Apparatus and method for diagnosing parkinson's disease

Abstract

The present invention relates to an apparatus and a method for diagnosing Parkinson's disease. The method for diagnosing Parkinson's disease according to the present embodiment may comprise: a step for presenting one or more athletic ability evaluation missions and one or more non-athletic ability evaluation missions to a user, and collecting, from the user, performance results for the one or more athletic ability evaluation missions and performance results for the one or more non-athletic ability evaluation missions; a step for extracting one or more athletic ability features and one or more non-athletic ability features from the performance results for the one or more athletic ability evaluation missions and the performance results for the one or more non-athletic ability evaluation missions; and a step for diagnosing the presence or absence of Parkinson's disease in the user on the basis of at least one of the results of extracting the one or more athletic ability features and the one or more non-athletic ability features.

Description

Device and method for diagnosing Parkinson's disease

[0001] The present invention relates to a device and method for diagnosing Parkinson's disease.

[0002] Parkinson's disease is a degenerative neurological disorder caused by the loss of dopamine-producing neurons, which can lead to motor dysfunction. Currently, diagnosis of Parkinson's disease relies primarily on clinical symptoms and medical observations. However, this relies heavily on subjective judgment and makes it difficult to detect the disease in its early stages.

[0003] The background technology described above is technical information that the inventor possessed for the purpose of deriving the present invention or acquired in the process of deriving the present invention, and cannot necessarily be considered as publicly known technology disclosed to the general public prior to the application for the present invention.

[0004] Prior art document: Korean Patent Publication No. 10-2019-0033802 (April 1, 2019)

[0005] One object of the present invention is to provide an objective and reliable device and method for early diagnosis and progression monitoring of Parkinson's disease.

[0006] One object of the present invention is to provide a device and method capable of accurately tracking the early detection and progression of Parkinson's disease through real-time data analysis.

[0007] One object of the present invention is to accurately diagnose Parkinson's disease through analysis of a subject's gestures and audio.

[0008] The problems addressed by the present invention are not limited to those mentioned above. Other problems and advantages of the present invention not mentioned above can be understood through the following description and will be more clearly understood through embodiments of the present invention. Furthermore, it will be appreciated that the problems and advantages addressed by the present invention can be realized by the means and combinations thereof set forth in the claims.

[0009] A method for diagnosing Parkinson's disease according to the present embodiment is a method for diagnosing Parkinson's disease performed by a processor of a Parkinson's disease diagnosis device, and may include a step of presenting one or more motor ability evaluation missions and one or more non-motor ability evaluation missions to a user, collecting a performance result for one or more motor ability evaluation missions and a performance result for one or more non-motor ability evaluation missions from the user, a step of extracting one or more motor ability features and one or more non-motor ability features from the performance result for one or more motor ability evaluation missions and the performance result for one or more non-motor ability evaluation missions, and a step of diagnosing whether the user has Parkinson's disease based on at least one of the results of extracting one or more motor ability features and one or more non-motor ability features.

[0010] A Parkinson's disease diagnosis device according to the present embodiment includes a processor and a memory operably connected to the processor and storing at least one code executed by the processor, and the memory can store code that, when executed through the processor, causes the processor to present one or more motor ability evaluation missions and one or more non-motor ability evaluation missions to a user, collect performance results for one or more motor ability evaluation missions and performance results for one or more non-motor ability evaluation missions from the user, extract one or more motor ability features and one or more non-motor ability features from the performance results for one or more motor ability evaluation missions and the performance results for one or more non-motor ability evaluation missions, and diagnose the presence or absence of Parkinson's disease in the user based on at least one of the results of extracting the one or more motor ability features and the one or more non-motor ability features.

[0011] In addition, other methods for implementing the present invention, other systems, and computer-readable recording media storing a computer program for executing the method may be further provided.

[0012] Other aspects, features and advantages other than those described above will become apparent from the following drawings, claims and detailed description of the invention.

[0013] According to the present invention, early diagnosis of Parkinson's disease can be made, thereby helping to initiate treatment at an appropriate time.

[0014] Additionally, simple, non-invasive tests that can be performed during the user's daily activities can improve the accuracy of diagnosis and monitoring of Parkinson's disease by precisely analyzing subtle changes in motor function.

[0015] Additionally, through continuous user data collection and analysis, the progression of Parkinson's disease can be monitored, providing improved information for disease management and treatment planning.

[0016] The effects of the present invention are not limited to those mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description below.

[0017] Figure 1 is an example diagram of a Parkinson's disease diagnosis environment according to the present embodiment.

[0018] FIG. 2 is a block diagram schematically illustrating the configuration of a Parkinson's disease diagnosis device according to the present embodiment.

[0019] Figure 3 is a block diagram schematically illustrating the configuration of the diagnosis management unit of the Parkinson's disease diagnosis device of Figure 2.

[0020] FIGS. 4A to 4I are block diagrams schematically illustrating the configuration of a diagnostic unit among the diagnostic management units according to the present embodiment.

[0021] FIGS. 5A to 5H are exemplary diagrams of a Parkinson's disease diagnosis screen provided to a user terminal (200) according to the present embodiment.

[0022] FIG. 6 is a block diagram schematically illustrating the configuration of a Parkinson's disease diagnosis device according to another embodiment.

[0023] Figure 7 is a flowchart for explaining a method for diagnosing Parkinson's disease according to the present embodiment.

[0024] The advantages and features of the present invention, and the methods for achieving them, will become clearer with reference to the embodiments described in detail with the accompanying drawings. However, the present invention is not limited to the embodiments presented below, but can be implemented in various different forms, and it should be understood that all modifications, equivalents, and substitutes included within the spirit and technical scope of the present invention are included. The embodiments presented below are provided so that the disclosure of the present invention is complete and so that those skilled in the art to which the present invention pertains will fully understand the scope of the invention. Describing the Invention The advantages and features of the present invention, and the methods for achieving them will become clearer with reference to the embodiments described in detail with the accompanying drawings. However, the present invention is not limited to the embodiments presented below, but can be implemented in various different forms, and it should be understood that all modifications, equivalents, and substitutes included within the spirit and technical scope of the present invention are included. The embodiments presented below are provided so that the disclosure of the present invention is complete and so that those skilled in the art to which the present invention pertains will fully understand the scope of the invention. In describing the present invention, if it is determined that a detailed description of a related known technology may obscure the gist of the present invention, the detailed description is omitted.

[0025] 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" are intended to 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. Terms such as first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used solely for the purpose of distinguishing one component from another.

[0026] Additionally, in the present application, a “part” may be a hardware component such as a processor or circuit, and / or a software component executed by a hardware component such as a processor.

[0027] Hereinafter, embodiments according to the present invention will be described in detail with reference to the attached drawings. In describing with reference to the attached drawings, identical or corresponding components are assigned the same drawing numbers, and redundant descriptions thereof will be omitted.

[0028] In the following examples, the terms first, second, etc. are not used in a limiting sense, but are used for the purpose of distinguishing one component from another.

[0029] In the examples below, singular expressions include plural expressions unless the context clearly indicates otherwise.

[0030] In the following examples, terms such as “include” or “have” mean that a feature or component described in the specification is present, and do not preclude the possibility that one or more other features or components may be added.

[0031] In some embodiments, where the implementation is otherwise feasible, a particular process sequence may be performed in a different order than described. For example, two processes described in succession may be performed substantially simultaneously, or in a reverse order from the described order.

[0032] FIG. 1 is an exemplary diagram of a Parkinson's disease diagnosis environment according to the present embodiment. Referring to FIG. 1, the Parkinson's disease diagnosis environment (1) may include a Parkinson's disease diagnosis device (100), a user terminal (200), and a network (300). In the following description, the expression in which the Parkinson's disease diagnosis device (100) transmits and receives signals with the user terminal (200) may be identical to the expression in which the Parkinson's disease diagnosis device (100) transmits and receives signals with the user.

[0033] A Parkinson's disease diagnosis device (100) can present one or more motor ability evaluation missions to a user and collect performance results for one or more motor ability evaluation missions from the user.

[0034] In this embodiment, the athletic ability assessment mission may include at least one of the first athletic ability assessment mission and the second athletic ability assessment mission. The first athletic ability assessment mission may include at least one of the first to fifth missions, and the second athletic ability assessment mission may include the sixth mission.

[0035] The first mission may include an action of flexing and extending a first object presented on the screen using a finger for a preset period of time (e.g., 20 seconds). In this embodiment, the first object may include a balloon. In this embodiment, the first object is not limited to a balloon, and may be any graphic element whose size is adjusted by touch in response to the flexion and extension of the finger within the screen. In this embodiment, the first mission may be named a balloon.

[0036] The second mission may include an action of repeatedly moving a second object displayed on the screen from the top to the bottom of the screen at maximum speed while touching it for a preset period of time. In this embodiment, the second object may include an image of a graphic element (e.g., a circle) that can be moved within the screen while being touched. In this embodiment, the second mission may be referred to as a "swipe."

[0037] The third mission may include an action of alternately touching two third objects presented on the screen at full speed using one finger for a preset amount of time. In this embodiment, the third object may include an image of a touchable graphic element (e.g., a circle) within the screen. In this embodiment, the third mission may be referred to as "touch."

[0038] The fourth mission may include an action of tracing a first-shaped guide pattern presented on the screen using a finger. In this embodiment, the first-shaped guide pattern may include a spiral pattern. In this embodiment, the first-shaped guide pattern is not limited to a spiral pattern, and any graphic pattern capable of performing touch tracing within the screen may be used. In this embodiment, the fourth mission may be named "drawing a spiral."

[0039] The fifth mission may include an action of tracing a second shape guide pattern, different from the first shape presented on the screen, using a finger. In this embodiment, the second shape guide pattern may include a wave pattern. In this embodiment, the second shape guide pattern is not limited to a wave pattern, and any graphic pattern capable of performing touch tracing within the screen may be used. In this embodiment, the fifth mission may be named "drawing a wave."

[0040] The sixth mission may include an action of listening to a guided speech voice corresponding to a script presented on the screen and speaking accordingly. In the present embodiment, the script may include a first script, a third script, and the like. The first script may include a basic set of vowels, for example, "aeiou." The second script may include syllables formed by combining specific consonants and specific vowels, for example, "pataka." The third script may include syllables formed by combining specific consonants and specific vowels, different from the second script, for example, "ga-na-da." In the present embodiment, the sixth mission may be named "speech."

