Abnormal data processing system

Abstract

Provided is a technique capable of processing abnormal data with high precision. An abnormal data processing system of the present invention that detects whether new data is abnormal and processes it includes: a storage unit configured to hold a single-subject database in which data of a single subject is accumulated; a single-subject database deviation degree calculation unit configured to calculate a single-subject database deviation degree that is a degree to which the new data deviates from the single-subject database; and a calculation unit configured to calculate the single-subject database deviation degree using the number of data of the single-subject database, and determine whether the new data is abnormal based on the deviation degree.

Description

[0001] This application is a divisional application of international application number PCT / JP2018 / 038709, which entered the Chinese national phase on April 20, 2020, and has application number 201880068182.2. Technical Field

[0002] This invention relates to information processing service technology. Additionally, this invention relates to technology for implementing abnormal data processing. Background Technology

[0003] Systems for measuring human data are increasing in fields such as healthcare, medicine, and nursing. These systems provide value to users by calculating and analyzing the data and providing feedback. One example of such a system is a system that easily evaluates cognitive and motor functions by measuring and analyzing a user's finger-tapping movements (a finger-tapping measurement and analysis system) (e.g., Patent Document 1). Here, finger-tapping movements refer to the repeated opening and closing of the thumb and forefinger. Regarding finger-tapping movements, it is known that the results vary depending on the presence and severity of brain dysfunctions such as dementia and Parkinson's disease. This demonstrates the possibility of early detection and assessment of the severity of brain dysfunctions in users based on the analysis results of finger-tapping movements using the aforementioned system.

[0004] Existing technical documents

[0005] Patent documents

[0006] Patent Document 1: Japanese Patent Application Publication No. 2017-140424

[0007] Patent Document 2: Japanese Patent Publication No. 2013-535268

[0008] Patent Document 3: Japanese Patent Application Publication No. 2013-039344 Summary of the Invention

[0009] The problem that the invention aims to solve

[0010] Imagine if human body data measurement services, such as finger tapping measurement and analysis systems, became widely available in ordinary households. This would lead to situations where users perform the measurements alone, or where family members who are not accustomed to measurement assist them. Without a skilled measurement specialist present, there is a possibility that the measurement process may be inappropriate, or the user may not perform the expected actions, resulting in unreliable data (hereinafter referred to as "abnormal data").

[0011] For example, when citing the aforementioned finger tapping measurement and analysis system, it is believed that there may be situations where the user interrupts the finger tapping movement during the specified measurement time, or where the user misunderstands the instructions and performs the finger tapping movement.

[0012] When using such anomalous data, problems arise where the analysis results cannot be reliably provided to the user. For example, in the case of the aforementioned finger tapping measurement and analysis system, there are instances where the results perceived as finger tapping movements are worse than the actual condition, resulting in an analysis indicating a high probability of brain dysfunction even though the actual likelihood of brain dysfunction is low.

[0013] Therefore, a mechanism for automatically handling abnormal data is needed. Methods for implementing this abnormal data handling mechanism include (A) situations where anomalies can be detected simply by examining the data as the object, and (B) situations where anomalies can only be detected by comparing with a past database (DB). In case (A), as long as the data as the object exists, it can be implemented using a personal computer (PC) terminal (local PC) connected to a local measurement device. In case (B), if the past DB is stored on the local PC, it can be implemented on the local PC; however, if the past DB is stored on a cloud server, it needs to be implemented on the server.

[0014] The situation described in (B) above will be described in detail. When comparing the object data with past databases, it is assumed that there are two types of past databases: (i) a database consisting solely of the user's data (hereinafter referred to as the "single-subject database") and (ii) a database that also includes data from other users (hereinafter referred to as the "multi-subject database"). The single-subject database (i) reflects the individuality of the user's data, so it is preferable to use the single-subject database (i) for proper anomaly detection. However, if the user has performed measurements multiple times, the single-subject database (i) can be used, but if measurements have been performed for the first time or only a few times, the database (i) has not yet accumulated sufficiently, so the multi-subject database (ii) needs to be used.

