A digital biomarker-based data processing system and method

Abstract

This invention relates to a data processing system and method based on digital biomarkers. The system includes at least a digital chessboard and a server. The digital chessboard is configured to collect motion data from the chessboard when a user uses it. The server is configured to acquire the motion data, calculate and analyze digital biomarkers corresponding to reference biomarkers based on the motion data, and assess / screen the user's brain health based on the digital biomarkers.

Description

Technical Field

[0001] This invention relates to the field of digital medical technology, and in particular to a data processing system and method based on digital biomarkers. Background Technology

[0002] Cognition is the process by which the brain receives and processes information from the outside world to actively understand the world. Cognitive functions involve multiple areas such as memory, attention, language, executive function, reasoning, calculation, and orientation. Cognitive impairment refers to damage to one or more of these areas, affecting patients' social functioning and quality of life to varying degrees, and in severe cases, even leading to death. Cognitive impairment is not only a purely medical problem but also a serious social one. Neurodegenerative diseases, cardiovascular and cerebrovascular diseases, nutritional and metabolic disorders (especially diabetes), infections, trauma, tumors, and drug abuse are among the many causes that can lead to cognitive impairment. Currently, the assessment of cognitive abilities in individuals mainly relies on cognitive impairment screening scales, where staff evaluate individuals to assess their cognitive abilities. Current digital assessment scales (such as the Savonix Digital Recognition Scale) and localized versions of these quantitative assessment scales in China typically embed the corresponding assessment software into mobile app applications, assessing the user's interaction with the application through gameplay. After careful research and evaluation of existing technologies related to cognitive assessment, the inventors found the following technical shortcomings: a. Traditional scales and cognitive assessments are time-consuming and labor-intensive, data collection is difficult, and the economic cost of testing is high. Furthermore, the elderly (especially those with lower levels of education) have low acceptance of this type of testing or find it difficult to use; b. Traditional cognitive assessments rely on the elderly seeking medical help independently, making the assessment method relatively passive; in addition, traditional scales monitor a limited number of data categories (for example, traditional scale assessments can only judge whether the test subject draws a clock correctly, the position of the hands and the outline of the clock are correct, and whether the numerical expression is correct, etc.); c. Traditional data analysis only focuses on the mining of correlations.

[0003] For example, Chinese patent document CN106327049A discloses a cognitive assessment system, including an information module, a testing module, and an analysis module. The information module is used to obtain medical information matching the testing module based on the subject's data, establishing a complete cognitive assessment database. The testing module obtains the subject's cognitive test data through testing, including the following five sub-modules: attention and executive function testing module, memory testing module, mathematical and calculation ability testing module, language testing module, and action and behavior control and planning testing module. The analysis module determines the subject's cognitive assessment results based on the medical information obtained by the information module and the cognitive test data obtained by the testing module. However, this invention still requires staff to perform measurements to achieve cognitive assessment, which places high demands on staff and is difficult to automate. Moreover, the decline in cognitive ability in an individual is a slow and imperceptible process. Cognitive ability scores obtained through one or two cognitive impairment screening scale assessments cannot accurately reflect changes in an individual's cognitive ability. As mentioned earlier, due to the time-consuming nature of the cognitive impairment screening scale assessment process, the high requirements for staff, and the difficulty for individuals with low levels of education to complete most of the assessment content, it is difficult to frequently obtain changes in cognitive ability through cognitive impairment screening scale assessments. Therefore, it is necessary to improve existing technologies.

[0004] In recent years, smart bracelets, Apple Watches, smart mattresses, pocket electrocardiographs, and other digital health and medical devices have sprung up like mushrooms after rain, many of which have already entered ordinary households. Besides helping people understand their health status more conveniently, these devices continuously collect health data, which, through the mobile internet, aggregates into astronomical health and medical data resources. Coupled with appropriate analytical methods, this can generate new insights to reveal the current state and development trends of the physical and mental health of groups, especially individuals. The resulting digital biomarkers hold the promise of becoming an effective means of gaining a deeper understanding of human health and disease. Simply put, digital biomarkers are objective data about an individual's physiology and behavior collected by users / consumers through interconnected digital health devices, used to explain, influence, and predict health outcomes. Traditional biomarkers, on the other hand, generally refer to indicators obtained through biochemical tests, used to mark changes in the structure or function of organs and other tissues. For example, blood tests in traditional hospitals can produce insightful data, but because they are obtained through biochemical tests rather than interconnected digital health devices, they are not digital biomarkers. On the one hand, the current development goal of digital biomarkers is to effectively complement, rather than replace, traditional biomarkers; on the other hand, digital biomarkers can powerfully drive the shift in healthcare models from reactive response to proactive prevention. By using digital biomarkers, researchers can not only better interpret diseases, but also leverage the ever-growing volume of health data to analyze the meaning of normal and healthy individuals' states, and more importantly, predict future health outcomes. Therefore, interest in digital biomarker research is expected to surge in the coming years.

[0005] Furthermore, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the applicant studied a large number of documents and patents when making this invention, but due to space limitations, not all details and contents were listed in detail. However, this does not mean that the present invention does not possess the features of these prior art. On the contrary, the present invention already possesses all the features of the prior art, and the applicant reserves the right to add relevant prior art to the background art. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a data processing system based on digital biomarkers. The system for screening brain health in a population includes at least:

[0007] The digital chess and card game system can at least collect motion data of the digital chess and card game when a user uses it. Preferably, the motion data includes at least the spatial coordinate data of the digital chess and card game during its use / movement by the user. Preferably, the motion data may also include the magnitude / direction of acceleration, the rotational angular velocity / angular acceleration along each coordinate axis, etc., during the use / movement of the digital chess and card game.

[0008] The server is at least able to acquire the motion data, calculate and analyze the motion data to derive digital biomarkers corresponding to the reference markers, and assess / screen the user's brain health based on the digital biomarkers.

[0009] When this system is used in a community card and board game entertainment venue, there are likely to be multiple users using the digital card and board games simultaneously (e.g., multiple users simultaneously arranging the cards, flipping them in their palms, or the same digital mahjong tile being moved or picked up by multiple different users at the same time). Each time a digital card and board game is used / moved, it must send motion data to the digital evaluation unit and / or server. However, since each digital card and board game can be used by multiple different users within a given time period, the intelligent trajectory sensor or server cannot accurately and quickly attribute the motion data of each digital card and board game in use / movement to the user's action database. In other words, the attribution problem of the motion trajectory and other data recorded by the digital card and board game remains unresolved. Therefore, equipping users of this system with wearable devices, which can at least send identifiers that identify the user to the intelligent trajectory analysis sensor and / or server within the digital card and board game, solves the attribution problem of the motion trajectory and other data recorded by the digital card and board game.

[0010] Preferably, the intelligent trajectory analysis sensor is capable of acquiring the identifier of the wearable device. Preferably, the intelligent trajectory analysis sensor is capable of measuring the distance between itself and the wearable device. Particularly preferably, the intelligent trajectory analysis sensor within each digital chessboard measures the distance to wearable devices surrounding the digital chessboard at a second moment after being moved by the user, and sends a data attribution request to the wearable device closest to the digital chessboard. Preferably, upon receiving the data attribution request, the wearable device closest to the digital chessboard sends its carried identifier to the intelligent trajectory analysis sensor to identify the user's identity information using the wearable device.

[0011] Preferably, the wearable device can manually input the user's identity information. Preferably, the user's identity information may include, but is not limited to, the user's name, age, gender, etc. Simultaneously, the wearable device can at least measure the distance between the moving digital chessboard and the wearable device.

[0012] Preferably, the intelligent trajectory analysis sensor can automatically time and record the time when the intelligent trajectory analysis sensor or the board moves / vibrates. Preferably, the motion data collected by the intelligent trajectory analysis sensor can be appended with a timestamp corresponding to the motion data.

[0013] Preferably, the first moment can be determined by the intelligent trajectory analysis sensor. Preferably, the intelligent trajectory analysis sensor determines the first moment when the vibration value measured by the intelligent trajectory analysis sensor and / or the card changes from zero to a non-zero value.

[0014] Particularly preferably, when the vibration value measured by the intelligent trajectory analysis sensor and / or the card game changes from zero to a non-zero value, and the height of the intelligent trajectory analysis sensor deviates from a first height, the intelligent trajectory analysis sensor determines this moment as the first instant. Preferably, the first height is the height of the digital card game platform. Preferably, the digital card game platform can be a mahjong table, etc. With this setting, the intelligent trajectory analysis sensor installed in the digital card game will only start recording its spatial coordinate information when moved by the user and only when moved outside the digital card game platform. This allows the intelligent trajectory analysis sensor to record a complete card-drawing action of the user (especially recording the pause and thinking process from drawing a card from the table to putting the card back on the table and the complete card-drawing trajectory). In other words, it avoids the intelligent trajectory analysis sensor breaking down a complete card-drawing action into multiple sub-parts for recording when the user is paused in mid-air, which could easily lead to the loss of potentially important digital biomarkers such as the pause / stagnation time in mid-air and the complete card-drawing trajectory.

[0015] Preferably, the second moment can be determined by the intelligent trajectory analysis sensor. Preferably, the intelligent trajectory analysis sensor determines the second moment when the vibration value measured by the intelligent trajectory analysis sensor and / or the card changes from a non-zero value to a zero value.

[0016] Particularly preferably, the intelligent trajectory analysis sensor acquires the identifier of the wearable device closest to the intelligent trajectory analysis sensor when and only when the intelligent trajectory analysis sensor in the digital chess and card game begins to search for the wearable devices around the intelligent trajectory analysis sensor at the second moment.

