with, and the story notification output process may further include a process for outputting a sentence including the title of the fairy tale or fable presented in the adapted story at the beginning of the story notification message.

[0094] The adapted story output step S6 may further include a question notification output step of outputting a question notification message by voice, which notifies the user that a predetermined number of questions will be presented after the adapted story output is completed.

[0095] In the dementia discrimination step S9, if the error rate is at least 25% or higher, it is determined to be dementia, if the error rate is 25-40%, it is determined to be mild dementia, and if the error rate is at least 40% or higher, it is determined to be severe dementia.

[0096] A plurality of actual stories are prepared and input, and a plurality of dramatized stories are formed, and questions related to each dramatized story are formed, respectively. The dramatized story output step S6 may further include a cognitive story confirmation process in which a story confirmation statement is output by voice to the test subject before the dramatized story is output, to confirm which of the plurality of dramatized stories the test subject already knows, and a story selection process in which the test subject selects an dramatized story that is confirmed to be already known by the test subject through the cognitive story confirmation process, so that the dramatized story confirmed through the cognitive story confirmation process is output.

[0097] The actual stories are prepared in multiple numbers and input, and multiple dramatized stories are formed, and questions related to each of the dramatized stories are formed, respectively. The dramatized story output step S6 may further include a used story confirmation process for confirming with the test subject which of the multiple dramatized stories has been previously used in a diagnostic test before the dramatized story is output, and a story elimination process for excluding from the output any dramatized story that has been confirmed to have been previously used through the used story confirmation process.

[0098] The dramatized story forming step S3 may further include a story checking and correcting step in which the dramatized story formed by text data is displayed on a display so that a producer or an operator can check and manually correct the story.

[0099] The question formation step S4 may further include a question confirmation and modification step in which the question formed by text data is displayed on a display so that a creator or an operator can confirm and manually modify the question.

[0100] By transmitting dementia information for each test subject using the data transmitter 80, information on diagnosed and screened dementia patients can be easily provided to the management servers of national and private institutions for dementia management and screening.

[0101] 11 and 12 are diagrams showing a dementia diagnosis system according to the present invention. As shown in the figures, the dementia diagnosis system using story-based voice extraction according to the present invention includes a data input device 100, a data storage device 200 that stores diagnostic voice data and patient voice feature data for diagnosing dementia, an audio output device 300 that outputs the diagnostic voice data stored in the data storage device 200 as audio to the outside, an audio receiver 400 that receives the voice of the person being tested, and a controller 500 that is connected to the data storage device 200, the audio output device 300, and the audio receiver 400 to control them and compare the voice features of the person being tested with the patient's voice feature data to determine dementia.

[0102] The data storage device 200 is capable of storing digital data like a memory or hard disk, and stores diagnostic voice data including story data for dementia diagnosis, patient voice feature data extracted from the voice features of dementia patients by comparing and analyzing the voice features of normal people with the voice features of dementia patients, and is capable of storing dementia information for each test subject.

[0103] Here, the patient voice feature data is data of common voice features extracted from the voices of many dementia patients, which are distinguished from the voice features of normal people.

[0104] The voice characteristics of the normal person and the dementia patient include one or more of fundamental frequency of voice, speaking rate, shimmer, jitter, formant and voice spectrum.

[0105] The fundamental frequency of the voice refers to a frequency that indicates the average and habitual pitch of the voice, and the speaking rate refers to the speed of words and the speed of articulation, and indicates the speed at which a person speaks.

[0106] The shimmer (shimmer0) is a numerical representation of the regularity of voice amplitude change, the jitter (jitter0) is a numerical representation of the rate of change of vocal cord vibration, the formant (formant0) refers to frequency components that characterize vowels, and the voice spectrum represents the distribution of voice frequencies.

[0107] Furthermore, the story data is in the form of a story centered on a person already known to the test subject and events related to this person, and forms an actual story consisting of a plurality of sentences. Some of the noun-form words contained in the actual story are extracted, and the extracted noun-form words are changed to other noun-form words of the same or similar type to form an dramatized story, and the dramatized story is converted into audio data.