[0015] In the case of using the multi-subject database (ii), the database is easy to prepare because data from users other than the current user only needs to be stored in advance, which is an advantage. However, the multi-subject database (ii) is a collection of data from multiple users and does not reflect the individuality of the user's data. Therefore, compared with the case of using the single-subject database (i), there is a possibility of incorrect detection of outlier data, which is a disadvantage.

[0016] Based on the above-mentioned issues, it is believed that a technique is needed to compensate for the strengths and weaknesses of (i) a single-subject database and (ii) a multi-subject database, and to utilize both databases simultaneously. Examples of prior art regarding abnormal data processing include Japanese Patent Application Publication No. 2013-535268 (Patent Document 2) and Japanese Patent Application Publication No. 2013-039344 (Patent Document 3). However, these two patent documents only describe methods for identifying whether test data is abnormal using a pre-provided database, and do not mention techniques for improving the accuracy of abnormal data detection by simultaneously using single-subject and multi-subject databases. Therefore, this invention proposes a technique for automatically processing abnormal data with high accuracy.

[0017] Technical solutions for solving the problem

[0018] An abnormal data processing system for detecting and processing new data according to one aspect of the present invention includes: a storage unit for storing a multi-subject DB containing data of multiple subjects and a single-subject DB containing data of a single subject; a single-subject DB deviation calculation unit for calculating a single-subject DB deviation as the degree to which new data deviates from the single-subject DB; a multi-subject DB deviation calculation unit for calculating a multi-subject DB deviation as the degree to which new data deviates from the multi-subject DB; and a composite deviation calculation unit for using the data count of the single-subject DB to obtain a composite deviation obtained by combining the single-subject DB deviation and the multi-subject DB deviation, and determining whether the new data is abnormal based on the composite deviation.

[0019] An abnormal data processing method using an input unit, an output unit, a control unit, and a storage unit to detect and process new data acquired by the input unit for anomalies, wherein the storage unit stores a multi-subject DB containing data from multiple subjects and a single-subject DB containing data from a single subject; calculates a single-subject DB deviation degree, which represents the degree to which new data deviates from the single-subject DB; calculates a multi-subject DB deviation degree, which represents the degree to which new data deviates from the multi-subject DB; and calculates a composite deviation degree based on the number of data in the single-subject DB using the single-subject DB deviation degree and the multi-subject DB deviation degree; and determines whether the new data is abnormal based on the composite deviation degree.

[0020] Invention Effects

[0021] A technology that enables high-precision automatic processing of abnormal data. Attached Figure Description

[0022] Figure 1 This is a block diagram illustrating the structure of the abnormal data processing system according to Embodiment 1 of the present invention.

[0023] Figure 2This is a block diagram representing the structure of the abnormal data processing system in Implementation 1.

[0024] Figure 3 This is a block diagram showing the structure of the measuring device in Embodiment 1.

[0025] Figure 4 This is a block diagram illustrating the structure of the terminal device in Embodiment 1.

[0026] Figure 5 This is a perspective view of a motion sensor worn on a finger, as shown in Embodiment 1.

[0027] Figure 6 This is a block diagram showing the structure of the motion sensor control unit and other components of the measuring device in Embodiment 1.

[0028] Figure 7 This is a flowchart illustrating the processing flow of the abnormal data processing system in Implementation Method 1.

[0029] Figure 8 This is a waveform diagram of an example of a waveform signal representing a characteristic quantity in Implementation 1.

[0030] Figure 9 This is a diagram illustrating a structural example of the feature quantity list in Implementation Method 1.

[0031] Figure 10 This is a waveform diagram illustrating an example of abnormal data detection without using DB in Implementation 1.