[0017] This configuration avoids situations where a single movement / motion of a digital chess piece corresponds to multiple users / wearable devices, as this could occur when measuring the distance between the intelligent trajectory analysis sensor and surrounding wearable devices at the first moment or between the first and second moments. This is because during the use of a digital chess piece, especially at the moment of use / movement (e.g., the second moment), the digital chess piece is located close to the user, maintaining a relatively large distance from other users' wearable devices. This prevents the intelligent trajectory analysis sensor from identifying the same minimum distance to multiple users' wearable devices, thus avoiding situations where a single movement / motion of a digital chess piece corresponds to multiple users / wearable devices. For example, when multiple users in the same group simultaneously approach / move one / several digital chess pieces, one / several digital chess pieces may measure the same minimum distance to multiple surrounding wearable devices. In this technical solution, the intelligent trajectory analysis sensor measures the distance between itself and surrounding wearable devices at a second moment. That is, when a digital card is used / moved, the digital card is already located close to the user who used it, thus maintaining a relatively large distance from other users' wearable devices. This results in a significant difference in the distance between the user who used the digital card, the user who did not use the digital card, and the digital card itself, thereby improving the accuracy of the attribution of spatial coordinate data collected by the digital card each time.

[0018] For example, the smart trajectory analysis sensor begins measuring its distance from surrounding wearable devices and sends an identity request to the wearable device closest to the digital game board. The wearable device closest to the digital game board then sends an identifier to the smart trajectory analysis sensor within the digital game board, identifying the user wearing the wearable device.

[0019] When the intelligent trajectory analysis sensor receives the identifier of the wearable device closest to the digital chessboard, the intelligent trajectory analysis sensor inserts the identifier into the motion data that the intelligent trajectory analysis sensor sends to the digital evaluation unit and / or server for that time, so as to identify which user the motion data collected at that time came from.

[0020] Particularly preferably, the intelligent trajectory analysis sensor can only measure the distance to wearable devices within a distance threshold range. Preferably, the distance threshold can be manually set according to the actual scenario, for example, a distance threshold of fifteen centimeters. This configuration significantly reduces the distance measurement requirements of the intelligent trajectory analysis sensor, thereby significantly reducing the amount of data collected by the sensor (e.g., distance to several surrounding wearable devices). This allows the intelligent trajectory analysis sensor to more quickly identify the wearable device closest to it in the current trajectory analysis, ultimately obtaining the user's identity information associated with that wearable device.

[0021] Preferably, the identifier can be composed of characters, numbers, etc. For example, the identifier can be A1, A2, A3, A4, where a group of users use the same set of digital chess and cards, the letter A can represent which group of users they belong to, and the number can identify the first user belonging to the corresponding user group.

[0022] Preferably, after the intelligent trajectory analysis sensor completes the transmission of motion data to the digital evaluation unit and / or cloud server via the wireless gateway, the intelligent trajectory analysis sensor automatically clears the recorded identifier, so that the intelligent trajectory analysis sensor can receive an identifier from the wearable device to identify the user's identity information during the next use / movement by the user.

[0023] The intelligent trajectory analysis sensor sends the motion data for the current movement to the digital evaluation unit and / or server.

[0024] When the digital evaluation unit and / or server receive the motion data, they first identify the identifier in the current motion data to identify the data source of the current motion data, and then enter the current motion data into the database of the user corresponding to the identifier in the motion data, thereby solving the problem of data ownership for each digital chess and card game.

[0025] Preferably, the motion data collected by the intelligent trajectory analysis sensor installed in the digital chess and card game can also be sent to the digital evaluation unit and / or server, or sent to the digital evaluation unit and / or server through the wearable device.

[0026] This configuration makes the screening / assessment of users' brain health (such as cognitive abilities) based on digital chess and card games more acceptable to the elderly population. This allows the system to be deployed in multiple communities as a preliminary screening tool for assessing the brain health (such as cognitive abilities) of elderly users, significantly reducing the economic and social costs of screening / assessing brain health (such as cognitive abilities) and avoiding the waste of medical resources. Compared to cognitive testing and assessment solely through an app, the digital chess and card games and digital assessment units provide more data dimensions from users. For example, an app only obtains two-dimensional data, while the intelligent trajectory analysis sensor within the digital chess and card games provides three-dimensional or even more dimensional data. This invention can assist in shifting the diagnostic threshold for brain health (such as cognitive impairment patients, especially Alzheimer's patients) earlier, serving as a simpler and more objective risk assessment (screening) tool for Alzheimer's disease, rather than a diagnostic tool for cognitive impairment.

[0027] Particularly preferably, the intelligent trajectory analysis sensor can also measure the vibration value of itself and / or the game board. Preferably, the intelligent trajectory analysis sensor can begin recording spatial coordinate data when the intelligent trajectory analysis sensor or the game board vibrates / moves. Preferably, the intelligent trajectory analysis sensor can stop recording spatial coordinate data when the intelligent trajectory analysis sensor or the game board stops vibrating / moving. Particularly preferably, the intelligent trajectory analysis sensor can begin recording spatial coordinate data when the intelligent trajectory analysis sensor or the game board vibrates / moves and the intelligent trajectory analysis sensor is not at a first height, where the first height is the height of the game board platform used by the user. For example, the game board platform can be a mahjong table, etc. This configuration avoids the intelligent trajectory analysis sensor within the digital mahjong set collecting data during non-independent user activities such as shuffling. Such data (e.g., when the digital mahjong set is automatically shuffled, or when multiple users manually shuffle the set) cannot be used as raw data for individual user behavior, thus further reducing the collection of invalid raw data, improving the quality of raw data, and reducing the server's data processing load. Preferably, the spatial coordinate data collected by the intelligent trajectory analysis sensor has a maximum accuracy of at least five millimeters. Preferably, the recording time for the spatial coordinate data collected by the intelligent trajectory analysis sensor can be up to ten seconds.

[0028] According to a preferred embodiment, the digital board game includes at least a board game and an intelligent trajectory analysis sensor disposed within or integrated into the board game. The intelligent trajectory analysis sensor is at least capable of acquiring the spatial coordinates of the board game during use by the user and / or a timestamp associated with those spatial coordinates.

[0029] According to a preferred embodiment, the user is capable of wearing a wearable device. The wearable device is configured to at least send an identifier capable of identifying the user's identity information to the user. The intelligent trajectory analysis sensor is capable of acquiring the identifier of the wearable device. Preferably, the wearable device can be worn on the user's wrist. Preferably, the wearable device can also be worn on other parts of the user's body.

[0030] According to a preferred embodiment, the wearable device can also collect or have input the user's individual characteristic data and / or physiological data, including at least heart rate, blood pressure, pulse oxygen saturation, body temperature, and associated timestamps. The individual characteristic data includes at least familiarity with card games, education level, and medical history.

[0031] The wearable device can transmit individual characteristic data and / or physiological data to the digital assessment unit or the server. Preferably, the physiological data may also include information related to user activity, such as the number of steps the user takes each day. Preferably, the physiological data may also include the frequency and amplitude of shaking of user body parts (e.g., hands). For example, the frequency and amplitude of shaking of user body parts (e.g., hands) can be obtained by an accelerometer within the wearable device.

[0032] Particularly preferably, the wearable device is capable of collecting at least the hand tremor data of the user wearing the device. Through this configuration, the server can combine / integrate the data collected by the wearable device with the motion data from digital chess and card games to jointly assess the cognitive abilities and / or other conditions (such as brain diseases, depression, schizophrenia, etc.) of multiple users.

[0033] According to a preferred embodiment, a digital assessment unit may also be included. The digital assessment unit is configured to monitor the interaction process between the user and the digital assessment unit to capture the user's interaction data, language data, eye-tracking data, and / or pupil data. The digital assessment unit can send the interaction data, language data, eye-tracking data, and / or pupil data of the aforementioned interaction process to the server, so that the server can calculate the user's assessment data based on the interaction data, language data, eye-tracking data, and / or pupil data of the interaction process.

[0034] Preferably, the interactive data is the data generated by the user during the interaction with the digital evaluation unit 1, such as including but not limited to click data, frequency data, continuous action data, trajectory data, information content selection data, swipe selection data, content change logic relationship data, etc.

[0035] Preferably, the digital assessment unit can identify and extract speech and text information from the language data used. Through the extraction of language data, it is possible to identify the user's emotional information, sentiment information, and so on.

[0036] Graphic information refers to text and image information, language and / or images or text used for communication. For example, ordinary users commonly use emoticons and emojis to express language and emotions in communication. Graphic information can also directly or indirectly extract users' emotional information, etc.

[0037] Speech information includes, for example, speech patterns in voice commands, words, and phrases. Preferably, after signal processing of the speech, the digital evaluation unit can extract phonemes and phoneme repetitions. Preferably, the digital evaluation unit can identify and extract features of each speech pattern, including pitch, amplitude, and spectrum.

[0038] Preferably, the eye-tracking data is data related to the user's eye movements. Preferably, the eye-tracking data may include, but is not limited to, data related to eye movements (or eye movement patterns) such as fixation, saccades, and tracking. Preferably, the eye-tracking data may also include: the user's fixation point position, fixation time, and pupil diameter. Preferably, the above-mentioned eye-tracking data can be obtained by monitoring the user's eye behavior using an electrooculogram (EOG) sensor or other related eye-tracking sensors. Pupil data includes, but is not limited to, pupil size data and pupil change data.

[0039] According to a preferred embodiment, the digital chess and card game can send the motion data to the digital evaluation unit and / or the server. The digital evaluation unit can acquire the motion data and send it to the server.

[0040] According to a preferred embodiment, it may also include a smart environment device unit. The smart environment device unit is capable of acquiring daily data from the user's daily life.