[0108] The story data is an adapted story converted into audio data, and is intended to enable the adapted story to be easily output and provided to the test subject later in audio format.

[0109] The actual story is preferably a fairy tale or fable that the test subject is already familiar with. For example, the actual story may be a fairy tale such as "Snow White" or a fable such as Aesop's Fables, as shown in FIG. 3, and a summary of the story may be used within 15 sentences.

[0110] In this case, when we say that the subject already knows the story, we mean that the story is a classic, or a traditional or widely known fairy tale or fable, and the subject has already known the plot for a long time and can easily recall it.

[0111] The dramatized story is an adaptation of the actual story, in which some of the noun words contained in the actual story are extracted and the extracted noun words are replaced with other words of similar types. As an example, as shown in Figure 3, the noun words extracted from the actual story are displayed in red squares. The extracted noun words displayed in red squares are "Snow White," "Stepmother," "Seven," "Dwarfs," "Apple," "Horse," and "Prince."

[0112] In this way, the noun-form words extracted from the actual story are changed to other words of similar types, such as “Kuroyukihime,” “Aunt,” “Five,” “Miner,” “Peach,” “Donkey,” and “Miner,” as shown in Figure 4 .

[0113] That is, "Snow White" in the actual story is changed to "Snow Black," "stepmother" in the actual story is changed to "aunt," "seven" in the actual story is changed to "five," "dwarfs" in the actual story are changed to "miners," "apples" in the actual story are changed to "peaches," "horses" in the actual story are changed to "donkeys," and "princes" in the actual story are changed to "miners."

[0114] When changing to such similar words, the change can be made automatically by extracting similar words, or the creator or administrator can directly intervene and change the words. Here, it is preferable that the creator or administrator directly change the words or intervene to change some words to make the adaptation more natural.

[0115] By changing some of the noun forms presented in the actual story, the dramatized story shown in Figure 4 is completed.

[0116] It is preferable that the dramatized story consists of 5 to 15 sentences and contains 20 to 50 different noun-shaped words. This is because if the number of sentences is less than 5 or if the number of noun-shaped words is less than 20, the overall length of the dramatized story is too short to be easily memorized, and therefore the short-term memory of the test subject cannot be effectively tested or dementia patients cannot be appropriately screened.

[0117] Furthermore, if the dramatized story contains more than 15 sentences or more than 50 noun-like words, the dramatized story itself becomes too long and difficult to remember, which could lead to misdiagnosis of ordinary people as having dementia.

[0118] Here, if the number of sentences in the actual story exceeds 15, it is preferable that the producer or manager intervene in the process of changing the story into the dramatized story to reduce the number of sentences, or prepare the actual story itself to have 15 sentences or less.

[0119] In addition, the dramatized story preferably includes at least one artificially created additional sentence that is different from the content of the actual story and is related to and linked to the content of the dramatized story.

[0120] An example of such an additional sentence is presented in the portion shown in Figure 4A, which serves to confuse and disorient the test subject, thereby hindering their memory. In other words, the test subject, especially a dementia patient, will be unable to remember the entire content of the dramatized story due to the memory interference caused by the additional sentence, and accordingly, screening for dementia patients will be more accurate.

[0121] Here, the additional text being different from the content of the actual story means that it is content that is not in the content of the actual story, and the additional text being content that is related to and connected to the content of the dramatized story means that it is content that is connected to the development of the story of the dramatized story. Such additional text may be created directly by the creator, or may be created automatically as it is content that adds excitement by including the main character.

[0122] Furthermore, the additional sentence is preferably placed at the end of the dramatized story as shown in FIG. 4, in order to effectively interfere with the test subject's memory of the dramatized story.

[0123] The dramatized story thus completed is converted into audio data and stored in the data storage device 200, and is later easily outputted and provided to the test subject via the audio output device 300.

[0124] This type of voice data conversion involves either having a person read and speak the dramatized story and converting it into digital data using various digital audio collection devices or recorders, or directly converting text into voice data using mechanical sound via a text-to-speech conversion device.

[0125] In addition, the data storage device 200 can store question data and answer data related to the dramatized story.

