Method for classifying aphasia using linguistic analysis and recording medium having same recorded therein

Abstract

The present invention relates to a method for classifying aphasia using linguistic analysis. The method comprises the steps of: when voice information of a patient is inputted, extracting a linguistic component from the inputted voice information; performing linguistic analysis by using the extracted linguistic component; diagnosing aphasia and non-aphasia through the linguistic analysis; and when the aphasia is diagnosed, classifying the aphasia into a detailed aphasia type. According to the present invention, aphasia can be diagnosed and treated at an early stage by diagnosing and classifying the aphasia by using language information.

Description

Aphasia classification method using language analysis and recording medium recording the method

[0001] The present invention relates to a technique for classifying aphasia using language analysis.

[0002] Aphasia is a type of language disorder that occurs after language acquisition, resulting in difficulties in understanding or expressing language due to acquired brain damage. Therefore, to diagnose aphasia, it must be established that normal language acquisition occurred prior to the language disorder. Unlike speech disorders, aphasia requires confirmation of damage to the brain, the central nervous system. The most common causes of aphasia include cerebral infarction, cerebral hemorrhage, brain trauma, brain tumors, and herpes encephalitis. Aphasia can also be observed in degenerative dementias such as Alzheimer's disease and frontotemporal dementia.

[0003] Primary progressive aphasia (PPA) is caused by selective neurodegenerative changes in the language areas of the brain. Language impairment appears as the initial symptom and progresses gradually, while other brain functions are relatively preserved. There are three types of PPA: nonfluent variant PPA, semantic variant PPA, and logopenic variant PPA.

[0004] Nonfluent PPA is characterized by grammatical errors, labored and hesitant speech, speech apraxia, and reduced fluency. Furthermore, complex sentence structures may impair comprehension, but single-word concepts are relatively well-maintained and object recognition is preserved. Early aphasia testing in nonfluent PPA primarily reveals reduced fluency.

[0005] Semantic variant PPA preserves fluency and grammar, but impairs noun naming and single-word comprehension. Object recognition is also impaired, but repetition and independent speech are preserved.

[0006] In contrast, logopenic PPA is characterized by a slow speech rate and subsequent speech impairment, while fluency, grammar, and word comprehension are relatively preserved. In addition, single-word memory retrieval is impaired during spontaneous speech, leading to hesitation in word search and frequent occurrence of phonemic paraphrases. However, unlike semantic variant PPA, single-word comprehension and object recognition ability are preserved, and motor speech impairment and true grammatical errors are less frequent than in nonfluent variant PPA, so fluency is relatively preserved compared to nonfluent variant PPA.

[0007] The process of producing speech sounds occurs when air from the lungs through respiration passes through the vocal cords in the larynx, causing them to vibrate. This process is called phonation, and these sounds are amplified and resonate as they pass through the pharynx. The resonated sounds then pass through the oral and nasal cavities. When the movements of the soft palate, tongue, and lips—the organs of articulation—are altered, various consonants and vowels are created. When intonation is added to the continuous production of these consonant and vowel sounds, prosody is formed. Therefore, if any of the above speech subsystems (respiration, phonation, resonance, articulation, prosody) are disrupted, problems with pronunciation and articulation arise, which is called a speech disorder. Therefore, speech disorders can be caused not only by damage to the central nervous system related to speech movement, but also by damage to the peripheral nerves that control the organs where speech is produced (e.g., vocal cords, tongue).

[0008] Conventional language disorder analysis techniques involve recording a patient's voice, playing the file back, transcribing the audio into text, and then analyzing the resulting transcription. Consequently, utilizing these techniques presents a significant challenge: speech analysis is time-consuming and expensive.

[0009] The present invention has been devised to solve the above problems, and its purpose is to provide a method for simply and quickly diagnosing potential aphasia, classifying aphasia, and screening and monitoring the cause of aphasia using a patient's language information.

[0010] The purpose of the present invention is not limited to the purposes mentioned above, and other purposes not mentioned will be clearly understood by those skilled in the art from the description below.

[0011] To achieve the above purpose, the present invention relates to a method for classifying aphasia using language analysis, comprising: a step of extracting a linguistic component from the input voice information when a patient's voice information is input; a step of performing language analysis using the extracted linguistic component; a step of diagnosing aphasia and non-aphasia through the language analysis; and a step of classifying into a detailed aphasia type when aphasia is diagnosed.