[0041] According to a preferred embodiment, the server is at least able to adjust the type / variety of the raw data collected by the digital chess and card game, wearable device, digital evaluation unit, and smart environment device unit based on the reference marker.

[0042] Preferably, the reference biomarker is a digital biomarker capable of effectively representing a user's target attribute of brain health. Preferably, the target attribute includes, but is not limited to, the ability to perform all cognitive processes. Preferably, the target attribute may also include the ability to perform mental processes. Preferably, the target attribute may further include the abilities of learning and judgment, language, and memory. For example, the reference biomarker can be a digital biomarker capable of effectively representing changes in a user's cognitive abilities or state. Preferably, the reference biomarker can also be a digital biomarker capable of effectively representing other levels of a user's physical health.

[0043] Digital biomarkers are not limited to a single data category; they can originate from one or more of the following: interaction data, language data, eye-tracking data, and everyday data.

[0044] For example, if the server analysis determines that the reference marker for cognitive decline (i.e., the effective digital biomarker) is the pause time and frequency of the user's hand when moving the digital chess piece, then the server can adjust the type of motion data of the user's hand collected by the digital chess piece based on this reference marker (i.e., the pause time and frequency of the user's hand). That is, the intelligent trajectory analysis sensor only needs to measure / collect the vibration value of the digital chess piece itself between the first moment and the second moment and the associated timestamp to calculate the digital biomarker (such as the pause time and frequency of the user's hand), without measuring / collecting the spatial coordinate data of the user's hand moving the digital chess piece. Alternatively, the intelligent trajectory analysis sensor can also collect the spatial coordinate data of the digital chess piece, but the intelligent trajectory analysis sensor may not send the collected spatial coordinate data to the digital evaluation unit or the server.

[0045] For example, when screening brain health in a population, if the server analysis finds that reference markers of cognitive decline also include the user's heart rate when using digital chess and cards, wearable devices (such as smart bracelets) only need to focus on measuring the user's heart rate data when using digital chess and cards.

[0046] For example, when screening brain health in a population, if the server analysis indicates that reference markers of cognitive decline include the frequency of forgetfulness, the smart environmental device unit only needs, or at least a water immersion sensor, to monitor the frequency of users forgetting to turn off the tap. This configuration allows the server to adjust the types and / or quantities of raw data collected by digital chess, digital assessment units, wearable devices, and smart environmental device units based on effective digital biomarkers inferred from causal analysis theory. This reduces the server's data processing load, increasing the speed at which the server processes health data collected from various sensors / devices in real time. This allows for faster analysis and feedback of relevant conclusions (e.g., inference of effective digital biomarkers) to doctors, consumers, researchers, and other relevant institutions. Ultimately, this enables the health data collected by digital chess, digital assessment units, wearable devices, and smart environmental device units to generate greater value, such as in large-scale brain health screenings in cities or communities, to identify users with abnormal brain health (e.g., cognitive decline or signs of mild cognitive impairment) early, and to promptly report these abnormalities to relevant medical personnel.

[0047] This invention also provides a data processing method based on digital biomarkers. The method is as follows:

[0048] The digital chess and card game collects the movement data of the digital chess and card game when the user uses it.

[0049] The server acquires the motion data collected by the digital chess and card game, and calculates and analyzes the motion data to obtain digital biomarkers with the same data type as the reference biomarkers. Based on the digital biomarkers, the server assesses / screens the user's brain health.

[0050] For example, the server can access previously recorded or analyzed valid reference biomarkers and compare the differences between the digital biomarker calculated by the server and the reference biomarkers or previously valid digital biomarkers, and assess / screen the user's brain health based on these differences. For example, the server can assess or screen users with declining cognitive abilities within a population based on these differences, or predict the user's trend of change within a specified future timeframe.

[0051] According to a preferred embodiment, the reference biomarker is obtained by the following steps: acquiring raw data; extracting feature data from the raw data based on machine learning, and further screening candidate digital biomarkers from the feature data; and performing causal inference from the candidate digital biomarkers based on causal learning to calculate and analyze the reference biomarker.

[0052] Preferably, the raw data is obtained through the following steps: a digital chess and card game collects motion data of the digital chess and card game (i.e., data required for reference to digital biomarkers) when the user uses the digital chess and card game; a wearable device collects other physiological data of the user; a digital assessment unit acquires the user's assessment data, language data, and eye-tracking data; a smart environment device unit collects daily data from the user's daily life; and a server acquires one or more of the motion data, other physiological data, assessment data, language data, eye-tracking data, and daily data as raw data for server analysis. Attached Figure Description

[0053] Figure 1 This is a simplified schematic diagram of the module connection relationship of a preferred embodiment provided by the present invention;

[0054] Figure 2 This is a schematic diagram of a preferred embodiment of the digital chess and card game provided by the present invention;

[0055] Figure 3 This is a schematic diagram of another preferred embodiment of the digital chess and card game and digital evaluation unit provided by the present invention.

[0056] List of reference numerals

[0057] 1. Digital chess and card games; 2. Servers; 3. Wearable devices;

[0058] 4: Digital assessment unit; 5: Intelligent environmental equipment unit; 101: Chess and card games;

[0059] 102: Intelligent trajectory analysis sensor. Detailed Implementation

[0060] The following is a detailed explanation with reference to the accompanying drawings.

[0061] Figure 1 , Figure 2 and Figure 3 A data processing system based on digital biomarkers is illustrated. The system includes at least a digital chessboard 1 and a server 2. Preferably, the digital chessboard 1 is configured to collect motion data of the digital chessboard 1 when a user uses it. Preferably, the server 2 is configured to acquire the motion data, calculate and analyze digital biomarkers corresponding to reference biomarkers based on the motion data, and assess / screen the user's brain health based on the digital biomarkers.

[0062] Preferably, server 2 can be a regular physical server 2. Preferably, server 2 can also be a cloud server 2.

[0063] Preferably, brain health includes at least cognitive abilities. Preferably, brain health may also include the abilities to learn, judge, speak, and remember, as well as the ability to perform mental processes. Preferably, brain health may also include other diseases.

[0064] According to a preferred embodiment, the digital chess and card game 1 includes at least: a chess and card game 101, and an intelligent trajectory analysis sensor 102 disposed or integrated within the chess and card game 101. The intelligent trajectory analysis sensor 102 is at least capable of acquiring the spatial coordinates and / or timestamps associated with the spatial coordinates during the process of the chess and card game 101 being used by the user.

[0065] When the digital chess and card game 1 is integrated into a mobile device, the evaluation unit is used to collect interaction data, language data, eye movement data and / or pupil data during the process of the chess and card game 101 being used by the user.

[0066] Preferably, the process of the user using the aforementioned chess and card game 101 includes at least: moving the chess and card game and rotating the chess and card game. Preferably, the process of the user using the aforementioned chess and card game may also include: the magnitude, direction, and frequency of the force applied to the chess and card game, and the magnitude and frequency of the force applied to the chess and card game when tapping it.

[0067] Particularly preferably, the board game 101 can be mahjong. Preferably, the board game 101 can also be other types of board games such as Chinese chess or international chess.

[0068] Preferably, the intelligent trajectory analysis sensor 102 is at least capable of acquiring the three-dimensional spatial coordinate data of the digital chessboard 1. Preferably, the NFC reader / writer is at least capable of configuring parameters for the intelligent trajectory analysis sensor 102. Preferably, the wireless gateway is used in conjunction with the intelligent trajectory analysis sensor 102. Preferably, the wireless charger is at least capable of charging the intelligent trajectory analysis sensor 102.

[0069] Preferably, the intelligent trajectory analysis sensor 102 is disposed within or integrated into the mahjong tile. Preferably, the intelligent trajectory analysis sensor 102 is configured to open the data channel for data transmission only when needed to connect with other devices such as a wireless gateway. Preferably, after the intelligent trajectory analysis sensor 102 is installed into the slot of the mahjong tile, the slot can be glued together using strong adhesive.

[0070] Preferably, the data transmitted by the intelligent trajectory analysis sensor 102 includes at least the three-dimensional spatial coordinate data of the digital chess and card game 1 (e.g., digital mahjong), that is, the coordinates of the digital chess and card game 1 corresponding to the x, y, and z axes respectively. Preferably, the coordinate system of the digital chess and card game 1 can be flexibly selected according to the actual application scenario.

[0071] Preferably, the NFC reader / writer supports the NFC protocol. Preferably, the NFC reader / writer can be used to configure parameters for the intelligent trajectory analysis sensor 102. Preferably, the NFC reader / writer can be connected to a PC. Preferably, the NFC reader / writer can also be used to read and write data acquired by the intelligent trajectory analysis sensor 102.

[0072] Preferably, the wireless charger can adopt the Qi standard wireless charging protocol. Preferably, the wireless charger can be used to charge the intelligent trajectory analysis sensor 102.

[0073] Preferably, the wireless gateway can be a wireless network, etc. Preferably, the wireless gateway can keep the data receiving channel always open, and can receive or transmit data sent by the intelligent trajectory analysis sensor 102 at any time. Preferably, a set of digital chess and card games 1 is equipped with a set of wireless gateways. Preferably, any one of the digital chess and card games 1 in a set of digital chess and card games 1 sends data only through the wireless gateway corresponding to that set of digital chess and card games 1.

[0074] Preferably, the intelligent trajectory analysis sensor 102 can be a triaxial accelerometer.

[0075] Preferably, the intelligent trajectory analysis sensor 102 does not need to maintain a long-term connection with the corresponding wireless gateway. The intelligent trajectory analysis sensor 102 can establish a data connection with the wireless gateway to open the data transmission channel when data needs to be sent.