[0032] Figure 11 This is a conceptual diagram in Implementation 1, representing anomaly detection using a single subject DB and anomaly detection using a multi-subject DB.

[0033] Figure 12 This is a conceptual diagram illustrating the calculation method for the degree of composite deviation in Implementation 1.

[0034] Figure 13 This is a conceptual diagram illustrating the calculation method for considering the deviation of timing in Implementation 1.

[0035] Figure 14 This is a diagram illustrating the structure of the table showing the correspondence between the reasons for anomaly detection and the processing in Implementation 1.

[0036] Figure 15 This is a plan view of a menu screen, which is an example of a display screen, in Implementation 1.

[0037] Figure 16 This is a plan view of a task measurement screen, which is an example of a display screen, in Implementation 1.

[0038] Figure 17 This is a plan view of the evaluation result screen, which is an example of a display screen, in Implementation 1.

[0039] Figure 18 This is a plan view of the first abnormal data detection / processing screen, which is an example of a display screen, in Implementation 1.

[0040] Figure 19 This is a plan view of the second abnormal data detection / processing screen, which is an example of a display screen, in Implementation 1.

[0041] Figure 20 This is a block diagram illustrating the structure of the abnormal data processing system according to Embodiment 2 of the present invention.

[0042] Figure 21 This is a planar view of a finger tapping on a screen as an example of motion in Embodiment 2.

[0043] Figure 22 In embodiment 2, this is a waveform diagram showing the distance between two fingers tapping on the screen.

[0044] Figure 23 This is a plan view showing the extension as an example of motion in Embodiment 2.

[0045] Figure 24 This is a plan view showing continuous striking as an example of motion in Embodiment 2.

[0046] Figure 25 This is a plan view showing the tapping as a motion example in Embodiment 2.

[0047] Figure 26 This is a plan view showing five-finger tapping as an example of motion in Embodiment 2.

[0048] Figure 27 This is a block diagram illustrating the structure of the abnormal data processing system according to Embodiment 3 of the present invention.

[0049] Figure 28 This is a block diagram showing the structure of the server as the exception data processing system in Implementation Method 3.

[0050] Figure 29 This is a diagram illustrating a structural example of user information, which serves as management information for the server, in Implementation Method 3. Detailed Implementation

[0051] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Furthermore, in principle, the same reference numerals will be used for the same parts in all the drawings used to describe the embodiments, and repeated descriptions will be omitted.

[0052] The embodiments are described in detail with reference to the accompanying drawings. However, the present invention is not limited to the description of the embodiments shown below. Changes can be made to the specific structure without departing from the spirit or essence of the invention, a point that will be readily understood by those skilled in the art.

[0053] When multiple elements have the same or identical function, different suffixes are sometimes added to the same reference numerals in the description. However, when it is not necessary to distinguish between multiple elements, the suffixes are sometimes omitted in the description.

[0054] The terms "first," "second," and "third" used in this specification are added for the purpose of identifying constituent elements and do not necessarily limit their quantity, order, or content. Furthermore, the numbers used to identify constituent elements are used for each context, and a number used in one context may not necessarily represent the same structure in another. Additionally, a constituent element identified by a particular number does not preclude it from also having the function of a constituent element identified by other numbers.

[0055] The positions, sizes, shapes, and extents of the structures shown in the accompanying drawings and other documents are sometimes not intended to represent actual positions, sizes, shapes, or extents for ease of understanding. Therefore, the present invention is not necessarily limited to the positions, sizes, shapes, and extents disclosed in the accompanying drawings and other documents.

[0056] This embodiment proposes a technique for automatically processing anomalous data. In anomalous data processing, there are cases where (A) anomalies can be detected simply by looking at the data as the object, and (B) anomalies can only be detected by comparing with past databases. This embodiment, particularly for (B), proposes a technique for detecting anomalous data by simultaneously using (i) a database consisting only of the user's data (single-subject database) and (ii) a database that also includes data from other users (multi-subject database). According to a representative implementation, high-precision anomalous data detection can be achieved by mutually compensating for the accuracy reduction caused by insufficient data volume when using (i) a database consisting only of the user's data and the accuracy reduction caused by the inability to reflect individual differences when using (ii) a database that also includes data from other users.