[0041] In this embodiment, the sixth mission may include missions 6-1 to 6-5 based on the first to third scripts described above. A detailed description of missions 6-1 to 6-5 will be provided below.

[0042] From the above-described contents, the Parkinson's disease diagnosis device (100) can present one or more of the first to sixth missions to the user and collect the performance results of one or more of the first to sixth missions from the user.

[0043] A Parkinson's disease diagnosis device (100) can present one or more non-motor ability evaluation missions to a user and collect performance results for one or more non-motor ability evaluation missions from the user.

[0044] In this embodiment, the non-motor ability assessment mission may include at least one of the first non-motor ability assessment mission and the second non-motor ability assessment mission. The first non-motor ability assessment mission may include the seventh mission, and the second non-motor ability assessment mission may include the eighth mission.

[0045] The seventh mission may include an action that presents a questionnaire on the screen to measure the user's non-motor abilities, including emotional responses, stress levels, symptoms of depression, and other emotional states, and requires the user to complete the questionnaire. In this embodiment, the seventh mission may be referred to as a "survey."

[0046] The eighth mission may include an action of presenting a first area on the screen with guide colors, a color selection palette in which a user can select an arbitrary color by moving a fourth object, a second area in which a color selected by moving the fourth object on the color selection palette is displayed, and moving the fourth object on the color selection palette so that a color matching the first area is displayed in the second area. In the present embodiment, the eighth mission may be named color matching.

[0047] From the above-described contents, the Parkinson's disease diagnosis device (100) can present one or more of the seventh mission and the eighth mission to the user, and collect the performance results of the seventh mission and one or more of the eighth mission from the user.

[0048] The Parkinson's disease diagnosis device (100) can extract one or more motor ability features and one or more non-motor ability features from the performance results of one or more motor ability evaluation missions and the performance results of one or more non-motor ability evaluation missions.

[0049] The Parkinson's disease diagnosis device (100) may perform preprocessing to remove unnecessary data from the performance results of one or more motor ability evaluation missions and one or more non-motor ability evaluation missions before extracting one or more motor ability features and one or more non-motor ability features. That is, the Parkinson's disease diagnosis device (100) may extract one or more motor ability features and one or more non-motor ability features based on the preprocessing results.

[0050] In the present embodiment, the motor skill feature may include one or more of the first motor skill feature and the second motor skill feature. The first motor skill feature may include one or more of the first to fifth features, and the second motor skill feature may include the sixth feature.

[0051] The first feature may include statistical analysis indices and frequency domain analysis indices for the first time series data collected from the results of performing the first mission, and the number of times and speed at which the first object is bent and extended using the fingers.

[0052] The second feature may include statistical analysis indicators and frequency domain analysis indicators for the second time series data collected from the results of performing the second mission, and the number of times the second object is repeatedly moved from the top to the bottom while being touched.

[0053] The third feature may include statistical analysis indicators and frequency domain analysis indicators for third time series data collected from the results of performing the third mission, and the number of times and touch speed of two third objects being alternately touched with one finger.

[0054] The fourth feature may include statistical analysis indices and frequency domain analysis indices for the fourth time series data collected from the results of the fourth mission, and a first curvature calculated in response to the results of tracing the first shape guide pattern using a finger.

[0055] The fifth feature may include statistical analysis indices and frequency domain analysis indices for fifth time series data collected from the results of performing the fifth mission, and a second curvature calculated in response to the results of tracing the second shape guide pattern using a finger.

[0056] The sixth feature may include the 6-1 feature and the 6-2 feature. The 6-1 feature may include one or more of the fundamental frequency, shimmer, harmonic to noise ratio (HNR), loudness, and speech rate of the audio signal included in the speech result collected from the execution result of the 6th mission. The 6-2 feature may include the classification result obtained by applying an artificial intelligence model (e.g., Wav2vec model) to the audio signal included in the speech result collected from the preprocessed execution result of the 6th mission.

[0057] In the present embodiment, the non-motor ability feature may include one or more of the first non-motor ability feature and the second non-motor ability feature. The first non-motor ability feature may include the seventh feature, and the second non-motor ability feature may include the eighth feature.

[0058] The seventh feature may include responses to a questionnaire collected from the results of the seventh mission.

[0059] The eighth feature may include a color error including the difference between the guide color presented in the first area and the color displayed in the second area calculated based on data collected from the performance results of the eighth mission, and a root mean square error (RMSE) calculated based on the color error.

[0060] The Parkinson's disease diagnosis device (100) can diagnose whether a user has Parkinson's disease based on at least one of the results of extracting one or more motor ability features and one or more non-motor ability features. In the present embodiment, the Parkinson's disease diagnosis device (100) can utilize an artificial intelligence model to diagnose whether a user has Parkinson's disease.

[0061] Artificial intelligence (AI) is a field of computer engineering and information technology that studies methods to enable computers to perform human-like tasks such as thinking, learning, and self-improvement. It means enabling computers to imitate human intelligent behavior.

[0062] Furthermore, artificial intelligence does not exist in isolation; rather, it is closely intertwined with other fields of computer science, both directly and indirectly. In particular, in modern times, there are active attempts to introduce AI elements into various fields of information technology and utilize them to solve problems in those fields.

[0063] Machine learning is a branch of artificial intelligence that develops the ability for computers to learn from data without explicit programming. Machine learning can be broadly categorized into supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning. Representative machine learning models include decision trees, support vector machines (SVMs), and Bayesian networks, which can be utilized for tasks such as classification, regression analysis, and data mining, respectively. Machine learning models receive input data, utilize statistical learning methods to learn data patterns, and then make predictions or decisions based on these patterns.

[0064] Deep learning is a branch of machine learning that utilizes artificial neural networks. It can learn and model complex data features through multi-layered neural networks (perceptrons). Deep learning is an artificial neural network that can include multiple hidden layers, consisting of an input layer, one or more hidden layers, and an output layer. Deep learning models are particularly useful in fields such as image recognition, natural language processing, and speech recognition due to their ability to recognize high-dimensional data patterns in large datasets. Artificial neural networks receive input data, pass it through each layer, learn increasingly abstract features, and ultimately provide a prediction result at the output layer.

[0065] In this embodiment, the Parkinson's disease diagnosis device (100) may exist independently in the form of a server, or the Parkinson's disease diagnosis function provided by the Parkinson's disease diagnosis device (100) may be implemented in the form of an application and loaded onto a user terminal (200).

[0066] The user terminal (200) can access the Parkinson's disease diagnosis application and / or Parkinson's disease diagnosis site provided by the Parkinson's disease diagnosis device (100) to receive Parkinson's disease diagnosis services.

[0067] The user terminal (200) may include a communication terminal capable of performing the functions of a computing device (not shown), and may be, but is not limited to, a tablet PC, a smart TV, a mobile phone, a PDA (personal digital assistant), a media player, a micro server, a GPS (global positioning system) device, an e-book reader, a digital broadcasting terminal, a navigation device, a kiosk, an MP3 player, a digital camera, home appliances, and other mobile or non-mobile computing devices, in addition to a desktop computer (201), a smart phone (202), and a laptop (203) operated by the user. In addition, the user terminal (200) may be a wearable terminal such as a watch, glasses, a hair band, and a ring, having a communication function and a data processing function. The user terminal (200) is not limited to the above-described contents, and a terminal capable of web browsing may be borrowed without limitation.

[0068] The network (300) may play a role in connecting the Parkinson's disease diagnosis device (100) and the user terminal (200). This network (300) may include wired networks such as a local area network (LAN), a wide area network (WAN), a metropolitan area network (MAN), and an integrated service digital network (ISDN), or wireless networks such as a wireless LAN (WLAN), code-division multiple access (CDMA), and satellite communication, but the scope of the present invention is not limited thereto. In addition, the network (300) may transmit and receive information using short-range communication and / or long-range communication. Here, short-range communication may include Bluetooth, RFID (radio frequency identification), IrDA (infrared data association), UWB (ultra-wideband), ZigBee, and Wi-Fi technologies, and long-range communication may include CDMA (code-division multiple access), FDMA (frequency-division multiple access), TDMA (time-division multiple access), OFDMA (orthogonal frequency-division multiple access), and SC-FDMA (single carrier frequency-division multiple access) technologies.

[0069] The network (300) may include a connection of network elements such as hubs, bridges, routers, and switches. The network (300) may include one or more connected networks, such as a multi-network environment, including a public network such as the Internet and a private network such as a secure corporate private network. Access to the network (300) may be provided via one or more wired or wireless access networks.

[0070] Furthermore, the network (300) can support CAN (controller area network) communication, V2I (vehicle to infrastructure) communication, V2X (vehicle to everything) communication, WAVE (wireless access in vehicular environment) communication technology, and IoT (Internet of Things) network and / or 5G communication that exchange and process information between distributed components such as objects.

[0071] FIG. 2 is a block diagram schematically illustrating the configuration of a Parkinson's disease diagnosis device according to the present embodiment. In the following description, any part that overlaps with the description of FIG. 1 will be omitted. Referring to FIG. 2, the Parkinson's disease diagnosis device (100) may include a communication unit (110), a storage medium (120), a program storage unit (130), a database (140), a diagnosis management unit (150), and a control unit (160).

[0072] The communication unit (110) may provide a communication interface necessary to provide transmission and reception signals in the form of packet data between the Parkinson's disease diagnosis device (100) and the user terminal (200) in conjunction with the network (300). Furthermore, the communication unit (110) may serve to receive a predetermined information request signal from the user terminal (200) and may serve to transmit information processed by the diagnosis management unit (150) to the user terminal (200). Here, the communication interface refers to a medium that serves to connect the Parkinson's disease diagnosis device (100) and the user terminal (200), and may include a path that provides a connection path so that the user terminal (200) can transmit and receive information after connecting to the Parkinson's disease diagnosis device (100). In addition, the communication unit (110) may be a device including hardware and software necessary to transmit and receive signals such as control signals or data signals through a wired or wireless connection with another network device.

[0073] The storage medium (120) performs the function of temporarily or permanently storing data processed by the control unit (160). Here, the storage medium (120) may include a magnetic storage medium or a flash storage medium, but the scope of the present invention is not limited thereto. The storage medium (120) may include a built-in memory and / or an external memory, and may include a volatile memory such as a DRAM, an SRAM, or an SDRAM, a non-volatile memory such as an OTPROM (one time programmable ROM), a PROM, an EPROM, an EEPROM, a mask ROM, a flash ROM, a NAND flash memory, or a NOR flash memory, a flash drive such as an SSD, a CF (compact flash) card, an SD card, a Micro-SD card, a Mini-SD card, an Xd card, or a memory stick, or a storage device such as an HDD.