[0012] By performing a given task on a patient, voice information can be input, and linguistic components can be extracted from the input voice information to perform language analysis.

[0013] Aphasia and non-aphasia can be diagnosed by performing language analysis by performing a predetermined task on the patient and comparing the extracted linguistic components from the patient's voice information with the normal distribution of linguistic components corresponding to the same age and education level as the patient.

[0014] According to the present invention, by diagnosing and classifying aphasia using language information, there is an effect of being able to diagnose and treat aphasia at an early stage.

[0015] In addition, according to the present invention, by diagnosing a specific cause of aphasia, there is an effect of preventing potential aphasia-related diseases and reducing the prevalence.

[0016] Figure 1 is a flowchart showing a method for classifying aphasia using language analysis according to one embodiment of the present invention.

[0017] Figure 2 is a flowchart showing a specific aphasia classification method using language analysis according to one embodiment of the present invention.

[0018] Figure 3 is a flowchart illustrating the classification process of aphasia in Figure 2.

[0019] Figures 4 to 10 show brain damage areas corresponding to aphasia classification according to one embodiment of the present invention.

[0020] Figure 11 is a graph illustrating a normal distribution.

[0021] Figure 12 is a diagram showing the standard deviation values ​​in Figure 11.

[0022] Figure 1 is a flowchart showing a method for classifying aphasia using language analysis according to one embodiment of the present invention.

[0023] Referring to Figure 1, when a patient's voice information is input (S110), a linguistic component is extracted from the voice information (S120). In one embodiment of the present invention, the linguistic component can be extracted using STT (Speech to Text) technology. That is, the voice information can be transcribed using STT technology to extract the linguistic component. In one embodiment of the present invention, a computer device including an audio collection device for inputting the patient's voice information may be provided. For example, the audio collection device may be implemented as a microphone.

[0024] Then, language analysis is performed using the extracted linguistic components (S130).

[0025] And, based on language analysis, aphasia is diagnosed (S140).

[0026] When diagnosed with aphasia, aphasia is classified into subtypes (S150).

[0027] In one embodiment of the present invention, a specific method is proposed for a process of extracting linguistic components from a patient's voice information, performing language analysis, and diagnosing aphasia based on the extracted linguistic components.

[0028] In one embodiment of the present invention, a method is proposed in which a patient is given four tasks (task 1, task 2, task 3, task 4) to input language information, linguistic components are extracted from the input language information to perform language analysis, and aphasia is diagnosed based on the results, and a specific type of aphasia is classified.

[0029] Task 1 is 'Name the objects (K-Boston naming test, short form)'.

[0030] In Task 1, analysis is performed by comparing the linguistic components extracted from the patient's voice information with the normal distribution of linguistic components corresponding to the same age and education level as the patient.

[0031] The present invention is susceptible to various modifications and embodiments. Specific embodiments are illustrated and described in detail in the drawings. However, this is not intended to limit the present invention to specific embodiments, but rather to encompass all modifications, equivalents, and alternatives falling within the spirit and technical scope of the present invention.

[0032] The terminology used in this application is only used to describe specific embodiments and is not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, it should be understood that the terms "comprise" or "have" indicate the presence of a feature, number, step, operation, component, part, or combination thereof described in the specification, but do not preclude the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0033] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Terms defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and will not be interpreted in an idealized or overly formal sense unless explicitly defined herein.

[0034] In addition, when describing with reference to the attached drawings, identical components will be assigned the same reference numerals regardless of the drawing numbers, and redundant descriptions thereof will be omitted. When describing the present invention, if a detailed description of a related known technology is judged to unnecessarily obscure the gist of the present invention, the detailed description will be omitted.

[0035] The subject performing the disease classification method using speech analysis of the present invention may be any computer device or server system performing the aphasia classification method using language analysis, or may be a control unit or processor that generally controls the system or device performing the aphasia classification method using language analysis. That is, the aphasia classification method using language analysis of the present invention may be composed of an algorithm, which is a type of software, and the software may be executed in the control unit or processor of the server system or device performing the aphasia classification method using language analysis.