[0057] Implementation Method 1

[0058] use Figures 1 to 19 The abnormal data processing system according to Embodiment 1 of the present invention will now be described. The abnormal data processing system of Embodiment 1 has the function of detecting abnormalities in data measured by the user and generating processing content for cases where an abnormality is detected. Through these functions, abnormal data can be detected and processed with high accuracy.

[0059] [Human Body Data Measurement System (1)]

[0060] Figure 1 This describes the structure of a human data measurement system, including the abnormal data processing system of Embodiment 1. In Embodiment 1, the human data measurement system is provided in facilities such as hospitals or elderly care facilities, or in a user's home. The human data measurement system includes an abnormal data processing system 1 and a measurement system 2, which is a magnetic sensor-type finger tapping motion system, connected via a communication line. The measurement system includes a measurement device 3 and a terminal device 4, connected via a communication line. Multiple measurement systems 2 may also be installed within the facility.

[0061] Measurement system 2 is a system that uses a magnetic sensor-type motion sensor to measure finger movements. A motion sensor is connected to measurement device 3. This motion sensor is worn on the user's finger. Measurement device 3 measures finger movements using the motion sensor and obtains measurement data including timing waveform signals.

[0062] The terminal device 4 displays various information for abnormal data processing on a display screen, including abnormal data detection results, reasons for abnormal detection, and abnormal data processing content, and accepts user input. In embodiment 1, the terminal device 4 is a PC.

[0063] The anomaly data processing system 1 provides anomaly data processing services as an information processing service. Its functions include anomaly data detection and anomaly data processing decision-making. The anomaly data detection function checks the measurement data obtained by the measurement system 2 for anomalies. The anomaly data processing decision-making function determines how to process the anomaly data detected by the anomaly data detection function.

[0064] The anomaly data processing system 1 inputs, for example, user instructions and measurement data, as input data to the measurement system 2. The anomaly data processing system 1 outputs, for example, anomaly data detection results and anomaly data processing content, as output data to the measurement system 2. The anomaly data detection results include not only whether the measurement data is abnormal, but also the reason for detecting the anomaly.

[0065] The human data measurement system of Implementation Method 1 is not limited to facilities such as hospitals or elderly care facilities and their examinees, but can be widely applied to general facilities or people. The measurement device 3 and the terminal device 4 can also be configured as an integrated measurement system. Similarly, the measurement system 2 and the abnormal data processing system 1 can be configured as an integrated device. The terminal device 4 and the abnormal data processing system 1 can also be configured as an integrated device, and the measurement device 3 and the abnormal data processing system 1 can also be configured as an integrated device.

[0066] [Abnormal Data Processing System]

[0067] Figure 2 The diagram illustrates the structure of the abnormal data processing system 1 according to Embodiment 1. The abnormal data processing system 1 includes a control unit 101, a storage unit 102, an input unit 103, an output unit 104, and a communication unit 105, which are connected via a bus. The input unit 103 is used for inputting operations performed by the administrator or other personnel of the abnormal data processing system 1. The output unit 104 is used for displaying images to the administrator or other personnel of the abnormal data processing system 1. The communication unit 105 has a communication interface and is used for communication processing with the measuring device 3 and the terminal device 4.

[0068] The control unit 101 controls the entire abnormal data processing system and consists of a Central Processing Unit (CPU), Read Only Memory (ROM), and Random Access Memory (RAM). Based on software program processing, it is a data processing unit that performs abnormal data detection and abnormal data processing decisions. The data processing unit of the control unit 101 includes a user information management unit 11, a task processing unit 12, an analysis and evaluation unit 13, an abnormal data detection unit 14, an abnormal data processing decision unit 15, an abnormal data processing execution unit 16, and a result output unit 17. The control unit 101 performs functions such as inputting measurement data from the measuring device 3, processing and analyzing the measurement data, outputting control instructions to the measuring device 3 and the terminal device 4, and outputting display data to the terminal device 4.