[0074] The program storage unit (130) is equipped with control software that performs tasks such as presenting a motor ability evaluation mission to a user and collecting the results of performing the motor ability evaluation mission from the user, presenting a non-motor ability evaluation mission to the user and collecting the results of performing the non-motor ability evaluation mission from the user, collecting and preprocessing the results of performing the motor ability evaluation mission and the non-motor ability evaluation mission, extracting motor ability features and non-motor ability features from the results of performing the motor ability evaluation mission and the non-motor ability evaluation mission, and diagnosing the presence or absence of the user's Parkinson's disease from the results of extracting the motor ability features and non-motor ability features.

[0075] The database (140) may include a management database storing various information for diagnosing Parkinson's disease. The management database may store the first to eighth missions presented to the user. The management database may store various algorithms for performing preprocessing on the performance results of the first to eighth missions, such as missing value processing, outlier detection and removal, normalization and standardization, data encoding, and feature scaling. The management database may store various algorithms for extracting features from the preprocessed performance results of the first to eighth missions, such as Fourier transform, wavelet transform, edge detection, corner detection, histogram analysis, principal component analysis, and artificial intelligence models. The management database may store an artificial intelligence model for diagnosing the presence or absence of Parkinson's disease in the user based on the results of feature extraction.

[0076] In addition, the database (140) may include a user database that stores information about users who will receive Parkinson's disease diagnosis services. Here, the user information may include basic information about the user, such as the user's name, affiliation, personal information, gender, age, contact information, email address, address, and image, as well as information about the user's authentication (login) such as the user ID (or email) and password, information about the country of access, access location, information about the device used for access, and information related to access such as the connected network environment.

[0077] Additionally, the user database may store user information, information and / or category history provided by users accessing the Parkinson's disease diagnosis application or Parkinson's disease diagnosis site, user-set environment settings, user-used resource usage information, and billing and payment information corresponding to the user's resource usage.

[0078] The diagnosis management unit (150) can present one or more motor ability evaluation missions to the user and collect the performance results of one or more motor ability evaluation missions from the user. The diagnosis management unit (150) can present one or more non-motor ability evaluation missions to the user and collect the performance results of one or more non-motor ability evaluation missions from the user. The diagnosis management unit (150) can collect and preprocess the performance results of one or more motor ability evaluation missions and the performance results of one or more non-motor ability evaluation missions. The diagnosis management unit (150) can extract one or more motor ability features and one or more non-motor ability features from the performance results of one or more motor ability evaluation missions and the performance results of one or more non-motor ability evaluation missions. The diagnosis management unit (150) can diagnose the presence or absence of Parkinson's disease in the user based on at least one of the results of extracting one or more motor ability features and one or more non-motor ability features.

[0079] The control unit (160) is a type of central processing unit and can control the operation of the entire Parkinson's disease diagnosis device (100) by driving the control software installed in the program storage unit (130). The control unit (160) may include all types of devices capable of processing data, such as a processor. Here, the 'processor' may refer to a data processing device built into hardware, which has a physically structured circuit to perform a function expressed by a code or instruction included in a program, for example. As an example of a data processing device built into hardware, it may include processing devices such as a microprocessor, a central processing unit (CPU), a processor core, a multiprocessor, an application-specific integrated circuit (ASIC), and a field programmable gate array (FPGA), but the scope of the present invention is not limited thereto.

[0080] FIG. 3 is a block diagram schematically illustrating the configuration of the diagnosis management unit of the Parkinson's disease diagnosis device of FIG. 2, and FIGS. 4a to 4i are block diagrams schematically illustrating the configuration of the diagnosis unit of the diagnosis management unit according to the present embodiment. In the following description, any part that overlaps with the description of FIGS. 1 and 2 will be omitted. Referring to FIGS. 3 to 4i, the diagnosis management unit (150) may include a collection unit (151), a preprocessing unit (152), an extraction unit (153), and a diagnosis unit (154).

[0081] The collection unit (151) can collect information about a user using the Parkinson's disease diagnosis service. In this embodiment, the information about the user may include an identification ID, whether or not there is Parkinson's disease, gender, age, academic years, and information about the hand that is primarily used between the two hands.

[0082] The collection unit (151) may present one or more athletic ability evaluation missions and one or more non-athletic ability evaluation missions to the user, and collect the performance results of one or more athletic ability evaluation missions and the performance results of one or more non-athletic ability evaluation missions from the user. In the present embodiment, the performance results of the athletic ability evaluation missions and the performance results of the non-athletic ability evaluation missions collected by the collection unit (151) may include raw data.

[0083] The collection unit (151) may present a first mission to the user and collect the result of performing the first mission from the user. In the present embodiment, the first raw data collected as the result of performing the first mission may include the time (timestamp) when the first mission starts and the data is collected. In addition, the first raw data may include a change rate (scale) of the distance between two fingers when the user manipulates the first object on the screen. Here, the change rate of the distance between the two fingers may be measured based on the initial distance between the two fingers set at the time of the initial manipulation of the first object. That is, the distance between the two fingers set when the user first touches and begins manipulating the first object may be set as the reference distance. Thereafter, the distance changing as the fingers move may be measured relative to the reference distance. In addition, the first raw data may include the x-axis and y-axis center coordinates (focal x, y) of the location where the two fingers touched the first object. In addition, the first raw data may include the intensity (force) of the force touching the first object. Additionally, the first raw data may include a count of touches of the first object on the screen (number of pointers).

[0084] The collection unit (151) may present a second mission to the user and collect the result of performing the second mission from the user. In the present embodiment, the second raw data collected as the result of performing the second mission may include the time (timestamp) when the second mission starts and the data is collected. In addition, the second raw data may include absolute coordinates (absolute x, y) of the x-axis and y-axis indicating the position on the screen where the second object is touched. The second raw data may include the change in the x-axis and y-axis from the position where the touch of the second object starts to the position where the touch ends (translation x, y). In addition, the second raw data may include the velocity (velocity x, y) at which the second object touched on the screen moves in the x-axis and y-axis directions. In addition, the second raw data may include relative x-coordinates and y-coordinates (X, Y) with respect to the touch point of the second object. Here, the relative x-coordinates and y-coordinates may be measured based on the touch start point of the second object. Additionally, the second raw data may include the strength of the force applied to the second object. Additionally, the second raw data may include whether the top of the screen was touched and whether the bottom of the screen was touched.

[0085] The collection unit (151) may present a third mission to the user and collect the result of performing the third mission from the user. In the present embodiment, the third raw data collected as the result of performing the third mission may include the time (timestamp) when the third mission starts and the data is collected. In addition, the third raw data may include information indicating whether the object touched by the user is the third object located on the left of two third objects (e.g., the 3-1 object) or the third object located on the right (e.g., the 3-2 object). In addition, the third raw data may include relative x-coordinates and y-coordinates (location x, y) for the touch points of the third objects. Here, the relative x-coordinates and y-coordinates may be measured based on the touch start point of one of the third objects (e.g., the 3-1 object or the 3-2 object). In addition, the third raw data may include absolute coordinates (page x, y) of the x-axis and y-axis indicating the location on the screen where the third objects are touched. Additionally, the third raw data may include the strength of the force applied to the third objects. Additionally, the third raw data may include the duration of the touch of the third objects.

[0086] The collection unit (151) may present a fourth mission to the user and collect the result of performing the fourth mission from the user. In the present embodiment, the fourth raw data collected as the result of performing the fourth mission may include the time (timestamp) when the fourth mission starts and the data is collected. In addition, the fourth raw data may include relative x-coordinates and y-coordinates (location x, y) for a point where a finger touches when tracing along a first-shaped guide pattern presented on the screen. Here, the relative x-coordinates and y-coordinates may be measured based on a point where the finger touches when tracing along the first-shaped guide pattern. In addition, the fourth raw data may include absolute coordinates (X, Y) of the x-axis and y-axis indicating a location on the screen where a finger touches during tracing. In addition, the fourth raw data may include the intensity (force) of a force that touches the first-shaped guide pattern during tracing. In addition, the fourth raw data may include the number of times tracing is attempted along the first-shaped guide pattern.

[0087] The collection unit (151) may present a fifth mission to the user and collect the result of performing the fifth mission from the user. In the present embodiment, the fifth raw data collected as the result of performing the fifth mission may include the time (timestamp) when the fifth mission starts and the data is collected. In addition, the fifth raw data may include relative x-coordinates and y-coordinates (location x, y) for a point where a finger touches when tracing along a second-shaped guide pattern presented on the screen. Here, the relative x-coordinates and y-coordinates may be measured based on a point where the finger touches when tracing along the second-shaped guide pattern. In addition, the fifth raw data may include absolute coordinates (X, Y) of the x-axis and y-axis indicating a location on the screen where a finger touches during tracing. In addition, the fifth raw data may include the intensity (force) of a force that touches the second-shaped guide pattern during tracing. In addition, the fifth raw data may include the number of times tracing is attempted along the second-shaped guide pattern.

[0088] The collection unit (151) may present a sixth mission to the user and collect the results of performing the sixth mission from the user. In the present embodiment, the sixth mission may include at least one of missions 6-1 to 6-5. Accordingly, the collection unit (151) may present missions 6-1 to 6-5 to the user and collect the results of performing missions 6-1 to 6-5 from the user. In the present embodiment, the sixth raw data collected as the results of performing the sixth mission may include an audio signal.

[0089] Mission 6-1 may include an action of listening to and speaking a first guide speech sound corresponding to a first script containing a basic vowel set (e.g., "aeiou"). In this embodiment, the 6-1 raw data collected as a result of performing Mission 6-1 may include a 6-1 audio signal.

[0090] Mission 6-2 may include an action of listening to and speaking a second guided speech sound corresponding to a second script containing syllables formed by combining specific consonants and specific vowels (e.g., Pataqa). In this embodiment, the 6-2 raw data collected as a result of performing Mission 6-2 may include a 6-2 audio signal.

[0091] Mission 6-3 may include an action to repeatedly utter the first script a maximum number of times for a preset period of time (e.g., 5 seconds). In this embodiment, the raw data 6-3 collected as a result of performing Mission 6-3 may include the audio signal 6-3.