[0036] The present invention relates to a method for classifying aphasia using language analysis, and comprises the steps of: when a patient's voice information is input, extracting a linguistic component from the input voice information; performing language analysis using the extracted linguistic component; diagnosing aphasia and non-aphasia through the language analysis; and, when aphasia is diagnosed, classifying the patient into a detailed aphasia type.

[0037] By performing a given task on a patient, voice information can be input, and linguistic components can be extracted from the input voice information to perform language analysis.

[0038] Aphasia and non-aphasia can be diagnosed by performing language analysis by performing a predetermined task on the patient and comparing the extracted linguistic components from the patient's voice information with the normal distribution of linguistic components corresponding to the same age and education level as the patient.

[0039] Figure 1 is a flowchart showing a method for classifying aphasia using language analysis according to one embodiment of the present invention.

[0040] Referring to Figure 1, when a patient's voice information is input (S110), a linguistic component is extracted from the voice information (S120). In one embodiment of the present invention, the linguistic component can be extracted using STT (Speech to Text) technology. That is, the voice information can be transcribed using STT technology to extract the linguistic component. In one embodiment of the present invention, a computer device including an audio collection device for inputting the patient's voice information may be provided. For example, the audio collection device may be implemented as a microphone.

[0041] Then, language analysis is performed using the extracted linguistic components (S130).

[0042] And, based on language analysis, aphasia is diagnosed (S140).

[0043] When diagnosed with aphasia, aphasia is classified into subtypes (S150).

[0044] In one embodiment of the present invention, a specific method is proposed for a process of extracting linguistic components from a patient's voice information, performing language analysis, and diagnosing aphasia based on the extracted linguistic components.

[0045] In one embodiment of the present invention, a method is proposed in which a patient is given four tasks (task 1, task 2, task 3, task 4) to input language information, linguistic components are extracted from the input language information to perform language analysis, and aphasia is diagnosed based on the results, and a specific type of aphasia is classified.

[0046] Task 1 is 'Name the objects (K-Boston naming test, short form)'.

[0047] In Task 1, analysis is performed by comparing the linguistic components extracted from the patient's voice information with the normal distribution of linguistic components corresponding to the same age and education level as the patient.

[0048] Figure 11 is a graph illustrating a normal distribution.

[0049] For example, in the case of normal distribution data as in Figure 11, the standard deviation (SD) is a value that indicates how far the normal distribution data is from the mean.

[0050] Figure 12 is a diagram showing the standard deviation values ​​in Figure 11.

[0051] Referring to Figure 12, it can be confirmed that the standard deviation, which is ±1SD of the mean, corresponds to ±1σ and includes 68.2%. In this case, the standard deviation is used only for data that is normally distributed.

[0052] Task 2 is "self-speaking." For example, ask, "Tell me about today's weather." If the patient can string together five or more words, it is considered a success. If the patient can string together fewer than five words, it is considered a failure.

[0053] Task 3 is ‘listening and understanding’.

[0054] In Task 3, a yes / no test (5 items) and a command compliance test (1 item) are administered.

[0055] The yes / no test (5 questions) is as follows:

[0056] (1) Is the door to this room closed now?

[0057] (2) Does a stone sink in water?

[0058] (3) Is the hammer used to cut wood?

[0059] (4) Do you put on socks after putting on your shoes?

[0060] (5) Are birds eaten by bugs?

[0061] The command follow-up test (1 question) is “Fold the paper in half, turn it over, and give it to me (the examiner).”

[0062] If the patient performs both the yes / no test (5 items) and the command compliance test (1 item), the test is considered successful; otherwise, it is considered unsuccessful.

[0063] In one embodiment of the present invention, a command compliance test can be performed by filming a patient's command compliance process through a camera in a computer device and analyzing the image to confirm whether the command has been complied with.

[0064] Task 4 is ‘Repeat’.

[0065] In Task 4, check whether the patient repeats 'Squirrel (1) / Returning migratory bird (2) / Sharp rock (3) / Korea High School Baseball League (4) / It's drizzling outside the window (5)'. If the patient repeats all of them, it is judged as success, otherwise it is judged as failure.

[0066] Figure 2 is a flowchart showing a specific aphasia classification method using language analysis according to one embodiment of the present invention.