[0076] Preferably, the intelligent trajectory analysis sensor 102 can be used to record the spatial trajectory data formed in mid-air and / or within a specific platform when a user grasps a digital chess and card game 1 integrated with or equipped with the intelligent trajectory analysis sensor 102. Preferably, the spatial trajectory data can include timing signals such as the duration of pauses during the user's grasping of the mahjong tile, the distance moved in a single action, and the movement trajectory of the user's hand from one position to another. Preferably, a single action is an action without any pauses throughout. Preferably, a single action includes pauses throughout.

[0077] Preferably, the working process of the intelligent trajectory analysis sensor 102 can also be divided into the following steps:

[0078] Once the intelligent trajectory analysis sensor 102 is activated (i.e., at the first moment), the intelligent trajectory analysis sensor 102 performs clock configuration and serial port configuration through STM32.

[0079] Set the transparent transmission mode via the 2.4G module;

[0080] 2.4G module sends data to wake up the sensor: If a sensor response signal is received, the intelligent trajectory analysis sensor 102 will go out of sleep; if no sensor response signal is received, the 2.4G module will resend data to wake up the sensor.

[0081] The intelligent trajectory analysis sensor 102 is initialized, that is, the time is reset to zero so that the reading time and x / y / z axis data are respectively zeroed, so as to read new coordinate data of x / y / z axis in subsequent processes;

[0082] When a user uses a digital chessboard 1 equipped with or integrated with an intelligent trajectory analysis sensor 102, the intelligent trajectory analysis sensor 102 calculates the output acceleration, angular velocity, and spatial coordinates of the digital chessboard 1.

[0083] The intelligent trajectory analysis sensor 102 performs data fusion, that is, it acquires the coordinate data of the x / y / z axes for the attitude judgment and setting of the attitude flag of the digital chess and card game 1, and acquires the identification code sent by the wearable device 3 or other devices to indicate the identity of the user currently using the digital chess and card game 1.

[0084] The intelligent trajectory analysis sensor 102 packages the aforementioned data fusion (e.g., the acceleration, angular velocity, and spatial coordinates of the digital chess piece 1, as well as an identification code for user identification). The packaged data can be configured with a checksum and sent to a gateway via a 2.4G module. The gateway then forwards the data to the cloud platform (i.e., cloud server 2). Alternatively, the intelligent trajectory analysis sensor 102 can send the data fusion and attitude determination results to a host computer (which can be a personal server 2) via a serial port. The host computer then reproduces all the data acquired by the intelligent trajectory analysis sensor 102 to create a spatial trajectory image of the currently used digital chess piece 1. Particularly preferably, this spatial trajectory image can also be sent to server 2 as a possible source of raw data for digital markers.

[0085] This configuration allows for the acquisition of more comprehensive data representing user decision-making processes and / or behavioral habits through the digital chess and card game 1 compared to simply relying on a mobile platform app. Specifically, the intelligent trajectory analysis sensor 102 collects data on the three-dimensional motion trajectory of the user's hand in mid-air and / or on a specific platform (e.g., the mahjong table) when using the digital chess and card game 1 (e.g., digital mahjong), including the force of the user's hand grasping the mahjong pieces, the frequency and intensity of tremors, the smoothness and continuity of the movements, pause times, and movement trajectories. This data can be transmitted to the server 2 via existing technologies such as Bluetooth, Wi-Fi, or wireless gateways. The actual process of users grabbing physical chess and card games such as mahjong is significantly different from simply touching a touchscreen with a game app. This is because a touchscreen with a game app can generally only obtain two-dimensional motion data of the user, while users generate three-dimensional spatial motion data, or even more dimensions of motion data, when actually grabbing physical chess and card games such as mahjong, resulting in richer data categories. In addition, obtaining user motion data through digital chess and card games is more acceptable to users.

[0086] Preferably, the intelligent trajectory analysis sensor 102 can transmit the acquired user action data to the digital evaluation unit 4 and / or server 2 on the mobile platform via existing technologies such as wireless networks.

[0087] According to a preferred embodiment, the user is capable of wearing a wearable device 3. The wearable device is configured to at least send an identifier to the user that identifies the user currently using the digital chess and card game 1. The intelligent trajectory analysis sensor 102 is capable of acquiring the identifier of the wearable device 3. Preferably, the intelligent trajectory analysis sensor 102 acquires the identifier of the wearable device 3 closest to it only when the intelligent trajectory analysis sensor 102 within the digital chess and card game 1 begins searching for the wearable devices 3 (around the intelligent trajectory analysis sensor 102) at the second moment.

[0088] Preferably, the wearable device 3 may include, but is not limited to, wristband smart devices, head-mounted smart devices, etc. For example, the wearable device 3 may be a smartwatch / bracelet or a smart helmet. Preferably, the wearable device 3 may be worn by the user 24 hours a day or during specific time periods.

[0089] Preferably, the wearable device 3 can be worn on the user's wrist. Preferably, the wearable device 3 can also be worn on other parts of the user's body.

[0090] According to a preferred embodiment, the wearable device 3 can also collect the user's physiological data. The physiological data includes at least heart rate, blood pressure, pulse oxygen saturation, body temperature, and associated timestamps. The wearable device 3 can send the physiological data to the server 2.

[0091] Preferably, the physiological data may also include information related to user activity, such as the number of steps a user takes each day.

[0092] Preferably, the physiological data may further include the frequency and amplitude of shaking of the user's body parts (e.g., hands). For example, the frequency and amplitude of shaking of the user's body parts (e.g., hands) can be obtained by an accelerometer within the wearable device 3.

[0093] According to a preferred embodiment, the system may further include an intelligent environment device unit 5, which is capable of acquiring daily data from the user's daily life. Preferably, the daily data refers to behavioral data collected by the intelligent environment device unit 5 from the user throughout the day or during specific time periods. Preferably, the behavioral data can be related to the user's cognitive ability assessment. Preferably, the behavioral data can be related to other physical health assessments of the user. Preferably, the intelligent environment device unit 5 may include, but is not limited to, the following devices: a water immersion sensor, an ultrasonic sensor, a voice recorder, and a microwave sensor. For example, the daily data may be behavioral data collected by a water immersion sensor in the toilet (e.g., monitoring the frequency of forgetting to turn off the tap) to indicate the user's memory status; an ultrasonic sensor located within the user's daily activity range can determine the user's distance, and a microwave sensor can determine the user's movement, thereby acquiring the user's distance and movement information to indicate the user's orientation and visuospatial ability; the user's voice information recorded by a voice recorder located within the user's daily activity range can be used to determine the user's attention based on its speech frequency, speech rate, and voice information, and can also assess the user's immediate or delayed verbal memory based on information such as repeated sentences and keywords. Preferably, daily data may also include risk factors and physiological indicators. Preferably, risk factors may include, but are not limited to, the frequency of smoking and the frequency of alcohol consumption. Preferably, physiological indicators may include, but are not limited to, the health status of the heart and the health status of the brain. Preferably, daily data may also include other motor data required for cognitive ability assessment.

[0094] According to a preferred embodiment, a digital assessment unit 4 may also be included. The digital assessment unit 4 is configured to monitor the interaction process between the user and the digital assessment unit 4 to capture the user's interaction data, language data, eye-tracking data, and / or pupil data. The digital assessment unit 4 can send the interaction data, language data, eye-tracking data, and / or pupil data to the server 2, so that the server 2 can calculate the user's assessment data based on the interaction process.

[0095] Preferably, the digitization evaluation unit 4 can identify and extract speech patterns in the used speech commands, words, and phrases. Preferably, after signal processing of the speech, the digitization evaluation unit 4 can extract phonemes and phoneme repetitions. Preferably, the digitization evaluation unit 4 can identify and extract features of each speech pattern, including pitch, amplitude, and spectrum.

[0096] Preferably, the eye-tracking data is data related to the user's eye movements. Preferably, the eye-tracking data may include, but is not limited to, data related to eye movements (or eye movement patterns) such as fixation, saccades, and tracking. Preferably, the eye-tracking data may also include: the user's fixation point position, fixation time, and pupil diameter. Preferably, the above-mentioned eye-tracking data can be obtained by monitoring the user's eye behavior using an electrooculogram (EOG) sensor or other related eye-tracking sensors.

[0097] Preferably, the linkage between the digital chess and card game 1 and the mobile device equipped with or integrated with the digital evaluation unit 4 can be at least in the following two ways: 1) The mobile device equipped with or integrated with the digital evaluation unit 4 and the intelligent trajectory analysis sensor 102, etc., transmit data to the gateway via Bluetooth or wirelessly, and then the gateway transmits the data to the server 2 via wired or wireless means; 2) The data collected by the intelligent trajectory analysis sensor 102 is transmitted to the digital evaluation unit 4 via Bluetooth or wirelessly, the digital evaluation unit 4 transmits the data to the gateway via Bluetooth or wirelessly, and the gateway transmits the data to the server 2 via wired or wireless means.

[0098] Preferably, the interaction data, language data, and eye-tracking data are acquired by the digital assessment unit 4. Acquisition methods include at least data reception and manual input. Preferably, the digital assessment unit 4 can be located within or integrated into a mobile platform. Preferably, the mobile platform includes, but is not limited to, mobile phones or tablets. Preferably, the digital assessment unit 4 can be equipped with or integrated with appropriate sensors to acquire the user's interaction data, language data, eye-tracking data, and / or pupil data, depending on actual needs.