[0126] The question data is data in which the modified noun form words contained in the dramatized story are used as correct answers, a plurality of questions related to the dramatized story are formed, and the questions are converted into audio data, which is used to test the memory of the test subject and to elicit speech.

[0127] The correct answer data is data of the correct answer to the question, and corresponds to the changed noun form words contained in the dramatized story.

[0128] The number of questions is preferably 3 to 12. This is because if there are fewer than 3 questions, the subject will not speak much, resulting in low discrimination power and making it impossible to properly test the subject's memory. If there are more than 12 questions, the subject's short-term memory may be exceeded, making it difficult to accurately test the memory.

[0129] Furthermore, the questions are preferably questions about whether the events, actions, situations, and contents presented in the dramatized story are factual, which allows for an appropriate test of short-term memory through simple, straightforward, fact-centered questions and allows the subject's speech to be easily guided.

[0130] The questions may be questions that request answers in sentences or words, which encourages the test subject to answer in short question or short answer format, thereby ensuring more voice recordings for dementia diagnosis and more accurate extraction of voice features.

[0131] Furthermore, the question may be a multiple-choice question in which the incorrect answers include noun words contained in the actual story that are the same as or similar to the correct answers, as shown in Figure 7. For example, in the case of a multiple-choice question like the question shown in Figure 7, noun words contained in the actual story, namely "Snow White," "Stepmother," "Seven," "Dwarfs," "Apple," "Horse," and "Prince," are included as incorrect answers among the four options, causing memory confusion in the test subject, and causing the test subject to select incorrect answers based on the content of a fairy tale that the test subject is already familiar with, thereby enabling the test subject to be properly identified as a dementia patient.

[0132] In other words, some test subjects do not fully remember the content of the dramatized story that contains the correct answer to the question, but instead memorize the content of the actual story that they have remembered for a long time and select the answer based on this, which induces the voice characteristics of dementia patients and incorrect answers, and test subjects who frequently induce such voice characteristics of dementia patients and incorrect answers are diagnosed and screened as dementia patients.

[0133] In addition, it is preferable that the last question be a question to check the subject's accuracy. Such a question to check the subject's accuracy is presented in the section shown in FIG. 7D, and allows the operator to evaluate the subject's metacognitive ability, which is the ability to recognize for oneself what the subject knows and what the subject does not know. Such an evaluation of metacognitive ability can be used as a reference to further confirm the accuracy of dementia diagnosis using this method.

[0134] The audio output device 300 can output digital data as audio like a speaker, and serves to output the diagnostic audio data stored in the data storage device 200 to the outside so that the subject can hear it.

[0135] While listening to the dramatized story of "Snow White" presented in FIG. 4 through the audio output device 300, the test subject confuses it with the actual story of "Snow White" presented in FIG. 3, which the test subject has already recognized and memorized, resulting in memory interference.

[0136] In this way, when the test subject is in a state where memory interference has occurred, the test subject is asked to recall the contents of the dramatized story that has been output to the test subject and speak it out loud. This creates an environment in which the test subject's memory is impaired or confusion is induced, and test subjects who cannot easily overcome this environment are quickly and accurately diagnosed and screened as having dementia.

[0137] In addition, the diagnostic test is conducted through audio output of the dramatized story, so even elderly test subjects who are not familiar with or have difficulty reading can easily be diagnosed for dementia.

[0138] The voice receiver 400 is capable of receiving voice and converting it into digital data, like a microphone, and serves to receive voice uttered by the subject and convert it into voice data.

[0139] That is, the voice of the test subject received by the voice receiver 400 is analyzed for voice features by the controller 500, and then compared with the patient voice feature data for use in dementia diagnosis.

[0140] The controller 500 is connected to the data input device 100, the data storage device 200, the audio output device 300 and the audio receiver 400, and controls the data storage device 200, the audio output device 300 and the audio receiver 400 to output story data and then receive the audio of the subject.