[0069] The User Information Management Department 11 processes and manages user information entered by users by registering it in the User Information 41 of the DB40, and verifies the User Information 41 in the DB40 when a user uses the service. User Information 41 includes each user's individual attribute values, usage history information, and user settings information. Attribute values ​​include gender, age, etc. Usage history information manages the user's history of using the services provided by this system. User settings information is the settings information configured by the user regarding the functions of this service.

[0070] The task processing unit 12 is responsible for processing tasks used for analysis and evaluation of motor functions, etc. In other words, a task is a specified finger movement. Based on the task data 42A in the DB40, the task processing unit 12 outputs the task to the screen of the terminal device 4. Additionally, the task processing unit 12 acquires measurement data of the task measured by the measuring device 3 and stores it in the DB40 as measurement data 42B.

[0071] The analysis and evaluation unit 13 calculates characteristic quantities representing the properties of the measurement data based on the user's measurement data 42B. The analysis and evaluation unit 13 stores the results of the analysis and evaluation processing, i.e., the analysis and evaluation data 43, in DB40.

[0072] The anomaly detection unit 14 processes anomaly data based on user-analyzed evaluation data 43, single-subject DB45A, multi-subject DB45B, and the anomaly detection reasons and processing correspondence table 50B in management table 50. It then processes the anomaly data and outputs the anomaly data detection result 44 to the screen of the terminal device 4. The anomaly detection unit 14 stores the anomaly detection result as anomaly data detection result 44 in DB40. The anomaly data detection result 44 includes not only whether the measured data is abnormal, but also the reason for the anomaly data detection. The anomaly data detection unit 14 sends the anomaly data detection result 44 from DB40 to the terminal device 4 and outputs it to the screen. The anomaly data detection unit 14 includes an anomaly data detection unit 14A that detects anomaly data without using DB, and an anomaly data detection unit 14B that detects anomaly data using DB. The abnormal data detection unit 14B using DB includes a single subject DB deviation calculation unit 14Ba, a multi-subject DB deviation calculation unit 14Bb, and a composite deviation calculation unit 14Bc.

[0073] The abnormal data processing decision unit 15 performs processing based on the abnormal data detection results 44 and the abnormal detection reasons and processing correspondence table 50B in the management table 50, generates abnormal data processing content 46 and saves it in DB40.

[0074] The abnormal data processing execution unit 16 performs processing based on the abnormal data processing content 46. At this time, depending on the abnormal data processing content 46, there may be cases where the measurement data 42B is stored in a single subject DB45A or multiple subject DB45B, or there may be cases where it is not stored.

[0075] Here, Single-Subject DB45A is a database consisting solely of data from a specific user, while Multi-Subject DB45B is a database that combines data from multiple users. Generally, the data in Single-Subject DB45A is a subset of that in Multi-Subject DB45B, but Multi-Subject DB45B may not include data from Single-Subject DB45A. Furthermore, Single-Subject DB45A and Multi-Subject DB45B can be treated as separate databases. However, it is also possible to attach user identification information to the data in Multi-Subject DB45B, enabling the extraction of data from specific users, thus allowing Multi-Subject DB45B to also perform the functions of Single-Subject DB45A.

[0076] The result output unit 17 processes the output of the user's analysis and evaluation data 43, abnormal data detection results 44, and abnormal data processing content 46 to the screen of the terminal device 4. The analysis and evaluation unit 13, the abnormal data detection unit 14, the abnormal data processing decision unit 15, the abnormal data processing execution unit 16, and the result output unit 17 cooperate to perform screen output processing.