[0099] Preferably, the user can also interact with the digital chess and card game 1 through the digital evaluation unit 4. For example, the digital evaluation unit 4 can issue evaluation instructions to the user; for example, the evaluation instruction could be "Please name the missing mahjong tile on the screen used by the digital evaluation unit 4, and remove the missing mahjong tile from the digital chess and card game 1", after which the digital evaluation unit 4 collects the user's interaction data and language data during the answer process; at the same time, the digital chess and card game 1, which is moved by the user, collects its own motion data.

[0100] Preferably, language data is not limited to the user's voice information, but also includes language and text information generated by user interaction. The language and text information is natural language information, including at least multiple languages ​​and scripts, numerical symbols, and image symbols with linguistic meaning, etc.

[0101] For example, the digital assessment unit 4 can directly transmit the user's language data it acquires to the server 2 wirelessly via Bluetooth technology or other means.

[0102] For example, the digital assessment unit 4 can acquire the user's motion data collected by the digital chess and card game 1, and evaluate the user's cognitive ability based on the aforementioned motion data and the language data collected by the digital assessment unit 4 itself. For example, the digital assessment unit 4 can evaluate the user's cognitive ability based on the clarity of the user's voice information, the smoothness of the chess and card game movement trajectory recorded by the intelligent trajectory analysis sensor 102, and the length of time it takes to complete an assessment unit.

[0103] For example, the digital assessment unit 4 can evaluate a user's cognitive ability by assessing whether the information input by the user conforms to the input standards. Input standards include, but are not limited to, speech clarity, clarity of trajectory memory, test duration, and whether the input information exceeds a specified range (e.g., the input should be numbers between 1 and 9, but the actual input exceeds this range). Through this configuration, the digital assessment unit 4 can be set up or integrated into a game app on a mobile device (e.g., a mobile phone, tablet, etc.). This allows the app to not only acquire language data or voice information from user interactions with the game app, but also to acquire user interaction data through the digital chess / card game 1, which is linked or connected to the digital assessment unit 4. Interaction data includes at least action information, trajectory information, and parameter information related to the action and trajectory information. Parameter information related to the action and trajectory information includes, for example, the interval between two actions, the continuity of actions, and the frequency of clicks. For example, when a user draws a clock using the digital assessment unit 4, some users may initially find it difficult. Compared to traditional scale assessments of drawing a clock, this method uses the screen of the digital assessment unit 4 and the electronic stylus to capture the following characteristics: the force applied by the user when using the electronic stylus to touch the touchscreen of the digital assessment unit 4, the speed and / or pause time of the electronic stylus tip traversing the touchscreen, and the shape of the handwriting formed by the user using the electronic stylus, etc. This data is saved to itself or sent to server 2 for a preliminary digital assessment of the user's cognitive abilities by server 2.

[0104] Preferably, the assessment data generated from the preliminary digital assessment of the user's cognitive abilities includes, but is not limited to: the force parameters of the electronic stylus when the user draws on the touchscreen of the digital assessment unit 4, the speed and / or pause time of the electronic stylus tip tracing across the touchscreen, the shape of the handwriting formed by the user using the electronic stylus, the clarity of the user's speech when interacting with the digital assessment unit 4 during the assessment process, the smoothness of the trajectory formed by the movement of the digital chessboard 1, and the time required to complete one assessment. This configuration not only allows for comparison of the user's gesture trajectories and the force of the electronic stylus when using the digital assessment unit 4 with other individuals with normal cognitive abilities, but more importantly, it enables the digital assessment unit 4 or server 2 to monitor and assess the changes in the elderly person's cognitive abilities over different periods, thereby avoiding the misclassification of test subjects with low cognitive abilities as having insufficient cognitive abilities.

[0105] Preferably, the digital assessment unit 4 focuses on assessing the user's cognitive abilities in the following areas: attention, visuospatial ability, abstract thinking, executive function, immediate memory, delayed memory, language ability, and orientation. For example, the digital assessment unit 4 can cut a mahjong tile image into several pieces, shuffle them, and then have the elderly person determine which mahjong tile each piece was cut from, thereby selecting the corresponding mahjong tile from a pool of mahjong tiles to obtain assessment data corresponding to the elderly person's visuospatial ability.

[0106] Preferably, the digital assessment unit 4 can be digitized or integrated into a mobile platform game app or other assessment device. With this configuration, user interaction data and / or language data can be obtained through user interaction with the digital assessment unit 4 or the mobile platform game app or other assessment device, thereby scoring or assessing the user's cognitive abilities.

[0107] Preferably, the main evaluation process of the digital evaluation unit 4 includes: data acquisition, speech recognition, logic verification, rule learning, and scoring.

[0108] Both the digital evaluation unit 4 and the server 2 may have data processing capabilities. When the data processing capabilities of the digital evaluation unit 4 and the server 2 differ, the process can be divided into two stages.

[0109] The first phase of the functionality of digital assessment unit 4 includes the following steps:

[0110] (1) In response to the user’s input / feedback on the questions raised by the digital assessment unit 4, the digital assessment unit 4 can determine whether the input is within the specified numerical range (such as whether the user’s birth date corresponds to the ID card, whether the scores of each cognitive assessment ability are within the specified numerical range, whether the input data is clear, whether the timeout occurs, etc.); if the input information is within the specified numerical range and there is no contradictory information, the logical verification is completed.

[0111] (2) The input information and the corresponding cognitive ability score of the cognitive domain are transmitted to the server 2 to obtain the total score of each digital assessment unit 4. The total score is compared with the cognitive assessment level corresponding to the set score range to realize the assessment of the subject's cognition.

[0112] The second phase includes the following steps:

[0113] (1) Based on multimodal data fusion technology, server 2 performs data fusion on language data acquired by voice acquisition device, sensor data and cognitive ability score data acquired by intelligent environment acquisition device;

[0114] (2) Extract features that may be used to characterize the user's cognitive state from the fused data using machine learning-based feature extraction methods (such as partial least squares, autoencoder algorithm and its derivative algorithms, adversarial network learning algorithm and its derivative algorithms, etc.).

[0115] (3) The extracted features are associated with the cognitive domain to obtain relevant digital biomarkers that can characterize the user's cognitive state. The importance of each digital biomarker is ranked according to its weight in the feature extraction algorithm to achieve cognitive screening and help detect cognitive impairment early.

[0116] This configuration integrates the digital assessment unit 4 into a mobile game app, enabling it to acquire not only question-and-answer information but also user action information and the trajectory information formed by the stylus on the screen. For example, the digital assessment unit 4 allows users to draw a clock on the game app (such as a tablet). The corresponding sensors on the mobile platform can capture the user's action and trajectory information while drawing the clock. For instance, some users may encounter difficulties drawing slowly. Compared to traditional scale assessments of clock drawing, this method, through the corresponding sensors on the mobile platform, captures at least (but not limited to) the following user characteristic data: stylus pressure, drawing speed, pause duration, accuracy of numerical expression, the trajectory formed by the stylus on the touchscreen, the dwell time for each stroke, and hand tremors. This characteristic data allows for a digital assessment of the user's cognitive abilities. This configuration, namely the digital assessment unit 4, allows for more diverse dimensions of user characteristic data and a more comprehensive collection of data categories, resulting in a more comprehensive and objective assessment of the user's cognitive abilities.

[0117] Preferably, the digital evaluation unit 4 within the mobile platform can establish interactive information such as voice with the user to obtain the user's language data (especially when capturing mahjong tiles embedded with the intelligent trajectory analysis sensor 102). For example, the digital evaluation unit 4 on the mobile platform asks the user, "Please name the missing mahjong tiles on the current page." The digital evaluation unit 4 then performs semantic conversion on the voice information used to answer to obtain the user's answer. Next, the digital evaluation unit 4 determines whether the answer is within a specified range. If it is, it determines the score (whether it is correct and the determination of lost points are preset by the digital evaluation unit 4) and transmits it to the server 2. For information used for input, it directly judges and transmits it to the server 2 (similar to the operation after semantic conversion of voice information). For trajectory information, it needs to be transmitted to the server 2, where a machine learning-based judgment algorithm determines the score.

[0118] According to a preferred embodiment, the digital assessment unit 4 is further equipped with or configured with an early warning module and a display module. The early warning module can compare the cognitive ability score of the user in the current period with the cognitive ability score of the user in the previous period. The early warning module can generate a first early warning message when the user's cognitive ability score decreases by more than a preset trigger threshold. The first early warning message can be displayed through the display module.

[0119] Preferably, the comparison period for the evaluation unit to periodically compare the cognitive ability score of the user in the current period with that of the user in the previous period can be manually set. Preferably, the preset trigger threshold can be manually set. For example, the preset trigger threshold could be 3% of the user's cognitive assessment score in the previous period. For example, the early warning module of the evaluation unit can be configured to use at least one of the following periods: day, week, month, quarter, and year. For example, the early warning module of the evaluation unit can be configured to compare the corresponding user's cognitive ability score on a daily basis. Or, for example, the early warning module of the evaluation unit can be configured to compare the corresponding user's cognitive ability score on both daily and weekly basis. That is, the early warning module of the evaluation unit compares the cognitive ability scores of the user in two consecutive days, and also compares the cognitive ability scores of the user in two consecutive weeks.

[0120] Preferably, the early warning module of the evaluation unit can periodically compare the cognitive ability score of the user in the current period with that of the user in the previous period at at least two different comparison periods. Different comparison periods can correspond to different preset trigger thresholds. For example, some users play cards quickly in the early stages of using the digital card game 1, but as their playing speed gradually slows down over time (e.g., a slowing trend over several months or one or two years), the early warning module of the evaluation unit may determine that the user has a cognitive decline. Through this configuration, the changes in the user's cognitive ability over different periods can be monitored and evaluated, i.e., longitudinal data on the user's cognitive ability. For example, comparing user A's data today with the data from next week's test, and then again the week after, may reveal the changing trend of evaluation ability, thus avoiding the misjudgment of users with already low cognitive ability as having insufficient cognitive ability.