[0092] Mission 6-4 may include an action to repeatedly utter the second script a maximum number of times for a preset period of time (e.g., 5 seconds). In this embodiment, the raw data 6-4 collected as a result of performing Mission 6-4 may include the audio signal 6-4.

[0093] Mission 6-5 may include an action of repeatedly uttering a third script comprising syllables (e.g., Ganada) formed by combining specific consonants and specific vowels different from the second script for a preset period of time (e.g., 5 seconds) a maximum number of times. In the present embodiment, the 6-5 raw data collected as a result of performing Mission 6-5 may include a 6-5 audio signal.

[0094] The collection unit (151) can present the seventh mission to the user and collect the results of the seventh mission from the user. In the present embodiment, the seventh raw data collected as a result of the seventh mission may include one or more of the following: the number of the questionnaire question (question ID), the response result to the questionnaire, the response time, and the category of the questionnaire question.

[0095] The collection unit (151) may present the eighth mission to the user and collect the result of performing the eighth mission from the user. In the present embodiment, the eighth raw data collected as the result of performing the eighth mission may include the question number (question ID) of the eighth mission. In addition, the eighth raw data may include a response result indicating whether the guide color displayed in the first area and the color displayed in the second area as a result of manipulating the fourth object on the color selection palette match. In addition, the eighth raw data may include the time (response time) taken for the user to check the color in the first area and for the correct color to be displayed in the second area as a result of manipulating the fourth object on the color selection palette. Here, the correct color may include the same color as the guide color.

[0096] The preprocessing unit (152) may perform processing and preprocessing on the raw data collected from the collection unit (151). In the present embodiment, processing and preprocessing may include removing unnecessary data from the performance results of one or more athletic ability evaluation missions and the performance results of one or more non-athletic ability evaluation missions.

[0097] The preprocessing unit (152) can convert the information about the user collected from the collection unit (151) into numeric data. The preprocessing unit (152) can convert the presence or absence of Parkinson's disease included in the information about the user into 1 if the user has Parkinson's disease and 0 if the user is normal. The preprocessing unit (152) can convert the gender included in the information about the user into 1 if the user is male and 0 if the user is female. The preprocessing unit (152) can convert the dominant hand among the two hands into 1 if the user is right-handed and 0 if the user is left-handed. The preprocessing unit (152) can omit the conversion of the age and years of education included in the information about the user into numeric data. This is because the age and years of education are already numbers.

[0098] The preprocessing unit (152) can perform processing and preprocessing on the first raw data collected from the collection unit (151). The preprocessing unit (152) can delete data if the number of pointers counted on the screen of the first object included in the first raw data is 1 or less. The reason for deleting the data in this way may be because the data is likely to be incomplete or contain errors. The preprocessing unit (152) can convert the first raw data into first time series data and extract statistical analysis indices and frequency domain analysis indices. Here, the statistical analysis indices may include the mean, standard deviation, maximum value, minimum value, skewness, kurtosis, and the range indicating the difference between the maximum and minimum values ​​of the first time series data. In addition, the frequency domain analysis indices may include the number of local maxima and local minima extracted from the first time series data, and the dominant frequency. Here, the dominant frequency may include the strongest frequency component in the first time series data. The preprocessing unit (152) may extract the number of times and speed at which the first object is bent and stretched using a finger for each section of the first time series data.

[0099] The preprocessing unit (152) can perform processing and preprocessing on the second raw data collected from the collection unit (151). The preprocessing unit (152) can convert the second raw data into second time series data and extract statistical analysis indices and frequency domain analysis indices. Here, the statistical analysis indices can include the mean, standard deviation, maximum value, minimum value, skewness, kurtosis, and range indicating the difference between the maximum and minimum values ​​of the second time series data. In addition, the frequency domain analysis indices can include the number of local maxima and local minima extracted from the second time series data, and the dominant frequency. The preprocessing unit (152) can extract the number of times the second object is repeatedly moved from the top to the bottom while being touched for each section of the second time series data.

[0100] The preprocessing unit (152) can perform processing and preprocessing on the third raw data collected from the collection unit (151). The preprocessing unit (152) can convert the third raw data into third time series data and extract statistical analysis indices and frequency domain analysis indices. Here, the statistical analysis indices can include the mean, standard deviation, maximum value, minimum value, skewness, kurtosis, and range indicating the difference between the maximum and minimum values ​​of the third time series data. In addition, the frequency domain analysis indices can include the number of local maxima and local minima extracted from the third time series data, and the dominant frequency. The preprocessing unit (152) can extract the number of times two third objects were touched and the speed of touching two third objects for each section of the second time series data.

[0101] The preprocessing unit (152) can perform processing and preprocessing on the fourth raw data collected from the collection unit (151). The preprocessing unit (152) can extract the number of times tracing the first shape guide pattern was attempted using a finger, and delete the rest, leaving only the tracing-related data of the longest tracing cycle. The preprocessing unit (152) can convert the fourth raw data into fourth time series data and extract statistical analysis indices and frequency domain analysis indices. Here, the statistical analysis indices can include the mean, standard deviation, maximum value, minimum value, skewness, kurtosis, and the range indicating the difference between the maximum and minimum values ​​of the fourth time series data. In addition, the frequency domain analysis indices can include the number of local maxima and local minima extracted from the fourth time series data, and the dominant frequency. The preprocessing unit (152) can extract the first curvature corresponding to the result of tracing the first shape guide pattern using a finger from the fourth raw data.

[0102] The preprocessing unit (152) can perform processing and preprocessing on the fifth raw data collected from the collection unit (151). The preprocessing unit (152) can extract the number of times the second shape guide pattern was attempted to be traced using a finger, and delete the rest, leaving only the tracing-related data of the longest tracing cycle. The preprocessing unit (152) can convert the fifth raw data into fifth time series data and extract statistical analysis indices and frequency domain analysis indices. Here, the statistical analysis indices can include the mean, standard deviation, maximum value, minimum value, skewness, kurtosis, and the range indicating the difference between the maximum and minimum values ​​of the fifth time series data. In addition, the frequency domain analysis indices can include the number of local maxima and local minima extracted from the fifth time series data, and the dominant frequency. The preprocessing unit (152) can extract a second curvature corresponding to the result of tracing a second shape guide pattern using a finger from the fifth raw data.

[0103] The preprocessing unit (152) can perform processing and preprocessing on the sixth raw data collected from the collection unit (151). In the present embodiment, the sixth raw data can include a sixth audio signal. In the present embodiment, the sixth raw data can include at least one of the 6-1 raw data to the 6-5 raw data. The preprocessing unit (152) can convert the sixth audio signal included in the sixth raw data into a mono channel audio signal. The sixth audio signal can include a stereo or multi-channel audio signal, and the preprocessing unit (152) can convert the stereo or multi-channel audio signal into a mono channel. In addition, the preprocessing unit (152) can remove noise from the sixth audio signal included in the sixth raw data. In addition, when there are multiple speakers in the sixth audio signal, the preprocessing unit (152) can separate the audio signal for each speaker.

[0104] The preprocessing unit (152) may perform processing and preprocessing on the seventh raw data collected from the collection unit (151). The preprocessing unit (152) may convert the seventh raw data into numeric data. In addition, if there are duplicate questionnaire question numbers (question IDs), the preprocessing unit (152) may keep the last questionnaire question and delete the rest. The preprocessing result of the seventh raw data may include the response results to the questionnaire.

[0105] The preprocessing unit (152) can perform processing and preprocessing on the 8th raw data collected from the collection unit (151). If there are duplicate question numbers (question IDs) of the 8th mission, the preprocessing unit (152) can keep the last question number and delete the rest. In addition, the preprocessing unit (152) can extract a color error including the difference between the guide color presented in the first area and the color displayed in the second area from the 8th raw data, and a root mean square error (RMSE) calculated based on the color error.

[0106] The extraction unit (153) can extract one or more motor ability features and one or more non-motor ability features from the data extracted from the preprocessing unit (152). The extraction unit (153) can extract one or more first motor ability features from the performance results of one or more preprocessed first motor ability evaluation missions. The extraction unit (153) can extract one or more second motor ability features from the performance results of one or more preprocessed second motor ability evaluation missions. The extraction unit (153) can extract one or more first non-motor ability features from the performance results of one or more preprocessed first non-motor ability evaluation missions. The extraction unit (153) can extract one or more second non-motor ability features from the performance results of one or more preprocessed second non-motor ability evaluation missions.

[0107] The extraction unit (153) can extract statistical analysis indices and frequency domain analysis indices for the first time series data from the performance results of the preprocessed first mission, and a first feature including the number of times and speed of bending and stretching the first object using a finger.

[0108] The extraction unit (153) can extract statistical analysis indicators and frequency domain analysis indicators for the second time series data collected from the performance results of the preprocessed second mission, and a second feature including the number of times the second object is repeatedly moved from the top to the bottom while being touched.

[0109] The extraction unit (153) can extract statistical analysis indicators and frequency domain analysis indicators for third time series data collected from the performance results of the preprocessed third mission, and third features including the number of times two third objects were touched alternately with one finger and the touch speed.

[0110] The extraction unit (153) can extract a fourth feature including a statistical analysis index and a frequency domain analysis index for the fourth time series data collected from the performance results of the preprocessed fourth mission, and a first curvature calculated in response to the result of tracing the first shape guide pattern using a finger.

[0111] The extraction unit (153) can extract a fifth feature including a statistical analysis index and a frequency domain analysis index for the fifth time series data collected from the performance results of the preprocessed fifth mission, and a second curvature calculated in response to the result of tracing the second shape guide pattern using a finger.

[0112] The extraction unit (153) can extract the 6-1 feature and the 6-2 feature from the audio signal included in the speech result collected from the performance result of the preprocessed 6th mission.

[0113] The extraction unit (153) can extract the 6-1 feature including at least one of a fundamental frequency including the lowest frequency that determines the pitch of speech in an audio signal included in a speech result collected from the performance result of the preprocessed 6th mission, an amplitude variation (shimmer) of the audio signal, a jitter including a periodic pitch variation of the audio signal, a harmonic to noise ratio (HNR) including a ratio of harmonic components and noise components in the audio signal, an intensity or loudness of the audio signal, and a speech rate in the audio signal.

[0114] In this embodiment, one or more of the fundamental frequency, shimmer, harmonic to noise ratio (HNR), intensity or loudness, and speech rate may be named as audio characteristic indicators.