[0067] Referring to Figure 2, as a result of performing task 1 (S201), if the result is less than -1SD compared to the normal distribution of the linguistic component corresponding to the same age and education level as the patient, aphasia is diagnosed (S203, S205), otherwise, non-aphasia is diagnosed (S207).

[0068] If aphasia is diagnosed (S205), tasks (task 2, task 3, task 4) are performed to classify specific aphasia types (S209-S251). The detailed process for classifying each aphasia type is described below.

[0069] First, if the patient fails to perform Task 2 (S209, S211), Task 3 (S213, S217), or Task 4 (S219, S229, S231), the patient is classified as having global aphasia. Global aphasia can be caused by brain lesions such as cerebral infarction.

[0070] If you fail to perform Task 2 (S209, S211), fail to perform Task 3 (S213, S217), and succeed in Task 4, you are classified as having mixed transcortical aphasia (S219, S229, S233). Mixed transcortical aphasia can be caused by brain lesions such as cerebral infarction.

[0071] If you fail to perform Task 2 (S209, S211), succeed in Task 3 (S213, S217), and fail to perform Task 4 (S221, S235, S237), you are classified as having Broca's aphasia. Broca's aphasia can be caused by brain lesions such as cerebral infarction, and non-fluent aphasia among degenerative dementias.

[0072] If Task 2 is not performed successfully (S209, S211), Task 3 is performed successfully (S213, S217), and Task 4 is performed successfully, it is classified as transcortical motor aphasia (S221, S235, S239). Transcortical motor aphasia can be caused by brain lesions such as cerebral infarction.

[0073] If Task 2 is successfully performed (S209, S211), Task 3 is not successfully performed (S215, S223), and Task 4 is not successfully performed (S225, S241, S243), it is classified as Wernicke's aphasia. Wernicke's aphasia can be caused by brain lesions such as cerebral infarction or by semantic dementia among degenerative dementias.

[0074] If Task 2 is successfully performed (S209, S211), Task 3 is not successfully performed (S215, S223), and Task 4 is successfully performed, it is classified as transcortical sensory aphasia (S225, S241, S245). Transcortical sensory aphasia can be caused by brain lesions such as cerebral infarction.

[0075] If Task 2 is successfully performed (S209, S211), Task 3 is successfully performed (S215, S223), and Task 4 is not successfully performed, it is classified as conduction aphasia (S227, S247, S249). Conduction aphasia can be caused by brain lesions such as cerebral infarction or by logopenic type Alzheimer's dementia among degenerative dementias.

[0076] If the patient successfully performs Task 2 (S209, S211), Task 3 (S215, S223), and Task 4 (S227, S247, S251), it is classified as naming aphasia. Naming aphasia is a type in which the patient is relatively fluent and can understand and repeat words relatively well, but the difficulty in naming is prominent, and mild decline in ability in all areas is often observed.

[0077] Figures 4 to 10 show brain damage areas corresponding to aphasia classification according to one embodiment of the present invention.

[0078] Figure 4 shows the brain damage areas corresponding to global aphasia. Global aphasia is a case in which severe impairment of language ability is observed across all areas, and is caused by brain damage in a wide area around the left Sylvian fissure.

[0079] Figure 5 shows the brain damage areas corresponding to mixed transcortical aphasia. Mixed transcortical aphasia is a case in which, similar to global aphasia, both the understanding and expression of language are severely impaired, but only the ability to repeat words is relatively preserved.

[0080] Figure 6 shows the brain damage areas corresponding to Broca's aphasia. Broca's aphasia is characterized by a decrease in fluency and the ability to repeat and name words, but relatively preserved ability to listen and understand. It is usually caused by damage to the lower left frontal lobe.

[0081] Figure 7 shows the brain damage areas corresponding to transcortical motor aphasia. Transcortical motor aphasia has similar characteristics to Broca's aphasia, but the ability to repeat words is relatively preserved. It is a disease caused by damage to the left frontal lobe or subcortical area.

[0082] Figure 8 shows the brain damage areas corresponding to Wernicke's aphasia. Wernicke's aphasia usually occurs due to damage to the upper part of the left temporal lobe and can produce speech.