[0121] According to a preferred embodiment, the digital assessment unit 4 or the server 2 can respectively acquire the user's motion data when using the digital chess and card game 1, the user's language data acquired by the digital assessment unit 4, and the user's daily data acquired by the intelligent environment device unit 5, so as to obtain digital biomarkers that may indicate changes in the user's cognitive ability or state through subsequent data analysis.

[0122] According to a preferred embodiment, the data analysis process performed by the digital evaluation unit 4 or the server 2 is as follows: the motion data of the user using the digital chess and card game 1 is input into a deep learning network to extract feature data; the user's language data obtained by the digital evaluation unit 4 is converted into feature data through Fourier transform and frequency domain analysis, and the feature data is further extracted from the above data through stack self-encoding; the user's daily data obtained by the intelligent environment device unit 5 is used to extract possible feature data through ontology learning.

[0123] The process of data analysis (i.e., extraction of digital biomarkers) when server 2 or cloud platform acquires user motion data (i.e., action-motion data) from using digital chess and card games 1, user interaction data, language data, eye-tracking data recorded by digital evaluation unit 4, and user daily data acquired by intelligent environment device unit 5 is as follows: The user's motion data (i.e., time-series signals) from using digital chess and card games 1 is input into a deep learning network to extract user feature data or indicators. The user's language data (i.e., time-series signals) acquired by digital evaluation unit 4 is converted into feature data using techniques such as Fourier transform and frequency domain analysis, and then further feature data is extracted from the above data using stacked self-encoding. The user's daily data acquired by intelligent environment device unit 5 (e.g., water immersion, ultrasound, and pressure data acquired when the user uses the toilet) is used to extract feature data that can characterize the target attribute through ontology learning. Preferably, the target attribute includes, but is not limited to, the ability to perform all cognitive processes. Preferably, the target attribute may also include the ability to perform mental processes. Preferably, the target attribute may also include the ability to learn and judge, and the ability to use language and memory.

[0124] According to a preferred embodiment, the digital evaluation unit 4 or the server 2 performs feature selection on all the aforementioned feature data using a feature selection method to screen out potential digital biomarkers, and then ranks the potential digital biomarkers by importance based on marginal contribution analysis. Preferably, the feature selection method may include at least one or more of partial least squares, variational autoencoder, and adversarial network learning. Preferably, other types of methods may also be used for feature selection. After the server 2 initially extracts the aforementioned feature data, the server 2 can perform feature selection on all the extracted features using partial least squares:

[0125] Y = X * (X T S(T T XX T S) -1 T T Y)+R e (1), Where S is the vector mapping the independent variable X, T is the vector mapping the dependent variable Y, and R... e The matrix consists of the corresponding residues. Server 2 performs feature selection on all the extracted feature data using partial least squares to screen out possible digital biomarkers, and then ranks the importance of the possible digital biomarkers based on marginal contribution analysis.

[0126] Preferably, the reference biomarkers can be selected based on specific medical screening / assessment needs. For example, effective digital biomarkers (in descending order of importance) analyzed by server 2 include: user hand movement trajectory, frequency and / or duration of user hand pauses. Server 2 can use these effective digital biomarkers as reference biomarkers for screening brain health.

[0127] Simultaneously, server 2 can acquire real-time scores of the user's cognitive abilities from the digital assessment unit 4 on the mobile platform, which can be used to assist in identifying or mining digital biomarkers that can be used to identify a decline (or change) in the user's cognitive level. Through this configuration, digital biomarkers that may be used to identify a decline or change in the user's cognitive level can be extracted from the user's motion data when using the digital chess and card game 1, the user's language data recorded when using the digital assessment unit 4, and the user's daily data acquired by the intelligent environment device unit 5.

[0128] According to a preferred embodiment, a causal analysis knowledge base is constructed based on the acquired feature data, and causal inference is performed on digital biomarkers based on the established causal analysis knowledge base using causal analysis theory, in order to discover digital biomarkers that can effectively identify changes in users' cognitive abilities or states.

[0129] The cognitive ability assessment method based on digital chess and card games 1 includes the following steps: acquiring at least the user's motion data through digital chess and card games 1; acquiring at least the user's language data through digital assessment unit 4; acquiring the user's daily data through intelligent environment device unit 5; server 2 acquiring the above-mentioned motion data, language data, and daily data to extract possible digital biomarkers; and establishing a causal relationship knowledge base based on the above-mentioned possible digital biomarkers to calculate and analyze reference biomarkers.

[0130] Preferably, the reference biomarker can be selected based on specific medical screening / assessment needs. For example, the reference biomarker can be a digital biomarker that can effectively identify changes in a user's cognitive abilities or state.

[0131] This invention provides a method for causal inference from candidate digital biomarkers based on causal learning.

[0132] The method includes:

[0133] S1: Constructing the original document database using document units;

[0134] S2: Data units construct datasets;

[0135] S3: Causal units construct causal relationships between symptoms;

[0136] S4: The knowledge unit stores the original literature database, the dataset, and / or the average causal effect to construct the knowledge base that can be read and / or displayed. Thus, the information provided by the knowledge base can be offered to medical professionals for reference, learning, and / or decision-making in a quantified data format.

[0137] To reduce the interference of numerous feature parameters generated from a large number of related documents on the causal relationship between disease pairs and to improve the utilization value of the original literature database, preferably, the literature unit can acquire numerous related documents containing multiple cognitive abilities and classify them into several literature unit bodies to construct the original literature database. This allows the data unit to obtain the main feature parameters based on the literature unit bodies and construct the dataset based on the main feature parameters.

[0138] Preferably, the causal unit constructs a Bayesian network based on key feature parameters and the dataset to analyze the average causal effect between cognitive abilities through data pattern analysis. This enables the knowledge unit to construct a knowledge base based on relevant literature, forming correspondences between cognitive abilities and their average causal effects. For example, the average causal effect between cognitive abilities can reflect whether cognitive abilities constitute complications or comorbidities.

[0139] Preferably, the key characteristic parameters may include, but are not limited to, possible digital biomarkers obtained in the aforementioned steps of server 2.

[0140] Preferably, the dataset may include, but is not limited to, voice feature data, behavioral feature data, risk factor data, and physiological indicator data acquired by server 2. Preferably, the behavioral feature data may include the motion data of the digital chess and card game 1. Preferably, the behavioral feature data may also include other user behavior information collected by wearable device 3, such as the number of steps the user takes each day and sleep duration.

[0141] Preferably, the causal unit can also analyze the direct causal effects between cognitive abilities through data pattern analysis, thereby enabling the knowledge unit to construct a knowledge base based on relevant literature in a way that forms a correspondence between cognitive abilities and the direct causal effects between cognitive abilities.

[0142] Preferably, the main characteristic parameters used by the causal unit may include, but are not limited to: the user's motion data when using the digital chess and card game 1, the user's language data recorded when using the digital evaluation unit 4, and the user's daily data obtained by the intelligent environment device unit 5.

[0143] Preferably, the literature unit is based on a large number of relevant documents containing multiple cognitive abilities. The literature unit categorizes the relevant documents to form several literature unit bodies to construct the original literature database. These relevant documents include medical records, research reports, conference proceedings, journal articles, books, academic papers, and patents. Given such a large volume of documents, they need to be classified according to certain methods. Literature classification aims to effectively observe the relationships between cognitive abilities and reduce the system load. For example, they can be classified according to digestive diseases, cardiovascular diseases, and neurological diseases. They can also be classified according to academic fields, such as rehabilitation medicine and psychology. However, given the large volume of literature, accurate and efficient classification directly affects the differentiation of complications and comorbidities. Preferably, literature classification can employ Bayesian methods, SVM methods, and k-NN methods.

[0144] Preferably, the relevant literature classification is performed as follows: S11: The frequency of words / phrases in each literature is statistically analyzed, and the joint occurrence probability of words / phrases is obtained according to the independence assumption. For example, for a specific literature, its joint occurrence probability distribution can be calculated using the Naive Bayes method.

[0145] S12: Calculate the correlation strength of words / phrases within a document unit. Calculating the correlation strength reflects the relevance of words / phrases, which is suitable for document classification. Preferably, during classification, N is defined as the set of document samples, V as the set of document types, and Vi as a subset of the i-th document type. W is the set of words / phrases, and Wi is a subset of the i-th word / phrase. Vi contains Sj samples, where the correlation reduction coordinate Tp of the p-th sample is an n-dimensional array:

[0146]

[0147] Where, ki (i = 1, 2, 3, ..., n) represents the number of occurrences of the i-th word. Normalization coefficient.

[0148] The association vector in Vi is the average of the reduced coordinates of the associations of all samples in Vi. This value reflects the strength of the associations between words / phrases in the literature.

[0149]

[0150] S13: Document unit 1 obtains the reduced coordinates of the document and, based on the classification function constructed from the reduced coordinates of all related documents, classifies the related documents according to an iterative algorithm to form several document unit bodies. Preferably, the reduced coordinates of any document are:

[0151]

[0152] In the formula, qi is the number of times the i-th word appears in the document. When classifying, the distance between the document to be classified and the support points (b1, b2, ..., bn) of each category of documents Vi is denoted as:

[0153]

[0154] Based on the strength of the correlation, construct a document classification function:

[0155]

[0156] In the formula, γi is related to the correlation strength.

[0157] Preferably, the iterative algorithm can employ a minimization iterative algorithm, a minimum optimization iterative algorithm, or an expectation-maximization iterative algorithm. Preferably, the classification function can be based on the sample size of relevant documents through deep learning, thereby enhancing the accuracy of document units.