[0115] In the present embodiment, the extraction unit (153) can extract the 6-2 feature by applying an artificial intelligence model to the audio signal included in the speech result collected from the execution result of the preprocessed 6th mission. The extraction unit (153) can receive a preprocessed audio signal by converting the audio signal included in the speech result collected from the execution result of the 6th mission from the preprocessing unit (152) into a mono channel audio signal, removing noise, or separating the audio signal by speaker. The extraction unit (153) can output a high-dimensional feature vector corresponding to the preprocessed audio signal by using a pre-trained self-supervised learning-based deep neural network model that takes an audio signal as an input and outputs a high-dimensional feature vector for the audio signal. The extraction unit (153) can determine the high-dimensional feature vector output from the deep neural network model as the 6-2 feature.

[0116] In this embodiment, a deep neural network model based on self-supervised learning may include the wave2vec model. In this embodiment, wave2vec is a self-supervised speech recognition model developed by Facebook AI Research. It uses a convolutional neural network (CNN) as an encoder and can extract high-quality audio speech features by detecting useful contextual information from the input audio signal. Based on the vector in the last hidden layer of the wave2vec model, which contains useful and abstract features of the audio signal, the sequence in the hidden state can be output to extract a 512-dimensional feature vector (a high-dimensional feature vector).

[0117] The extraction unit (153) can extract the seventh feature including the answer to the above questionnaire collected from the performance results of the preprocessed seventh mission.

[0118] The extraction unit (153) can extract the color error including the difference between the guide color presented in the first area and the color displayed in the second area based on the data collected from the performance results of the preprocessed eighth mission, and the eighth feature including the root mean square error (RMSE) calculated based on the color error.

[0119] The diagnostic unit (154) can diagnose the presence or absence of Parkinson's disease in a user based on at least one of the extracted motor ability features and one or more non-motor ability features. The diagnostic unit (154) can utilize various features (the first to eighth features) singly or in combination and apply them to an artificial intelligence model.

[0120] In this embodiment, the diagnostic unit (154) can diagnose the presence or absence of Parkinson's disease using a single feature. For example, when diagnosing the presence or absence of Parkinson's disease using the first feature or the third feature, the diagnostic unit (154) can utilize the first diagnostic model (410 of FIG. 4A, machine learning model). Accordingly, regardless of which evaluation mission among the first to eighth evaluation missions is performed, the diagnostic unit (154) can diagnose the presence or absence of Parkinson's disease for the performed evaluation mission.

[0121] In the present embodiment, the diagnosis unit (154) can diagnose the presence or absence of Parkinson's disease by utilizing a combination of multiple features. For example, when the diagnosis unit (154) diagnoses the presence or absence of Parkinson's disease by combining at least one of the first to eighth features, the diagnosis unit (154) can utilize the second diagnosis model (420 of FIG. 4b, deep learning model). In addition, when the diagnosis unit (154) diagnoses the presence or absence of Parkinson's disease by combining at least one of the first to eighth features and a sixth image converted into an image based on the performance result of the sixth mission, the diagnosis unit (154) can utilize the fourth diagnosis model (440 of FIG. 4f, deep learning model). From this, the diagnosis unit (154) can provide a more comprehensive and accurate diagnosis result than when using a single feature through a wide range of data obtained by combining various features.

[0122] The diagnostic unit (154) can comprehensively utilize multiple features to more precisely evaluate and diagnose various symptoms of Parkinson's disease. During this process, at least one of the first to eighth features is processed through a diagnostic model, thereby deriving a diagnostic result most appropriate for the user's condition. The diagnostic unit (154) utilizes these features to enable a Parkinson's disease diagnosis tailored to the user's characteristics, and by selecting and applying an optimized model based on the user's characteristics, it can support early detection of the disease and effective treatment strategies.

[0123] Hereinafter, the operation of the diagnostic unit (154) will be described in detail with reference to FIGS. 4a to 4i. In the present embodiment, at least one of FIGS. 4a to 4i may be configured as a diagnostic unit (154), and the entirety of FIGS. 4a to 4i may be configured as one diagnostic unit (154).

[0124] FIG. 4A illustrates the configuration of a diagnostic unit (154) according to one embodiment. Referring to FIG. 4A, the diagnostic unit (154) may include a first diagnostic model (410). In the present embodiment, the first diagnostic model (410) may include a machine learning model.

[0125] The diagnosis unit (154) can diagnose the presence or absence of Parkinson's disease in a user corresponding to at least one of the information about the user and the results of extracting one or more motor ability features and one or more non-motor ability features by using a first diagnosis model (410) that diagnoses the presence or absence of Parkinson's disease based on information about the user and at least one of the results of extracting one or more motor ability features and one or more non-motor ability features. In the present embodiment, the first diagnosis model (410) can include a machine learning model that is trained in a supervised learning manner using training data that takes as input information about the user and at least one of the one or more motor ability features and one or more non-motor ability features, and labels the presence or absence of Parkinson's disease.

[0126] The diagnostic unit (154) can train the initially set first diagnostic model (410) using the labeled training data in a supervised learning manner. Here, the initially set first diagnostic model (410) is an initial model designed to be configured as a model capable of diagnosing the presence or absence of Parkinson's disease, and the parameter values ​​are set to arbitrary initial values. As the initial model is trained using the above-described training data, the parameter values ​​are optimized, so that the first diagnostic model (410) can be completed to accurately predict the presence or absence of Parkinson's disease.

[0127] In the present embodiment, one or more motor ability features and one or more non-motor ability features may be identical to at least one of the first to eighth features described above. For convenience of explanation, the one or more motor ability features and the one or more non-motor ability features will be referred to as at least one of the first to eighth features below.

[0128] FIG. 4B illustrates a configuration of a diagnostic unit (154) according to another embodiment. Referring to FIG. 4B, the diagnostic unit (154) may include a second diagnostic model (420). In the present embodiment, the second diagnostic model (420) may include a first deep learning model.

[0129] The diagnosis unit (154) can diagnose the presence or absence of Parkinson's disease in a user corresponding to information about the user and at least one of the first to eighth features by using a second diagnosis model (420) that diagnoses the presence or absence of Parkinson's disease based on information about the user and at least one of the first to eighth features. In the present embodiment, the second diagnosis model (420) can include a first deep learning model trained in a supervised learning manner using training data that takes information about the user and at least one of the first to eighth features as inputs and labels the presence or absence of Parkinson's disease.

[0130] The diagnostic unit (154) can train the initially set second diagnostic model (420) using the labeled training data in a supervised learning manner. Here, the initially set second diagnostic model (420) is an initial model designed to be configured as a model capable of diagnosing the presence or absence of Parkinson's disease, and the parameter values ​​are set to arbitrary initial values. As the initial model is trained using the above-described training data, the parameter values ​​are optimized, so that the second diagnostic model (420) can be completed to accurately predict the presence or absence of Parkinson's disease.

[0131] Fig. 4c illustrates the configuration of the second diagnostic model (420) illustrated in Fig. 4b. Referring to Fig. 4c, the second diagnostic model (420) may include a first FC layer (421).

[0132] The second diagnostic model (420) can generate a diagnostic result including the presence or absence of Parkinson's disease based on the combined result of at least one of the first to eighth features and information about the user through the first fully connected (FC) layer (421). In the present embodiment, the second diagnostic model (420) does not require separate image processing, and a diagnosis can be made using only numerical features.

[0133] Typically, FC layers are utilized at the end of a neural network to integrate features extracted through the network into a single vector, which is then used to derive the final output. Because every input neuron in the FC layer is connected to every neuron in the next layer, it possesses the ability to effectively learn complex relationships and patterns between input data.

[0134] In this embodiment, the first FC layer (421) can process complex interactions between at least one of the first to eighth features and information about the user, and generate a diagnostic result including the presence or absence of Parkinson's disease based on the complex interactions.

[0135] FIG. 4d illustrates a configuration of a diagnostic unit (154) according to another embodiment. Referring to FIG. 4d, the diagnostic unit (154) may include a third diagnostic model (430). In the present embodiment, the third diagnostic model (430) may include a second deep learning model.

[0136] The diagnosis unit (154) can diagnose the presence or absence of Parkinson's disease of a user corresponding to information about the user, at least one of the first to eighth features, a fourth image generated based on the performance result of the fourth mission included in the motor ability feature, and at least one of the fifth images generated based on the performance result of the fifth mission included in the motor ability feature, by using a third diagnosis model (430) that diagnoses the presence or absence of Parkinson's disease of the user corresponding to at least one of the fourth image generated based on the performance result of the fifth mission included in the motor ability feature, and a fifth image generated based on the performance result of the fifth mission included in the motor ability feature. In the present embodiment, the third diagnostic model (430) may include a second deep learning model trained in a supervised learning manner using training data labeled with the presence or absence of Parkinson's disease, which inputs information about the user, at least one of the first to eighth features, and at least one of a fourth image generated based on the performance result of the fourth mission included in the motor ability feature and a fifth image generated based on the performance result of the fifth mission included in the motor ability feature.

[0137] In this embodiment, the diagnostic unit (154) can generate at least one of the fourth image and the fifth image and input it into the third diagnostic model (430). The diagnostic unit (154) can perform preprocessing on the execution result of the fourth mission and convert the generated fourth time series data into the third image. The diagnostic unit (154) can perform preprocessing on the execution result of the fifth mission and convert the generated fifth time series data into the fifth image.

[0138] At least one of the fourth and fifth images generated by the diagnostic unit (154) may be stacked for each channel. Here, the expression "stacked for each channel" may refer to a process in which channels are sequentially stacked in image data having multiple channels. For example, in image processing, each of the RGB (red, green, blue) colors has a separate channel, and these channels can be combined to form an entire color image. This process enables analysis by stacking complex information in image data, and enables effective processing of various data dimensions.

[0139] Fig. 4e illustrates the configuration of the third diagnostic model (430) illustrated in Fig. 4d. Referring to Fig. 4e, the third diagnostic model (430) may include a first convolutional neural network (CNN) layer (431) and a second FC layer (432).

[0140] The third diagnostic model (430) can convert at least one of the fourth image and the fifth image stacked for each channel into a first one-dimensional vector through the first CNN layer (431).

[0141] The first CNN layer (431) can analyze at least one of the fourth and fifth images stacked for each channel to extract visual features such as boundaries, color changes, shapes, sizes, and patterns, and can condense the visual features into a first one-dimensional vector. The first CNN layer (431) can transfer the converted first one-dimensional vector to the second FC layer (432).