[0141] The controller 500 also has a speech request unit 410 that requests the test subject to recall the content of the story that has been output and speak it out loud after the output of the story data is completed, a voice feature analysis unit 420 that analyzes voice features from the voice data received from the test subject, and a dementia determination unit 430 that compares the voice features of the test subject analyzed through the voice feature analysis unit 420 with the patient voice feature data to determine the presence or absence of dementia and / or the state of dementia.

[0142] The controller 500 analyzes the voice characteristics of the test subject, performs dementia diagnosis based on the analyzed voice characteristics, and controls the storage of dementia information for each test subject determined by the dementia diagnosis unit 430 in the data storage device 200.

[0143] The speech request unit 410 requests the subject to recall the contents of the dramatized story and speak aloud, and serves to prompt and induce the subject to speak by informing the subject of the time to speak. The speech request unit 410 outputs a speech request content such as "Let's recall the contents of the story and speak aloud."

[0144] The voice feature analysis unit 420 serves to extract voice features of the test subject by analyzing the voice of the test subject collected through the voice receiver 400. That is, the voice feature analysis unit 420 serves to analyze and extract fundamental frequency, speaking rate, shimmer, jitter, formant, and voice spectrum of the voice from the voice data of the test subject.

[0145] The dementia detection unit 430 plays a role in determining whether the test subject has dementia by comparing the voice features of the test subject with patient voice feature data, which is the voice features of dementia patients. Here, when the test subject recalls and speaks the dramatized story of "Snow White," the test subject confuses it with the actual story of "Snow White" that the test subject already knows, which induces the voice features of dementia patients, and the test subject is determined to have dementia.

[0146] The controller 500 further includes an answer request unit 440 that outputs audio data of the questions to the outside after completing the output of the dramatized story to the test subject, and asks the test subject questions to request answers. The answer request unit 440 plays a role of checking the test subject's memory through questions and encouraging speech through answers.

[0147] The controller 500 further includes an auxiliary dementia determination unit 450 that compares and analyzes answers collected from the test subject to the questions with correct answers to check the error rate and determines the presence or absence of dementia and / or the state of dementia according to the degree of the error rate. The auxiliary dementia determination unit 450 plays a role in auxiliary determination of dementia based on the error rate of the test subject's answers to the questions.

[0148] Here, the correct answer to the question is shown in a red box in Figure 7, and any answer other than the correct answer is naturally treated as an incorrect answer. In Figure 7, the correct answer to "Question 8" is confirmed by evaluating "Questions 1" to "Questions 7," and the incorrect answer rate is the percentage of incorrect answers to all questions.

[0149] It is preferable that the controller 500 performs a combined dementia diagnosis based on the voice characteristics of the test subject and a dementia diagnosis based on the error rate of the test subject to determine whether the test subject has dementia, in order to improve the accuracy of the dementia diagnosis.

[0150] The controller 500 further includes a memory disruption unit 460 that disrupts the test subject's memory by allowing a certain time to pass after the output of the dramatized story is completed and before requesting the test subject to speak aloud. The memory disruption unit 460 disrupts the test subject's memory of the dramatized story by delaying the time, making it difficult for the test subject to remember the entire content of the dramatized story, thereby enabling more accurate screening of dementia patients.

[0151] In this case, the predetermined time period may be from a few seconds to several tens of seconds, and during this time, a sound such as music may be output to more actively interfere with the memory of the subject.

[0152] The dementia diagnosis system configured as described above may be configured as an integrated dementia diagnosis machine in which the data storage device 200, audio output device 300, audio receiver 400 and controller 500 are modularized and integrated, or may be configured as an automatic dementia diagnosis machine in which all functions are programmed or automated using artificial intelligence or the like.

[0153] The dementia diagnosis system further includes a data input device 100 connected to the data storage device 200 and controlled by the controller 500, and a data transmitter 600 connected to the data storage device 200 and controlled by the controller 500.

[0154] The data input device 100 is controlled by the controller 500 and allows various data, including diagnostic voice data, to be input and stored in the data storage device 200. In other words, it allows the operator to easily input various data, including diagnostic voice data, into the system, and allows changes, additions, and updates to diagnostic voice data to be made appropriately.