[0077] The data and information stored in DB40 of storage unit 102 include user information 41, task data 42A, measurement data 42B, analysis and evaluation data 43, abnormal data detection results 44, single subject DB45A, multiple subject DB45B, abnormal data processing content 46, and management table 50. Control unit 101 maintains and manages the management table 50 in storage unit 102. The administrator can set the contents of management table 50. Management table 50 stores a feature quantity list 50A for setting feature quantities, an abnormal detection reason and processing correspondence table 50B for setting the processing corresponding to the abnormal detection reason, etc.

[0078] [Measuring device]

[0079] Figure 3 The structure of the measuring device 3 according to Embodiment 1 is shown. The measuring device 3 includes a motion sensor 20, a housing 301, a measuring unit 302, and a communication unit 303. The housing 301 has a motion sensor interface 311 connected to the motion sensor 20 and a motion sensor control unit 312 for controlling the motion sensor 20. The measuring unit 302 measures waveform signals through the motion sensor 20 and the housing 301 and outputs them as measuring data. The measuring unit 302 includes a task measuring unit 321 for obtaining measuring data. The communication unit 303 has a communication interface and communicates with the abnormal data processing system 1 to send the measuring data to the abnormal data processing system 1. The motion sensor interface 311 includes an analog-to-digital conversion circuit that samples and converts the analog waveform signal detected by the motion sensor 20 into a digital waveform signal. This digital waveform signal is input to the motion sensor control unit 312.

[0080] Alternatively, the measuring device 3 can store each measuring data in the storage unit, or the measuring device 3 can store the measuring data only in the abnormal data processing system 1 instead of storing each measuring data.

[0081] [Terminal Device]

[0082] Figure 4The structure of the terminal device 4 according to Embodiment 1 is shown. The terminal device 4 includes a control unit 401, a storage unit 402, a communication unit 403, an input device 404, and a display device 405. The control unit 401 performs control processing based on software program processing, including displaying abnormal data detection results, displaying abnormal data processing content, and executing abnormal data processing content. The storage unit 402 stores user information, task data, measurement data, analysis and evaluation data, abnormal data detection results, and abnormal data processing content obtained from the abnormal data processing system 1. The communication unit 403 has a communication interface, communicates with the abnormal data processing system 1 to receive various data from the abnormal data processing system 1, and sends user input instructions to the abnormal data processing system 1. The input device 404 includes a keyboard and mouse. The display device 405 displays various information on a display screen 406. Alternatively, a touch panel can be used as the display device 405.

[0083] [Finger, motion sensor, finger tapping meter]

[0084] Figure 5 This indicates that a magnetic sensor, which functions as a motion sensor 20, is worn on the user's finger. The motion sensor 20 has a pair of coil sections, namely a transmitting coil section 21 and a receiving coil section 22, connected to the measuring device 3 via a signal line 23. The transmitting coil section 21 generates a magnetic field, which is detected by the receiving coil section 22. Figure 5 In this example, on the user's right hand, the transmitting coil 21 is worn near the thumbnail, and the receiving coil 22 is worn near the index fingernail. The finger on which it is worn can be changed to another finger. The location of wear is not limited to near the fingernail.

[0085] like Figure 5 As shown, the motion sensor 20 is worn on the user's fingers, such as the thumb and index finger of the left hand. In this state, the user repeatedly opens and closes their fingers, i.e., taps their fingers. During the tapping motion, the user moves between a closed state (fingertips touching) and an open state (fingertips separated). With this movement, the distance between the coils of the transmitting coil 21 and the receiving coil 22, corresponding to the distance between the fingertips, changes. The measuring device 3 measures the waveform signal that changes in response to the magnetic field between the transmitting coil 21 and the receiving coil 22 of the motion sensor 20.