[0142] The third diagnostic model (430) can generate a diagnostic result including the presence or absence of Parkinson's disease based on a combination of at least one of the first to eighth features, the first one-dimensional vector, and information about the user through the second FC layer (432).

[0143] In this embodiment, the second FC layer (432) processes a complex interaction between at least one of the first to eighth features, the first one-dimensional vector output from the first CNN layer (431), and information about the user, and based on this, can generate a diagnostic result including the presence or absence of Parkinson's disease.

[0144] FIG. 4F illustrates the configuration of a diagnostic unit (154) according to another embodiment. Referring to FIG. 4F, the diagnostic unit (154) may include a fourth diagnostic model (440). In the present embodiment, the fourth diagnostic model (440) may include a third deep learning model.

[0145] The diagnosis unit (154) can diagnose the presence or absence of Parkinson's disease of a user corresponding to information about the user, at least one of the first to eighth features, and at least one of the sixth images generated based on the performance result of the sixth mission included in the motor ability feature, by using a fourth diagnosis model (440) that diagnoses the presence or absence of Parkinson's disease based on information about the user, at least one of the first to eighth features, and the sixth image generated based on the performance result of the sixth mission included in the motor ability feature. In the present embodiment, the fourth diagnosis model (440) may include a third deep learning model that is trained in a supervised learning manner by training data that takes as input information about the user, at least one of the first to eighth features, and at least one of the sixth images generated based on the performance result of the sixth mission included in the motor ability feature, and labels the presence or absence of Parkinson's disease.

[0146] In the present embodiment, the diagnostic unit (154) can generate a sixth image and input it into the fourth diagnostic model (440). The diagnostic unit (154) can perform one or more preprocessing operations on the audio signal collected from the execution result of the sixth mission, including conversion into a mono channel audio signal, noise removal, and speaker separation. The diagnostic unit (154) can convert the audio signal on which the preprocessing has been performed into a sixth image. In the present embodiment, the sixth image can include a Mel-spectogram image. The diagnostic unit (154) can generate a Mel-spectogram image as an image that visually represents changes in frequency components over time by converting the audio signal into a Mel-scale frequency spectrum.

[0147] Fig. 4g illustrates the configuration of the fourth diagnostic model (440) illustrated in Fig. 4f. Referring to Fig. 4e, the fourth diagnostic model (440) may include a second CNN layer (441) and a third FC layer (442).

[0148] The fourth diagnostic model (440) can convert the sixth image into a second one-dimensional vector through the second CNN layer (441).

[0149] The second CNN layer (441) analyzes the sixth image (Mel-spectogram image) to extract the frequency change over time, which is one of the important features of the voice, as a visual feature, and can convert the visual feature into a second one-dimensional vector by condensing it. The second CNN layer (441) can transfer the converted second one-dimensional vector to the third FC layer (442).

[0150] The fourth diagnostic model (440) can generate a diagnostic result including the presence or absence of Parkinson's disease based on a combination of at least one of the first to eighth features, the first one-dimensional vector, and information about the user through the third FC layer (442).

[0151] In this embodiment, the third FC layer (442) processes a complex interaction between at least one of the first to eighth features, the second one-dimensional vector output from the second CNN layer (441), and information about the user, and based on this, can generate a diagnostic result including the presence or absence of Parkinson's disease.

[0152] FIG. 4h illustrates the configuration of a diagnostic unit (154) according to another embodiment. Referring to FIG. 4h, the diagnostic unit (154) may include a fifth diagnostic model (450). In the present embodiment, the fifth diagnostic model (450) may include a fourth deep learning model.

[0153] The diagnosis unit (154) uses a fifth diagnosis model (450) that diagnoses the presence or absence of Parkinson's disease based on information about the user, at least one of the first to eighth features, a fourth image generated based on the performance result of the fourth mission included in the motor ability feature, at least one of the fifth images generated based on the performance result of the fifth mission included in the motor ability feature, and a sixth image generated based on the performance result of the sixth mission included in the motor ability feature, to diagnose the presence or absence of Parkinson's disease in the user corresponding to information about the user, at least one of the first to eighth features, a fourth image generated based on the performance result of the fourth mission included in the motor ability feature, at least one of the fifth images generated based on the performance result of the fifth mission included in the motor ability feature, and a sixth image generated based on the performance result of the sixth mission included in the motor ability feature. In the present embodiment, the fifth diagnostic model (450) may include a third deep learning model trained in a supervised learning manner by training data labeled with the presence or absence of Parkinson's disease, which inputs information about the user, at least one of the first to eighth features, a fourth image generated based on the performance result of the fourth mission included in the motor ability feature, at least one of the fifth images generated based on the performance result of the fifth mission included in the motor ability feature, and at least one of the sixth images generated based on the performance result of the sixth mission included in the motor ability feature.

[0154] In this embodiment, the diagnostic unit (154) can generate at least one of the fourth image and the fifth image and input it into the third diagnostic model (430). The diagnostic unit (154) can generate the sixth image and input it into the fifth diagnostic model (450).

[0155] Fig. 4i illustrates the configuration of the fifth diagnostic model (450) illustrated in Fig. 4h. Referring to Fig. 4i, the fifth diagnostic model (450) may include a first CNN layer (451), a second CNN layer (452), and a fifth FC layer (453).

[0156] The fifth diagnostic model (450) can convert at least one of the fourth image and the fifth image stacked for each channel into a first one-dimensional vector through the first CNN layer (451).

[0157] The fifth diagnostic model (450) can convert the sixth image into a second one-dimensional vector through the second CNN layer (452).

[0158] The fifth diagnostic model (450) can generate a diagnostic result including the presence or absence of Parkinson's disease based on a combination of at least one of the first to eighth features, the first one-dimensional vector, the second one-dimensional vector, and information about the user through the fifth FC layer (453).

[0159] In this embodiment, the fifth FC layer (453) processes a complex interaction between at least one of the first to eighth features, the first one-dimensional vector output from the first CNN layer (451), the second one-dimensional vector output from the second CNN layer (452), and information about the user, and based on this, can generate a diagnostic result including the presence or absence of Parkinson's disease.

[0160] In this embodiment, the structures of the first diagnostic model (410) to the fifth diagnostic model (450) can be designed for the purpose of comprehensively analyzing various characteristics of Parkinson's disease through various input data and increasing diagnostic accuracy.

[0161] Additionally, in the structures of the second diagnostic model (420) to the fifth diagnostic model (450), the CNN layer is an optional element and can only be applied when using image data. That is, if the mission results are not converted into an image, the CNN layer is not selected, and the presence or absence of Parkinson's disease can be diagnosed using the FC layer. The FC layer performs the final diagnosis based on the integrated features, and can output a one-dimensional vector containing two numbers as the final output.

[0162] The output one-dimensional vector can be used as a criterion for making a diagnostic decision, including whether or not Parkinson's disease is present. By comparing the first and second numbers, if the first number is greater than the second number, the user can be diagnosed with Parkinson's disease. Conversely, if the first number is less than the second number, the user can be diagnosed as normal. In this way, the FC layer effectively synthesizes the features extracted through the input neural network, providing crucial diagnostic information for accurately diagnosing the presence or absence of Parkinson's disease.

[0163] Figures 5a to 5h are exemplary diagrams of Parkinson's disease diagnosis screens provided to a user terminal (200) according to the present embodiment. In the following description, any parts that overlap with the descriptions of Figures 1 to 4 will be omitted.

[0164] Referring to FIG. 5, FIG. 5a illustrates a method for performing a first mission presented on a screen of a user terminal (200). For the first mission, the diagnostic management unit (150) can guide a user to flex and extend a first object (511, e.g., a balloon) presented on the screen as much as possible using a finger for a preset period of time (e.g., 20 seconds). The user can perform the first mission according to the guidance.

[0165] Figure 5b illustrates a method for performing a second mission presented on the screen of a user terminal (200). For the second mission, the diagnostic management unit (150) may guide the user to repeatedly move the second object (521) presented on the screen from the top to the bottom of the screen at maximum speed while touching it for a preset period of time (e.g., 20 seconds). The user may perform the second mission according to the guidance.

[0166] Figure 5c illustrates a method for performing the third mission presented on the screen of a user terminal (200). For the third mission, the diagnostic management unit (150) may guide the user to alternately touch two third objects (531, 532) presented on the screen at maximum speed using one finger for a preset period of time. The user may perform the third mission according to the guidance.

[0167] Figure 5d illustrates a method for performing the fourth mission presented on the screen of a user terminal (200). The diagnostic management unit (150) can guide the user to trace the first shape guide pattern (541, e.g., spiral) presented on the screen using a finger for the fourth mission. The user can perform the fourth mission according to the guidance.

[0168] Figure 5e illustrates a method for performing the fifth mission presented on the screen of a user terminal (200). For the fifth mission, the diagnostic management unit (150) can guide the user to trace a second-shaped guide pattern (551, e.g., a wave) presented on the screen using a finger. The user can perform the fifth mission according to the guidance.

[0169] Figure 5f illustrates a method for performing the sixth mission presented on the screen of the user terminal (200). For the sixth mission, the diagnostic management unit (150) can guide the user to listen to the guide speech voice corresponding to the script (e.g., "Aeiou") presented on the screen and speak accordingly. Furthermore, the diagnostic management unit (150) can guide the user to press the recording start button (561) and speak when ready. The user can perform the sixth mission according to the guidance.

[0170] Figure 5g illustrates an example of the seventh mission presented on the screen of a user terminal (200). For the seventh mission, the diagnostic management unit (150) may provide a questionnaire on the screen to measure the user's non-motor ability status, including emotional responses, stress levels, depression symptoms, and other emotional states, and guide the user to answer the questionnaire. Following the guidance, the user may select a "Yes" button (571) or a "No" button (572) for the presented questionnaire. The user may also perform the seventh mission by entering answers to other questionnaires, following the guidance.

[0171] FIG. 5h illustrates a method for performing the 8th mission presented on the screen of the user terminal (200). For the 8th mission, the diagnostic management unit (150) may present a first area (581) on the screen where guide colors are presented, a color selection palette (583) for selecting an arbitrary color by moving a fourth object (582), and a second area (584) for displaying a color selected by moving the fourth object (582) on the color selection palette (583). The diagnostic management unit (150) may guide the user to move the fourth object (582) on the color selection palette (583) so that a color matching the first area (581) is displayed in the second area (584). The user may perform the 8th mission according to the guidance.