[0155] The data transmitter 600 is, for example, a wired or wireless transceiver that is controlled by the controller 500 and enables the dementia information for each test subject stored in the data storage device 200 to be transmitted to an external device via a transmission line or a network.

[0156] By transmitting dementia information for each test subject using the data transmitter 600, information on diagnosed and screened dementia patients can be easily provided to the management servers of national and private institutions for dementia management and screening.

[0157] This dementia diagnosis system, which is configured in this way, replaces the main words of a classic real story that the test subject already knows and is familiar with with new words, presents it as an dramatized story, and has the subject recall and recite the dramatized story in a state that hardens the memory. This maximizes the memory interference effect between the dramatized story and the classic real story, making it a groundbreaking invention that enables rapid and accurate screening for dementia.

[0158] The device includes a data storage device (20) that stores diagnostic voice data including story data for diagnosing dementia, and that stores patient voice feature data obtained by extracting voice features of dementia patients by comparing and analyzing voice features of normal people and voice features of dementia patients, and that can store dementia information for each test subject; a voice output device (30) that outputs the diagnostic voice data stored in the data storage device (20) to the outside so that the test subject can hear it; a voice receiver (40) that receives voices uttered by the test subject and converts them into voice data; and a speech requesting unit (41) that is connected to the data storage device (20), the voice output device (30), and the voice receiver (40) and controls the data storage device (20), the voice output device (30), and the voice receiver (40) to receive the voices of the test subject after outputting the story data, and requests the test subject to recall the content of the output story and speak aloud after completing the output of the story data. a dementia discrimination unit (43) that compares the voice features of the test subject analyzed by the voice feature analysis unit (42) with the patient voice feature data to determine the presence or absence of dementia and / or the state of dementia; and a controller (50) that controls the storage of dementia information for each test subject determined by the dementia discrimination unit (43) in the data storage device (20), wherein the story data is in the form of a narrative centered on a person already known to the test subject and events related to this person, forming an actual story consisting of a plurality of sentences, extracting some of the noun-form words included in the actual story, and dramatizing the extracted noun-form words by changing them to other noun-form words of the same or similar type, to form an dramatized story, and converting the dramatized story into voice data.

[0159] The data storage device (20) stores question data and correct answer data related to the dramatized story, the question data being data obtained by setting modified noun-form words included in the dramatized story as correct answers, forming a plurality of questions related to the dramatized story, and converting the questions into audio data, and the controller (50) further includes an answer request unit (44) that, after the output of the dramatized story to the test subject is completed, sequentially outputs the audio data of the questions to the outside, asks the test subject questions, and requests an answer.

[0160] A dementia diagnosis system using story-based speech extraction, characterized in that the questions consist of 3 to 12 questions requesting answers as to whether the events, actions, situations, and content presented in the dramatized story are factual.

[0161] A dementia diagnosis system using story-based speech extraction, characterized in that the questions are questions that request answers in sentences or words.

[0162] A dementia diagnosis system using story-based voice extraction, characterized in that the questions are multiple-choice questions in which the incorrect answers include noun-form words contained in the actual story of the same or similar type as the correct answers, and the final question is one that verifies the subject's accuracy.

[0163] The dementia diagnosis system using story-based voice extraction is characterized in that the controller 50 further includes an auxiliary dementia discrimination unit 45 that compares and analyzes the answers collected from the test subject to the questions with the correct answers to check the error rate, and determines the presence or absence of dementia and / or the state of dementia depending on the degree of the error rate.

[0164] The controller 50 is a dementia diagnosis system using story-based voice extraction, characterized in that it determines whether the test subject has dementia by performing a combined dementia determination based on the test subject's voice characteristics and the test subject's error rate.

[0165] A dementia diagnosis system using story-based voice extraction, wherein the actual story is a fairy tale or fable that the subject is already familiar with.

[0166] A dementia diagnosis system using story-based speech extraction, wherein the dramatized story consists of 5 to 15 sentences and contains 20 to 50 different noun-form words.

[0167] A dementia diagnosis system using story-based speech extraction, characterized in that the dramatized story includes at least one additional sentence that is artificially created and associated with and linked to the content of the dramatized story, different from the content of the actual story, and placed at the end of the dramatized story.