[0158] Preferably, the data unit can obtain key feature parameters based on the document unit body and construct a dataset based on these key feature parameters. This reduces the interference of numerous feature parameters formed by numerous related documents on the causal relationship between cognitive abilities and improves the utilization value of the original document database. Preferably, when the data unit obtains the document unit body, the data unit obtains the dataset by pairing cognitive ability pairs. The data unit extracts the relationship between cognitive ability pairs in each related document using syntactic analysis of natural language processing to establish a relational knowledge base for cognitive ability pairs. The relationships between cognitive ability pairs include positive, negative, and vertical relationships. Furthermore, based on the relational knowledge table, the data unit searches for documents containing cognitive ability pairs within the document unit body and obtains the relational reliability values ​​of cognitive ability pairs in a fusion manner to establish a relational reliability value database for cognitive ability pairs. The relationships between cognitive ability pairs include positive, negative, and vertical relational reliability values. Thus, the data unit constructs a dataset based on the relational knowledge base and relational reliability value database established by pairwise pairing of all cognitive abilities. For example, in related documents, a certain disease or behavior L1 and cognitive state L2 are obtained. The relationship between a certain symptom or behavior L1 and cognitive state L2 can be positive, meaning that symptom or behavior L1 influences cognitive state L2, denoted as L1→L2. The relationship between symptom or behavior L1 and cognitive state L2 can also be negative, meaning that cognitive state L2 influences symptom or behavior L1, denoted as L2→L1. The relationship between symptom or behavior L1 and cognitive state L2 can also be vertical, meaning that cognitive state L2 and symptom or behavior L1 do not influence each other (L1⊥L2). Since complications or comorbidities are multiple, it can also include another symptom or behavior L3 and several cognitive abilities such as cognitive state L4. Based on the above construction of cognitive ability relationships, a knowledge base can be constructed for the relationship between symptom or behavior L1 and another symptom or behavior L3, the relationship between cognitive state L2 and another symptom or behavior L3, and so on. Then, within a unit document, a relationship reliability value database is constructed based on the above relationship knowledge base according to the content of different documents. Preferably, the sum of the positive relation reliability value, negative relation reliability value, and vertical relation reliability value is normalized. That is, within the unit document body, all documents are traversed and queried, and the positive relation reliability value, negative relation reliability value, and vertical relation reliability value are weighted according to frequency. The data unit inputs the dataset constructed from the above relation knowledge base and relation reliability value database into the causal unit for the next step.

[0159] Preferably, for journal articles, the reliability value of the positive relationship between L1 and L2 can also be defined as follows:

[0160]

[0161] Where C(Xi) represents the credibility of document Xi, calculated as: C(Xi) = (IFi+1) × (CIi+1), where Xi represents the i-th document, IFi is the standardized impact factor of the journal containing document Xi, and CIi is the standardized citation count. If no document has an L1-L2 relationship, then KL(L1→L2) = 0, KL(L2→L1) = 0, and KL(L1⊥L2) = 1. Other types of documents can be defined in the same way; for example, medical records can be defined based on the doctor's authority, and conference articles can be defined based on the conference's authority, and so on.

[0162] Preferably, the causal unit constructs a Bayesian network based on key feature parameters and the dataset. Preferably, the key feature parameters include positive relation confidence values, negative relation confidence values, and vertical relation confidence values. The causal unit constructs the Bayesian network as follows:

[0163] S31: Preferably, a dataset D = (D1, D2, ..., Di) is defined as several groups of cognitive abilities, and L = (L1, L2, ..., Ln) is the specific set of cognitive abilities for a certain group of cognitive abilities. A Bayesian network evaluation function is constructed based on the aforementioned relational knowledge base.

[0164] logP(G,D,KL)=logP(G)+logP(D|G)+logP(KL|G)

[0165] In the formula, G is a Bayesian grid, whose values ​​include a directed acyclic graph with a set of specific cognitive abilities (L = (L1, L2, ..., Ln)) as nodes. P(G) is the prior distribution. Based on existing knowledge, the maximum value of logP(G) + logP(D|G) is equivalent to logP(G|D). logP(G|D) can be scored according to the Bayesian Information Criterion (BIC).

[0166]

[0167] Wherein, if any edge in structure G is represented as Lm→Ln, then KL(GLm, GLn) = KL(Lm→Ln). KL(Lm→Ln) is the relation confidence value. The summation in the formula sums the literature knowledge confidence of the positive relations corresponding to all directed edges in structure G.

[0168] S32: Construct an undirected graph structure constraint based on the aforementioned relational knowledge base; for a given dataset D, for any cognitive ability pair Lm and Ln in D, obtain the cognitive ability pair IDs of attribute pairs Lm and Ln by retrieving them from the relational knowledge base. Based on these IDs, retrieve the relational reliability values ​​of Lm→Ln and Ln→Lm from the relational reliability value table of cognitive ability pairs Lm and Ln in the literature. If L1 influences L2, the connection is L1 connecting to L2 and pointing to L2, constructing a directed edge between L1 and L2, and assigning a positive relational reliability value. If L2 influences L1, the connection is L2 connecting to L1 and pointing to L2, constructing a directed edge between L2 and L1, and assigning a negative relational reliability value. If L2 does not influence L1, the two are not connected, and a vertical relational reliability value is assigned.

[0169] S33: Construct a Bayesian network based on the Bayesian network evaluation function and the undirected graph structure constraints. After determining the undirected graph structure constraints of the Bayesian network, a heuristic search algorithm, such as the K2 algorithm, can be executed to seek the network structure with the optimal evaluation function. The general steps are as follows: start the search from the initial model. At each step of the search, first, the current model is locally modified using search operators to obtain a series of candidate models. Then, the score of each candidate model is calculated, and the optimal candidate model is compared with the current model. If the score of the optimal candidate model is larger, it is used as the next current model, and the search continues; otherwise, the search stops, and the current model is returned. According to the Bayesian principle, the candidate model with the largest score is the Bayesian network. Preferably, the Bayesian network evaluation function is constructed based on the established Bayesian network and Bayesian rules. The Bayesian network evaluation function can be constructed based on classic heuristic structure learning algorithms, such as the K2 algorithm, the Max-Min Parents and Children algorithm, and Markov chain Monte Carlo search, etc.

[0170] The causal unit is based on mining the average causal effect between cognitive abilities through data patterns, thereby enabling the determination of whether cognitive abilities constitute complications or comorbidities based on the average causal effect. In calculating the average causal effect, the causal unit uses Pearl's principle and Bayesian network structure to calculate the average causal effect between cognitive abilities. When exploring whether event X is the cause of event Y, Pearl needs to intervene in X to implement event X, calculating E(Y|do(X)). That is, if the average change of event Y under the intervention of X is greater than the significance level, then X is considered the cause of Y. Specifically, in a given dataset D or Di, the cognitive abilities to be studied are first selected. These cognitive abilities include the target cognitive ability and other cognitive abilities that affect the target cognitive ability. For example, to study whether a certain condition or behavior L1 is a complication of cognitive state L2, all edges pointing to cognitive abilities of L1 are truncated. The average causal effect of a certain condition or behavior L1 on cognitive state L2 is then observed. If this change is greater than a set causal effect threshold, then the condition or behavior L1 and cognitive state L2 constitute a complication; otherwise, they constitute a comorbidity.

[0171] When using causal units to mine the average causal effect between cognitive abilities through data patterns, the sheer volume of literature results in a massive Bayesian grid. Therefore, the backdoor criterion is employed to calculate the average causal effect. The backdoor criterion states that the Bayesian grid G ​​is a directed acyclic graph, (Lm, Ln) is a pair of nodes in G, and the set of nodes Z is the backdoor of (Lm, Ln), where all nodes in Z are not descendants of Z, and Z blocks all paths connecting Lm to Ln. Therefore, the causal relationship between cognitive abilities and Lm and Ln can be inferred using the backdoor principle.

[0172] To simplify undirected graph constraints using independence tests without affecting the causal relationships between cognitive ability pairs, causal units can be used. For example, the chi-square independence test can be employed.

[0173] In this invention, the independence test can also be performed in the following ways:

[0174] For cognitive ability Lm, the nodes connected to Lm are obtained in an array based on the constructed undirected graph, forming its node set. The correlation between each node and cognitive ability Lm is calculated sequentially, and the node with the highest correlation is selected, with an independence assumption, and nodes independent of Lm under a given subset D are deleted. In this invention, entropy is used to measure the uncertainty of the random variable with respect to Lm. Given the random variable Lm, the uncertainty of the random variable Ln can be measured using conditional entropy as follows:

[0175]

[0176] The degree of correlation between random variables Ln and Lm can be measured by mutual information:

[0177]

[0178] If the mutual information exceeds the mutual information threshold, Ln and Lm are considered correlated. If the mutual information does not exceed the mutual information threshold, Ln and Lm are considered uncorrelated. This configuration method allows for the extraction of effective / important digital biomarkers for identifying changes in human users' cognitive levels or other diseases from data sources such as the user's motion data while using the digital chess and card game 1, the user's language data recorded while using the digital assessment unit 4, and the user's daily data acquired by the intelligent environment device unit 5.

[0179] For example, a causal knowledge base might infer that depression and habitually inefficient sleep are both effective / important digital biomarkers of cognitive decline, but the causal relationship between depression and cognitive decline is stronger than that between habitually inefficient sleep and cognitive decline.

[0180] For example, increased intensity of learning activities and increased frequency of physical exercise are effective / important digital biomarkers of improved cognitive abilities.