[0172] Fig. 6 is a block diagram schematically illustrating the configuration of a Parkinson's disease diagnosis device according to another embodiment. In the following description, any part that overlaps with the description of Figs. 1 to 5 will be omitted. Referring to Fig. 6, a Parkinson's disease diagnosis device (100) according to another embodiment may include a processor (170) and a memory (180).

[0173] In this embodiment, the processor (170) can process functions performed by the communication unit (110), storage medium (120), program storage unit (130), database (140), diagnostic management unit (150), and control unit (160) disclosed in FIGS. 2 and 3.

[0174] The processor (170) can process computer program commands by performing basic arithmetic, logic, and input / output operations. Here, the commands may be provided from one or more of the program storage unit (130), the database (140), and the control unit (160). In addition, the processor (170) can generally control the operations of other components associated with the Parkinson's disease diagnosis device (100).

[0175] Meanwhile, the processor (170) may perform at least a portion of the data analysis, processing, and result information generation for performing the above-described operations using at least one of a machine learning, neural network, or deep learning algorithm as a rule-based or artificial intelligence (AI) algorithm. Examples of the neural network may include models such as a CNN (Convolutional Neural Network), a DNN (Deep Neural Network), and an RNN (Recurrent Neural Network).

[0176] For example, the processor (170) may be implemented as an array of a plurality of logic gates, or it may be implemented as a combination of a general-purpose microprocessor and a memory storing a program executable on the microprocessor. For example, the processor may include a general-purpose processor, a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a controller, a microcontroller, a state machine, etc. In some environments, the processor (170) may include an application-specific integrated circuit (ASIC), a programmable logic device (PLD), a field programmable gate array (FPGA), etc. For example, the processor (170) may refer to a combination of processing devices, such as a combination of a digital signal processor (DSP) and a microprocessor, a combination of a plurality of microprocessors, a combination of one or more microprocessors coupled with a digital signal processor (DSP) core, or any other such combination of configurations.

[0177] The memory (180) is operably connected to the processor (170) and can store at least one code associated with an operation performed by the processor (170).

[0178] In addition, the memory (180) may perform a function of temporarily or permanently storing data processed by the processor (170), and may include data constructed as a database (140). Here, the memory (180) may include a magnetic storage medium or a flash storage medium, but the scope of the present invention is not limited thereto. The memory (180) may include a built-in memory and / or an external memory, and may include a volatile memory such as DRAM, SRAM, or SDRAM, a non-volatile memory such as OTPROM, PROM, EPROM, EEPROM, mask ROM, flash ROM, NAND flash memory, or NOR flash memory, a flash drive such as an SSD, a CF card, an SD card, a Micro-SD card, a Mini-SD card, an xD card, or a memory stick, or a storage device such as an HDD.

[0179] Fig. 7 is a flowchart illustrating a Parkinson's disease diagnosis method according to the present embodiment. In the following description, any parts that overlap with the descriptions of Figs. 1 to 6 will be omitted. The Parkinson's disease diagnosis method according to the present embodiment will be described assuming that the Parkinson's disease diagnosis device (100) is performed by a processor (170) with the assistance of peripheral components.

[0180] Referring to FIG. 7, in step S710, the processor (170) may present one or more motor ability evaluation missions and one or more non-motor ability evaluation missions to the user, and collect performance results for one or more motor ability evaluation missions and performance results for one or more non-motor ability evaluation missions from the user.

[0181] In step S720, the processor (170) may extract one or more motor ability features and one or more non-motor ability features from the performance results of one or more motor ability evaluation missions and the performance results of one or more non-motor ability evaluation missions. In the present embodiment, the processor (170) may perform preprocessing on the performance results of the motor ability evaluation missions and the performance results of the non-motor ability evaluation missions before extracting the motor ability features and the non-motor ability features.

[0182] In step S830, the processor (170) can diagnose whether the user has Parkinson's disease based on at least one of the results of extracting one or more motor ability features and one or more non-motor ability features.

[0183] The embodiments of the present invention described above may be implemented in the form of a computer program that can be executed through various components on a computer, and such a computer program may be recorded on a computer-readable medium. At this time, the medium may include a magnetic medium such as a hard disk, a floppy disk, and a magnetic tape, an optical recording medium such as a CD-ROM and a DVD, a magneto-optical medium such as a floptical disk, and a hardware device specifically configured to store and execute program instructions, such as a ROM, a RAM, a flash memory, etc.

[0184] Meanwhile, the computer program may be specifically designed and constructed for the present invention, or may be one known and available to those skilled in the computer software field. Examples of computer programs may include not only machine language code, such as that generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like.

[0185] The use of the term "above" and similar referential terms in the specification of the present invention (especially in the claims) may refer to both singular and plural. Furthermore, when a range is described in the present invention, it is intended that the invention encompasses inventions that apply individual values ​​falling within the range (unless otherwise stated), and is equivalent to describing each individual value constituting the range in the detailed description of the invention.

[0186] Unless the steps constituting the method according to the present invention are explicitly described in a specific order or are not described to the contrary, the steps may be performed in any appropriate order. The present invention is not necessarily limited to the order in which the steps are described. The use of all examples or exemplary terms (e.g., "for example," etc.) in the present invention is merely intended to illustrate the present invention in detail, and the scope of the present invention is not limited by the examples or exemplary terms unless otherwise defined by the claims. Furthermore, those skilled in the art will appreciate that various modifications, combinations, and variations can be configured according to design conditions and factors within the scope of the appended claims or their equivalents.

[0187] Therefore, the idea of ​​the present invention should not be limited to the embodiments described above, and all scopes equivalent to or equivalently modified from the following claims as well as the claims are considered to fall within the scope of the idea of ​​the present invention.

Claims

1. A method for diagnosing Parkinson's disease performed by a processor of a Parkinson's disease diagnosis device, A step of presenting one or more motor ability evaluation missions and one or more non-motor ability evaluation missions to a user, and collecting performance results for the one or more motor ability evaluation missions and performance results for the one or more non-motor ability evaluation missions from the user; A step of extracting one or more motor ability features and one or more non-motor ability features from the performance results of one or more motor ability evaluation missions and the performance results of one or more non-motor ability evaluation missions; and Comprising a step of diagnosing whether the user has Parkinson's disease based on at least one of the results of extracting the one or more motor ability features and the one or more non-motor ability features. How to diagnose Parkinson's disease.

2. In paragraph 1, The above collecting steps are: A step of presenting one or more first exercise ability evaluation missions to a user and collecting performance results for the one or more first exercise ability evaluation missions from the user, The step of collecting the results of the above first exercise ability evaluation mission is: Including a step of presenting at least one mission among the first to fifth missions and collecting the results of performing said one mission, The above first mission includes an action of flexing and extending the first object presented on the screen as much as possible using a finger for a preset period of time. The second mission includes the action of repeatedly moving the second object presented on the screen from the top to the bottom at the highest speed while touching it for a preset period of time. The third mission involves the action of alternately touching two third objects presented on the screen at the highest speed using one finger for a preset period of time. The fourth mission includes an action of tracing the first shape guide pattern presented on the screen using a finger. The fifth mission includes an action of tracing a second shape guide pattern different from the first shape presented on the screen using a finger. How to diagnose Parkinson's disease.

3. In paragraph 1, The above collecting steps are: A step of presenting a second exercise ability evaluation mission to a user and collecting the performance results of the second exercise ability evaluation mission from the user, The step of collecting the results of the above second exercise ability evaluation mission is: Including a step of presenting a sixth mission and collecting the results of performing the sixth mission, The above 6th mission includes an action of listening to a guide speech voice corresponding to a script presented on the screen and speaking accordingly. How to diagnose Parkinson's disease.

4. In paragraph 3, The step of collecting the results of the above 6th mission is: Including a step of presenting one or more missions from Mission 6-1 to Mission 6-5 and collecting the performance results for said one or more missions, The above 6-1 mission includes an action of listening to the first guide speech sound corresponding to the first script including the basic vowel set and speaking accordingly. Mission 6-2 includes the action of listening to and speaking a second guide speech sound corresponding to a second script containing syllables formed by combining specific consonants and specific vowels. Mission 6-3 includes an action to repeat the first script as many times as possible for a preset amount of time. Mission 6-4 includes an action to repeat the above second script the maximum number of times for a preset amount of time. The above 6-5 mission includes an action to repeatedly utter a third script containing syllables formed by combining specific consonants and specific vowels different from the second script for the maximum number of times during a preset period of time. How to diagnose Parkinson's disease.

5. In paragraph 1, The above collecting steps are: A step of presenting a first non-motor ability evaluation mission to a user and collecting the performance results of the first non-motor ability evaluation mission from the user, The step of collecting the results of the above first non-exercise ability evaluation mission is: Including a step of presenting the 7th mission and collecting the results of performing the 7th mission, The above 7th mission includes an action that provides a questionnaire to measure the user's non-motor ability status on the screen and asks the user to answer the questionnaire. How to diagnose Parkinson's disease.

6. In paragraph 1, The above collecting steps are: A step of presenting a second non-motor ability evaluation mission to a user and collecting the performance results of the second non-motor ability evaluation mission from the user, The step of collecting the results of the above second non-exercise ability evaluation mission is: Including a step of presenting the 8th mission and collecting the results of performing the 8th mission, The above 8th mission includes an action of presenting a first area on the screen where guide colors are presented, a color selection palette in which a color can be selected by moving a fourth object, a second area in which a color selected by moving the fourth object on the color selection palette is displayed, and moving the fourth object on the color selection palette so that a color matching the first area is displayed in the second area. How to diagnose Parkinson's disease.