[0168] The dementia diagnosis system using story-based voice extraction is characterized in that the controller 50 further includes a memory disturbance unit 46 that disturbs the memory of the test subject by allowing a certain period of time to pass after the output of the dramatized story is completed and before requesting the test subject to speak aloud.

[0169] A dementia diagnosis system using story-based voice extraction, wherein the voice features are one or two of the fundamental frequency of the voice, speaking rate, shimmer, jitter, formant, and voice spectrum.

[0170] The dementia diagnostic system further includes a data input device 10 connected to the data storage device 20, controlled by the controller 50, and capable of inputting and storing various data including diagnostic voice data into the data storage device 20, and a data transmitter 80 connected to the data storage device 20, controlled by the controller 50, and transmitting the dementia information for each test subject stored in the data storage device 20 to an external device.

[0171] FIG. 13 is a flow chart showing a method for diagnosing dementia using story-based speech extraction according to the present invention.

[0172] Referring to FIG. 13, the dementia diagnosis method using story-based voice extraction according to the present invention includes a patient data preparation step S10 of extracting voice features of a dementia patient and preparing patient voice feature data, an actual story preparation step S20 of preparing an actual story, an dramatized story formation step S30 of dramatizing the actual story to prepare an dramatized story and converting it into voice data, an dramatized story output step S40 of outputting the voice data of the dramatized story to an external device, a speech request step S50 of requesting the test subject to speak about the content of the dramatized story, a voice collection step S60 of receiving and collecting voices spoken by the test subject, and a dementia diagnosis step S70 of comparing the collected voice of the test subject with the patient's voice feature data to determine whether or not dementia is present. These steps are performed sequentially to diagnose dementia.

[0173] The dementia diagnosis method may further include a question formation step S30a, following the dramatized story formation step S30, of determining the changed noun form words included in the dramatized story as correct answers, preparing a plurality of questions related to the dramatized story, and converting the plurality of questions into voice data. The speech request step S50 may further include a question output process of sequentially outputting voice data of the questions to the outside to ask the questions to the test subject after the output of the dramatized story to the test subject is completed.

[0174] In the question formation step S30a, the questions may consist of 3 to 12 questions requesting answers as to whether the events, actions, situations and contents presented in the dramatized story are factual.

[0175] The question may be a question that requests a sentence or word answer.

[0176] The questions may be multiple choice questions in which the incorrect answers include noun-form words contained in the actual story of the same or similar type as the correct answers, and the final question may be a question to check the subject's accuracy rate.

[0177] The dementia determination step S70 may further include an auxiliary dementia determination process for comparing and analyzing the answers collected from the test subject to the questions with the correct answers to check the error rate, and determining the presence or absence of dementia and / or the state of dementia depending on the degree of the error rate.

[0178] The dementia determination step S70 can determine whether the subject has dementia by performing a combination of dementia determination based on the voice characteristics of the subject and dementia determination based on the subject's error rate.

[0179] In the real story preparation step S20, the real story may be a fairy tale or fable that the test subject is already familiar with.

[0180] In the adaptation story formation step S30, the adaptation story may consist of 5 to 15 sentences and may contain 20 to 50 different noun-form words.

[0181] The dramatized story formation step S30 may further include a text addition process for adding at least one artificially created additional text that is different from the content of the actual story and is related to and linked to the content of the dramatized story, and the additional text may be placed at the end of the dramatized story, using story-based audio extraction.

[0182] The speaking request step S50 may further include a memory disturbance step of disturbing the test subject's memory by allowing a certain time to pass before requesting the test subject to speak aloud after the output of the dramatized story is completed.

[0183] In the patient data preparation step S10, the voice features may be one or more of voice fundamental frequency, speaking rate, shimmer, jitter, formant and voice spectrum.

[0184] Embodiments of the present invention may be realized in the form of a computer program executable on a computer via various components, and such a computer program may be recorded on a computer-readable medium, which may include hardware devices specifically configured to store and execute program instructions, such as magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, ROMs, RAMs, flash memories, etc.