[0181] For example, through a causal knowledge base, it may be inferred that the increased frequency of hand tremors when a user uses digital chess and card games is an effective digital biomarker of a decline in the user's cognitive ability over a certain period or short period.

[0182] For example, the decision-making time for each user during a certain period or a short period of time when using digital chess and card games shows a significant increasing trend.

[0183] For example, the frequency of a user's hand tremors can be directly related to symptoms such as reduced movement, rigidity, tremor, and postural instability, which can lead to paralysis angitas, also known as Parkinson's disease, and may have a direct causal effect on the patient's mobility, attention, orientation, and visuospatial abilities.

[0184] For example, if the digital assessment unit 4 or wearable device 3 detects that the number of times a user pauses in speech, frequently makes consonants such as "uh-huh," "ah," or has obvious difficulty expressing themselves exceeds the normal reference value in a short period of time / recently, it may be a sign of a decline in the user's language nervous system and can serve as an effective digital biomarker of cognitive decline. The normal reference value can be obtained from server 2 or other means, which can be the average / median of normal cognitive abilities or normal individuals.

[0185] Through this configuration, server 2 can also generate a causal relationship network model based on relevant variable features to perform causal reasoning, in order to identify the causal relationship between relevant feature variables related to changes in user cognitive ability and changes in cognitive ability, as well as the strength of the causal relationship. This can provide medical personnel or medical researchers with an effective way to understand effective digital biomarkers or pathological causes that cause changes in human individual cognitive ability.

[0186] Through this configuration, server 2 can also update the digital assessment unit 4 or server 2 based on the discovered effective digital biomarkers to further accurately screen / assess brain health. Simultaneously, the discovered effective digital biomarkers can also serve as the basis for subsequent digital assessment unit 4 or server 2 as a brain health (e.g., Alzheimer's disease) risk assessment platform.

[0187] This invention also provides a data processing method based on digital biomarkers. The method is as follows:

[0188] Digital chess and card game 1 collects motion data of the digital chess and card game 1 when a user uses it.

[0189] Server 2 acquires the motion data collected by the digital chess and card game 1, calculates and analyzes the motion data to obtain digital biomarkers corresponding to reference markers, and assesses / screens the user's brain health based on the digital biomarkers.

[0190] Preferably, server 2 can access previously recorded or analyzed valid reference biomarkers, and compare the differences between the digital biomarkers calculated by server 2 and the reference biomarkers or previously valid digital biomarkers, and assess / screen the user's brain health based on these differences. For example, server 2 can assess the cognitive decline of users in a population based on these differences, or predict the user's trend of change within a specified future timeframe.

[0191] For example, the reference marker is the pause time and / or pause frequency of the user's hand when moving the digital chess and card game 1, as determined by server 2's prior analysis of the raw data.

[0192] Preferably, server 2 is able to periodically or irregularly compare the frequency and / or intensity of digital biomarkers of users in the current period with the frequency and / or intensity of digital biomarkers of users in the previous period. The period can be manually set, such as weekly.

[0193] For example, if server 2 identifies that the frequency of digital biomarkers (the intensity and / or frequency of hand tremors when the user uses digital chess and card games 1, or the pause time / decision time during each use of digital chess and card games 1) exceeds a preset threshold within a certain time period / cycle, server 2 determines that the user's cognitive ability is likely to decline at this time. The preset threshold can be inferred by server 2 through a causal relationship knowledge base.

[0194] For example, when server 2 detects that the frequency of a certain digital biomarker (such as multiple pauses in speech, frequent use of consonants like "uh-huh" or "ah-ah") exceeds a preset threshold within a certain time period / cycle, server 2 determines that the user's cognitive ability is likely to decline at this time.

[0195] According to a preferred embodiment, the reference marker is obtained by the following steps:

[0196] Obtain the raw data;

[0197] Based on machine learning, feature data is extracted from the raw data, and candidate digital biomarkers are further screened from the feature data;

[0198] Causal inference is performed from the candidate digital biomarkers based on causal learning to calculate and analyze the reference biomarker.

[0199] Preferably, the raw data is obtained by the following steps:

[0200] Digital chess and card game 1 collects motion data of the digital chess and card game 1 when the user uses it.

[0201] Wearable device 3 collects users' daily physiological data;

[0202] Digital assessment unit 4 acquires user assessment data, language data, and eye-tracking data;

[0203] Intelligent environment device unit 5 collects daily data from the user's daily life;

[0204] Server 2 acquires one or more of the following as raw data: motion data, other physiological data, assessment data, language data, eye-tracking data, and daily ....

[0205] It should be noted that the specific embodiments described above are exemplary. Those skilled in the art can devise various solutions inspired by the disclosure of this invention, and these solutions all fall within the scope of this invention and its protection. Those skilled in the art should understand that this specification and its accompanying drawings are illustrative and do not constitute a limitation on the claims. The scope of protection of this invention is defined by the claims and their equivalents. This specification contains multiple inventive concepts; terms such as "preferredly," "according to a preferred embodiment," or "optionally" indicate that the corresponding paragraph discloses an independent concept. The applicant reserves the right to file divisional applications based on each inventive concept. Throughout the text, features introduced by "preferredly" are merely optional and should not be construed as mandatory. Therefore, the applicant reserves the right to abandon or delete relevant preferred features at any time.

Claims

1. A digital biomarker based data processing system, characterized in that, include: Digital chess and card games include chess and card games and intelligent trajectory analysis sensors. The system collects motion data of the digital chess and card games when users use them and sends the motion data to the digital evaluation unit and server. The intelligent trajectory analysis sensors are used to obtain the spatial coordinates of the chess and card games during the user's use and the timestamps associated with the spatial coordinates. Servers and digital assessment units set up within mobile platforms, Wearable devices send identifiers that identify the user to intelligent trajectory analysis sensors and servers. And an intelligent environment device unit, the intelligent environment device unit being used to acquire daily data of the user's daily life, the daily data being behavioral data of the user collected by the intelligent environment device unit throughout the day or during a specific period, the behavioral data being behaviors related to the user's cognitive ability assessment; The digital assessment unit monitors the interaction process between the user and the digital assessment unit to capture the user's interaction data, language data, eye movement data, and pupil data. The digital assessment unit sends the interaction data, language data, eye movement data, and pupil data of the interaction process to the server so that the server can calculate the user's assessment data based on the interaction data, language data, eye movement data, and pupil data of the interaction process. The digital assessment unit and the server respectively acquire the user's motion data when using digital chess and card games, the user's language data acquired by the digital assessment unit, and the user's daily data acquired by the intelligent environment device unit, in order to obtain digital biomarkers that correspond to reference markers that effectively represent changes in the user's cognitive ability or state, and to assess or screen the user's brain health based on the digital biomarkers. The intelligent trajectory analysis sensor automatically times the movement or vibration of the intelligent trajectory analysis sensor or the card. When the vibration value of the intelligent trajectory analysis sensor and the card changes from zero to a non-zero value, the intelligent trajectory analysis sensor determines this to be the first moment. When the vibration value of the intelligent trajectory analysis sensor and the card changes from a non-zero value to zero, the intelligent trajectory analysis sensor determines this to be the second moment. The intelligent trajectory analysis sensor obtains the identifier of the nearest wearable device if and only if it starts searching for wearable devices around it at the second moment. The reference markers are obtained through the following steps: Obtain the raw data; Based on machine learning, feature data is extracted from the raw data, and candidate digital biomarkers are further screened from the feature data; Causal inference is performed from the candidate digital biomarkers based on causal learning, and the reference biomarker is calculated and analyzed. The raw data is obtained through the following steps: The digital chess and card game collects the motion data of the digital chess and card game when the user uses it. Wearable devices collect users' daily physiological data; The digital assessment unit acquires the user's assessment data, language data, and eye-tracking data; The intelligent environment device unit collects daily data from the user's daily life. The server acquires one or more of the following as raw data: motion data, daily physiological data, assessment data, language data, eye-tracking data, and daily ....

2. The data processing system of claim 1, wherein, The intelligent trajectory analysis sensor (102) is disposed in or integrated into the chessboard (101).

3. The data processing system of claim 1, wherein, Wearable device (3) collects or is input into the user’s individual characteristic data and daily physiological data, the daily physiological data including at least heart rate, blood pressure, pulse oxygen saturation, body temperature and associated timestamps, and the individual characteristic data including at least familiarity with chess and card games, education level and medical history; The wearable device (3) sends individual characteristic data and daily physiological data to the server (2).

4. The data processing system of claim 3, wherein, The server (2) adjusts the type of motion data collected by the digital chess and card game (1) based at least on the reference marker.

5. A method using the data processing system based on digital biomarkers as described in any one of claims 1 to 4, characterized in that, The method is as follows: The digital chess and card game (1) collects the motion data of the digital chess and card game (1) when the user uses the digital chess and card game (1); The server (2) acquires the motion data collected by the digital chess and card game (1), calculates and analyzes the motion data to obtain digital biomarkers corresponding to the reference markers, and evaluates or screens the user's brain health based on the digital biomarkers. The reference marker is obtained through the following steps: Obtain the raw data; Based on machine learning, feature data is extracted from the raw data, and candidate digital biomarkers are further screened from the feature data; Causal inference is performed from the candidate digital biomarkers based on causal learning, and the reference biomarker is calculated and analyzed. The raw data is obtained through the following steps: The digital chess and card game (1) collects the motion data of the digital chess and card game (1) when the user uses it; Wearable devices (3) collect users' daily physiological data; The digital assessment unit (4) acquires the user's assessment data, language data, and eye-tracking data; The intelligent environment device unit (5) collects daily data from the user's daily life. The server (2) acquires one or more of the motion data, daily physiological data, assessment data, language data, eye movement data and daily data as raw data.

Images