[0083] Figure 9 shows the brain damage areas corresponding to transcortical sensory aphasia. Transcortical sensory aphasia has many similar characteristics to Wernicke's aphasia, but the ability to repeat speech is relatively preserved, and it is often caused by damage around Wernicke's area or subcortical areas.

[0084] Figure 10 shows the brain damage areas that correspond to conduction aphasia. Conduction aphasia is a condition in which the patient is relatively fluent and has relatively good listening and understanding skills, but is impaired in the area of ​​following others' speech. It occurs when there is damage to the connecting fibers called the arcuate bundle, which connects the Broca and Wernicke areas of the left brain, which are responsible for language production and comprehension.

[0085] In one embodiment of the present invention, while classifying the type of aphasia, the brain damage area of ​​the corresponding aphasia type, as shown in FIGS. 4 to 10, can be displayed on the screen together.

[0086] If diagnosed with aphasia (S207), tasks (task 5, task 6, task 7, task 8) can be performed to diagnose apraxia of speech.

[0087] Task 5 examines whether there is a difference in performance between automatic speech (such as reciting numbers or days of the week) and intentional speech.

[0088] Task 6 involved repeating appended words (e.g., wedding-wedding-wedding hall-wedding location) to determine whether performance deteriorates as the words become longer.

[0089] Task 7 involves checking whether the repetition of the same word (e.g., wedding-wedding-wedding) produces errors that involve inconsistent errors or exploratory behavior.

[0090] Task 8 involves assessing whether the patient has orofacial motor impairment, for example, when performing tasks such as tongue protrusion, lip pursing, mimicking whistling, mimicking blowing out a candle, or coughing.

[0091] Figure 3 is a flowchart illustrating the classification process of aphasia in Figure 2.

[0092] Referring to Figure 3, if all of Task 5, Task 6, Task 7, and Task 8 are met, it is diagnosed as apraxia of speech (S301 to S317), otherwise it is diagnosed as other, including stuttering, repetitive speech, echolalia, and mutism (S319).

[0093] Apraxia of speech (AOS) refers to a disorder in programming the sensorimotor commands for muscle coordination and movement required for voluntary speech production. Apraxia of speech occurs without muscle paralysis or language or cognitive impairment. It is most often caused by stroke damage to the left cerebral hemisphere, often accompanying nonfluent aphasia. It is characterized by inappropriate pauses or groping during speech production and frequent articulation errors. Unlike dysarthria, apraxia of speech is characterized by inconsistent speech, primarily consisting of inversions or substitutions within words or phrases. Repeated attempts to recognize and correct these inconsistencies result in different errors each time. Performance may improve in automatic speech, such as counting or stating the days of the week. Complex consonant sequences are more difficult to produce, and errors are more common when coding targets.

[0094]

[0095] Meanwhile, the method for classifying aphasia using language analysis according to an embodiment of the present invention can be implemented as computer-readable code on a computer-readable recording medium. A computer-readable recording medium includes any type of recording device that stores data that can be read by a computer system.

[0096] For example, computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, hard disk, floppy disk, removable storage device, non-volatile memory (flash memory), and optical data storage device.

[0097] Additionally, the computer-readable recording medium can be distributed across computer systems connected to a computer network, and stored and executed as readable code in a distributed manner.

[0098] While the present invention has been described using several preferred embodiments, these embodiments are illustrative and not limiting. Those skilled in the art will appreciate that various changes and modifications can be made without departing from the spirit and scope of the invention as set forth in the appended claims.

Claims

1. When a patient's voice information is input, a step of extracting a linguistic component from the input voice information; A step of performing language analysis using the extracted linguistic components; A step of diagnosing aphasia and non-aphasia through the above language analysis; and When diagnosed with aphasia, the next step is to classify it into specific aphasia types. A method for classifying aphasia including .

2. In claim 1, A method for classifying aphasia, characterized by performing a predetermined task on a patient to input speech information, extracting linguistic components from the input speech information, and performing language analysis.

3. In claim 2, A method for classifying aphasia, characterized in that it diagnoses aphasia and non-aphasia by performing language analysis by comparing linguistic components extracted from the patient's voice information by performing a predetermined task on the patient with a normal distribution of linguistic components corresponding to the same age and education level as the patient.

4. A computer-readable recording medium having recorded thereon a program capable of executing the method of any one of claims 1 to 3 by a computer.

Images