7. In paragraph 1, The above extraction step is, A step of extracting one or more first motor ability features from the performance results of one or more first motor ability evaluation missions among the performance results of one or more motor ability evaluation missions, The step of extracting one or more first motor ability features comprises: comprising a step of extracting one or more features from the first to fifth features, The first feature includes statistical analysis indices and frequency domain analysis indices for the first time series data collected from the results of performing the first mission of bending and stretching the first object presented on the screen using a finger for a preset period of time, and the number of times and speed of bending and stretching the first object using a finger. The second feature includes statistical analysis indices and frequency domain analysis indices for the second time series data collected from the results of performing the second mission of repeatedly moving the second object presented on the screen from the top to the bottom at the highest speed while touching it for a preset period of time, and the number of times the second object was repeatedly moved from the top to the bottom while touching it. The third feature includes statistical analysis indices and frequency domain analysis indices for third time series data collected from the results of performing the third mission of alternately touching two third objects presented on the screen at the highest speed using one finger for a preset period of time, and the number of times the two third objects were alternately touched using one finger and the touch speed. The fourth feature includes statistical analysis indices and frequency domain analysis indices for the fourth time series data collected from the results of the fourth mission of tracing the first shape guide pattern presented on the screen using a finger, and a first curvature calculated in response to the results of tracing the first shape guide pattern using a finger. The fifth feature includes statistical analysis indices and frequency domain analysis indices for fifth time series data collected from the results of performing the fifth mission to trace a second shape guide pattern different from the first shape presented on the screen using a finger, and a second curvature calculated in response to the results of tracing the second shape guide pattern using a finger. How to diagnose Parkinson's disease.

8. In paragraph 1, The above extraction step is, A step of extracting a second motor ability feature from the performance results of a second motor ability evaluation mission among the performance results of one or more of the above motor ability evaluation missions, The step of extracting the above second exercise ability feature is: Step 6-1 includes extracting features, The above 6-1 feature includes at least one of the fundamental frequency, shimmer, harmonic to noise ratio (HNR), loudness, and speech rate of the audio signal included in the speech result collected from the performance result of the 6th mission to speak the script presented on the screen. How to diagnose Parkinson's disease.

9. In paragraph 1, The above extraction step is, A step of extracting a second motor ability feature from the performance results of a second motor ability evaluation mission among the performance results of one or more of the above motor ability evaluation missions, The step of extracting the above second exercise ability feature is: Includes a step of extracting features 6-2, The step of extracting the above 6-2 feature is: A step of generating a preprocessed audio signal by converting the audio signal included in the speech result collected from the performance result of the 6th mission to speak the script presented on the screen, removing noise, or separating the audio signal by speaker; and A step of outputting a high-dimensional feature vector corresponding to the preprocessed audio signal using a pre-trained self-supervised learning-based deep neural network model that takes an audio signal as input and outputs a high-dimensional feature vector for the audio signal; and A step of determining the output result of the high-dimensional feature vector corresponding to the above preprocessed audio signal as the 6-2 feature, How to diagnose Parkinson's disease.

10. In paragraph 1, The above extraction step is, A step of extracting a first non-motor ability feature from the performance result of a first non-motor ability evaluation mission among the performance results of one or more non-motor ability evaluation missions, The step of extracting the above first non-motor ability feature is: Including a step of extracting the seventh feature, The seventh feature includes a questionnaire for measuring the user's non-motor ability status on the screen and answers to the questionnaire collected from the results of the seventh mission. How to diagnose Parkinson's disease.

11. In paragraph 1, The above extraction step is, A step of extracting a second non-motor ability feature from the performance results of a second non-motor ability evaluation mission among the performance results of one or more non-motor ability evaluation missions, The step of extracting the second non-motor ability feature is: Including a step of extracting the 8th feature, The eighth feature comprises a first area where a guide color is presented on the screen, a color selection palette for selecting an arbitrary color by moving a fourth object, a second area where a color selected by moving the fourth object on the color selection palette is displayed, and a color error including a difference between the guide color presented in the first area and the color displayed in the second area, which is calculated based on data collected from the results of performing an eighth mission of moving the fourth object on the color selection palette so that a color matching the first area is displayed in the second area, and a root mean square error (RMSE) calculated based on the color error. How to diagnose Parkinson's disease.

12. In paragraph 1, The above diagnosing steps are: A method for diagnosing the presence or absence of Parkinson's disease in a user corresponding to information about the user and at least one of the results of extracting one or more motor ability features and one or more non-motor ability features, comprising: using a first diagnostic model for diagnosing the presence or absence of Parkinson's disease based on information about the user and at least one of the results of extracting one or more motor ability features and one or more non-motor ability features; The above first diagnostic model is, A machine learning model trained in a supervised learning manner by training data that inputs information about the user, at least one of one or more motor ability features and one or more non-motor ability features, and labels the presence or absence of Parkinson's disease. How to diagnose Parkinson's disease.

13. In paragraph 1, The above diagnosing steps are: A step of diagnosing the presence or absence of Parkinson's disease of a user corresponding to information about the user and at least one of the results of extracting one or more motor ability features and one or more non-motor ability features, using a second diagnostic model that diagnoses the presence or absence of Parkinson's disease based on information about the user and at least one of the results of extracting one or more motor ability features and one or more non-motor ability features, The above second diagnostic model is, A deep learning model trained in a supervised learning manner by training data that inputs information about the user, at least one of one or more motor ability features and one or more non-motor ability features, and labels the presence or absence of Parkinson's disease. How to diagnose Parkinson's disease.

14. In paragraph 13, The above diagnosing steps are: A step of generating a diagnostic result including the presence or absence of Parkinson's disease based on a combined result of extracting at least one of the one or more motor ability features and the one or more non-motor ability features through a first FC (fully connected) layer and information about the user. How to diagnose Parkinson's disease.

15. In paragraph 1, The above diagnosing steps are: A step of diagnosing the presence or absence of Parkinson's disease of the user corresponding to at least one of the information about the user, the results of extracting at least one motor ability feature and at least one non-motor ability feature, and at least one of a fourth image generated based on a performance result of a fourth mission included in the motor ability feature and a fifth image generated based on a performance result of a fifth mission included in the motor ability feature, using a third diagnostic model that diagnoses the presence or absence of Parkinson's disease, the third diagnostic model comprising: information about the user, at least one of the results of extracting at least one motor ability feature and at least one non-motor ability feature, and at least one of a fourth image generated based on a performance result of a fourth mission included in the motor ability feature and a fifth image generated based on a performance result of a fifth mission included in the motor ability feature; The third diagnostic model above is, A deep learning model that is trained in a supervised learning manner by training data that has information about a user, at least one of one or more motor ability features and one or more non-motor ability features, at least one of a fourth image generated based on the performance result of a fourth mission included in the motor ability features, and a fifth image generated based on the performance result of a fifth mission included in the motor ability features as input, and has Parkinson's disease as a label. How to diagnose Parkinson's disease.

16. In paragraph 15, The above diagnosing steps are: A step of converting at least one of the fourth image and the fifth image into a first one-dimensional vector through a first CNN (convolutional neural network) layer; and A step of generating a diagnostic result including the presence or absence of Parkinson's disease based on a combined result of extracting at least one of the one or more motor ability features and the one or more non-motor ability features through a second FC layer, the first one-dimensional vector, and information about the user. How to diagnose Parkinson's disease.

17. In paragraph 1, The above diagnosing steps are: A step of diagnosing the presence or absence of Parkinson's disease of the user corresponding to the sixth image generated based on the information about the user, the results of extracting one or more motor ability features and one or more non-motor ability features, and the results of performing the sixth mission included in the motor ability features, using a fourth diagnostic model that diagnoses the presence or absence of Parkinson's disease based on information about the user, at least one of the results of extracting one or more motor ability features and one or more non-motor ability features, and the results of performing the sixth mission included in the motor ability features, The fourth diagnostic model above is, A deep learning model trained in a supervised learning manner by training data having as input information about the user, at least one of one or more motor ability features and one or more non-motor ability features, and a sixth image generated based on the performance result of the sixth mission included in the motor ability features, and labeled with the presence or absence of Parkinson's disease. How to diagnose Parkinson's disease.

18. In paragraph 17, The above diagnosing steps are: A step of converting the sixth image into a second one-dimensional vector through a second CNN layer; and A step of generating a diagnostic result including the presence or absence of Parkinson's disease based on a combined result of extracting at least one of the one or more motor ability features and the one or more non-motor ability features through a third FC layer, the second one-dimensional vector, and information about the user. How to diagnose Parkinson's disease.

19. In paragraph 1, The above diagnosing steps are: A step of diagnosing the presence or absence of Parkinson's disease of the user corresponding to at least one of the results of extracting one or more motor ability features and one or more non-motor ability features, a fourth image generated based on a performance result of a fourth mission included in the motor ability features, a fifth image generated based on a performance result of a fifth mission included in the motor ability features, and a sixth image generated based on a performance result of a sixth mission included in the motor ability features, using a fifth diagnostic model that diagnoses the presence or absence of Parkinson's disease, the fifth diagnostic model comprising: information about the user; at least one of the results of extracting one or more motor ability features and one or more non-motor ability features; at least one of the fourth image generated based on a performance result of the fourth mission included in the motor ability features and a fifth image generated based on a performance result of the fifth mission included in the motor ability features; and a sixth image generated based on a performance result of the sixth mission included in the motor ability features, The fifth diagnostic model above is, A deep learning model that is trained in a supervised learning manner by training data that has as input information about a user, at least one of one or more motor ability features and one or more non-motor ability features, a fourth image generated based on the performance result of a fourth mission included in the motor ability features, at least one of a fifth image generated based on the performance result of a fifth mission included in the motor ability features, and a sixth image generated based on the performance result of a sixth mission included in the motor ability features, and has Parkinson's disease as a label. How to diagnose Parkinson's disease.

20. In paragraph 19, The above diagnosing steps are: A step of converting at least one of the fourth image and the fifth image into a first one-dimensional vector through a first CNN layer; A step of converting the sixth image into a second one-dimensional vector through a second CNN layer; and A step of generating a diagnostic result including the presence or absence of Parkinson's disease based on a result of combining at least one of the results of extracting the one or more motor ability features and the one or more non-motor ability features, the first one-dimensional vector, the second one-dimensional vector, and information about the user through the fourth FC layer. How to diagnose Parkinson's disease.

21. A computer-readable recording medium storing a computer program for executing the method of claim 1 using a computer.

22. As a device for diagnosing Parkinson's disease, processor; and A memory operably connected to the processor and storing at least one code to be executed by the processor, The above memory, when executed through the processor, causes the processor to present one or more motor ability evaluation missions and one or more non-motor ability evaluation missions to the user, and collect the performance results for the one or more motor ability evaluation missions and the performance results for the one or more non-motor ability evaluation missions from the user. Extracting one or more motor ability features and one or more non-motor ability features from the performance results of one or more motor ability evaluation missions and the performance results of one or more non-motor ability evaluation missions, Storing a code that causes the user to diagnose whether or not the user has Parkinson's disease based on at least one of the results of extracting the one or more motor ability features and the one or more non-motor ability features. Parkinson's disease diagnostic device.

Images