[0185] The computer program may be one specially designed and constructed for the present invention, or one that is well known and available to those skilled in the art of computer software. Examples of the computer program include not only machine code such as that produced by a compiler, but also high-level language code that can be executed on a computer using an interpreter, etc.

[0186] According to one embodiment, methods according to various embodiments of the present disclosure may be provided in a computer program product. The computer program product may be traded as a commodity between a seller and a buyer. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., a compact disc read-only memory (CD-ROM)) or may be distributed online (e.g., downloaded or uploaded) via an application store (e.g., Play Store™) or directly between two user devices. In the case of online distribution, at least a portion of the computer program product may be at least temporarily stored or temporarily generated in a machine-readable storage medium, such as the memory of a manufacturer's server, an application store server, or an intermediary server.

[0187] Unless explicitly stated or stated to the contrary, steps constituting a method according to the present invention may be performed in any suitable order. The present invention is not necessarily limited to the order of steps described above. The use of all examples or exemplary terms (such as, for example, etc.) in the present invention is merely for the purpose of explaining the present invention in detail, and the scope of the present invention is not limited by such examples or exemplary terms unless otherwise limited by the claims. Furthermore, those skilled in the art will understand that various modifications, combinations, and variations can be made within the scope of the claims or their equivalents, depending on design conditions and factors.

[0188] On the other hand, although specific embodiments have been described in the detailed description of the present invention, it will be obvious to those skilled in the art that various modifications are possible without departing from the scope of the present invention.

Claims

1. An apparatus for performing a cognitive decline diagnosis method using story-based speech extraction, comprising: a controller for controlling the device; The controller generating an adapted story based on an actual story consisting of a plurality of sentences, and converting the generated adapted story into audio data; outputting audio data of the dramatized story to an external device and providing it to a test subject; providing a distracting stimulus to interfere with the test subject's memory, and then collecting speech from the test subject; Analyzing voice features from the collected voice of the test subject, and comparing the analyzed voice features of the test subject with patient voice feature data to determine the presence or absence of cognitive decline and / or the state of cognitive decline; The patient voice feature data is generated based on voice features of patients with cognitive impairment extracted by comparative analysis of voice features of normal individuals and voice features of patients with cognitive impairment.

2. The controller providing a distracting stimulus to interfere with the test subject's memory, and then providing a question related to the dramatized story; The device according to claim 1 , wherein speech uttered by the test subject is collected as a response to the question.

3. The controller The device of claim 1, wherein the dramatized story is generated by generating artificially created additional text that is different from the content of the actual story and that is related to and linked to the content of the dramatized story, and placing the additional text at the end of the dramatized story.

4. The controller The device described in claim 1, which compares and analyzes the answers collected from the test subject to the questions with the correct answers to determine the error rate, and determines the presence or absence of dementia and / or the state of dementia based on the degree of the error rate.

5. The controller The device according to claim 4, wherein dementia detection is performed in a combined manner based on the voice characteristics of the test subject and the error rate of the test subject, thereby determining whether the test subject has dementia.

6. The audio features include: The device of claim 1 , wherein the characteristics are any one or more of fundamental frequency of the voice, speaking rate, shimmer, jitter, formants, and voice spectrum.

7. 1. A computer-implemented method for diagnosing cognitive decline using story-based audio extraction, comprising: generating an adapted story based on an actual story consisting of a plurality of sentences, and converting the generated adapted story into audio data; outputting audio data of the dramatized story to an external device and providing it to a test subject; providing a distracting stimulus to interfere with the test subject's memory, and then collecting speech uttered by the test subject; Analyzing voice features from the collected voice of the test subject, and comparing the analyzed voice features of the test subject with patient voice feature data to determine the presence or absence of cognitive decline and / or the state of cognitive decline; A computer-implemented method, wherein the patient voice feature data is generated based on voice features of patients with cognitive impairment extracted by comparative analysis of voice features of normal individuals and voice features of patients with cognitive impairment.

8. A computer-readable recording medium having recorded thereon a program for executing the computer-implemented method according to claim 7.

Images