[0086] Finger tapping specifically includes the following tasks. Examples of these movements include free single-hand movement, single-hand rhythmic movement, simultaneous free movement of both hands, alternating free movement of both hands, simultaneous rhythmic movement of both hands, and alternating rhythmic movement of both hands. Free single-hand movement refers to tapping with two fingers of one hand as quickly as possible. Rhythmic single-hand movement refers to tapping with two fingers of one hand in response to a rhythmic stimulus. Simultaneous free movement of both hands refers to tapping with two fingers of both hands at the same time. Alternating free movement of both hands refers to tapping with two fingers of both hands at alternating intervals.

[0087] [Motion sensor control unit and finger tapping meter]

[0088] Figure 6 This section shows a detailed structural example of the motion sensor control unit 312 and the like in the measuring device 3. In the motion sensor 20, the distance D between the transmitting coil 21 and the receiving coil 22 is shown. The motion sensor control unit 312 includes an AC generating circuit 312a, a current generating amplifier circuit 312b, a preamplifier circuit 312c, a detector circuit 312d, an LPF circuit 312e, a phase adjustment circuit 312f, an amplifier circuit 312g, and an output signal terminal 312h. The current generating amplifier circuit 312a is connected to the current generating amplifier circuit 312b and the phase adjustment circuit 312f. The current generating amplifier circuit 312b is connected to the transmitting coil 21 via a signal line 23. The preamplifier circuit 312c is connected to the receiving coil 22 via a signal line 23. At the rear end of the preamplifier circuit 312c, the detector circuit 312d, the LPF circuit 312e, the amplifier circuit 312g, and the output signal terminal 312h are connected sequentially. The phase adjustment circuit 312f is connected to the detector circuit 312d.

[0089] AC generating circuit 312a generates an AC voltage signal of a predetermined frequency. Current generating amplifier circuit 312b converts the AC voltage signal into an AC current of a predetermined frequency and outputs it to transmitting coil section 21. Transmitting coil section 21 generates a magnetic field due to the AC current. This magnetic field induces an electromotive force in receiving coil section 22. Receiving coil section 22 outputs an AC current generated by the induced electromotive force. This AC current has the same frequency as the predetermined frequency of the AC voltage signal generated in AC generating circuit 312a.

[0090] The preamplifier circuit 312c amplifies the detected AC current. The detector circuit 312d detects the amplified signal based on the reference signal 312i from the phase adjustment circuit 312f. The phase adjustment circuit 312f adjusts the phase of the AC voltage signal at a specified frequency or twice the frequency from the AC generator circuit 312a and outputs it as the reference signal 312i. The LPF circuit 312e limits the frequency band of the detected signal and outputs it. The amplifier circuit 312g amplifies the signal to a specified voltage. Then, an output signal equivalent to the measured waveform signal is output from the output signal terminal 312h.

[0091] The output signal, i.e., the waveform signal, is a signal with a voltage value representing the distance D between two fingers. The distance D and the voltage value can be transformed based on a prescribed formula. This formula can be obtained through calibration. Calibration, for example, involves measuring while the user holds a block of a specified length with two fingers of the target hand. Based on the dataset of voltage and distance values ​​under this measurement, the prescribed formula is obtained using an approximate curve that minimizes error. Furthermore, the size of the user's hand can be determined through calibration for normalization of characteristic quantities, etc. In Embodiment 1, the aforementioned magnetic sensor is used as the motion sensor 20, and a measurement unit corresponding to this magnetic sensor is used. However, it is not limited to this; other detection units and measurement units such as accelerometers, strain gauges, and high-speed cameras can also be applied.

[0092] [Processing Flow]

[0093] Figure 7 This describes the overall process flow of the human data measurement system in Implementation 1, which is mainly performed by the abnormal data processing system 1. Figure 7 This includes steps S1 to S10. The following explanation will follow the steps in sequence.

[0094] (S1) User operates the measurement system 2. The terminal device 4 displays an initial screen on the display screen. The user selects the desired operation item on the initial screen. For example, selecting an operation item for abnormal data detection / processing. The terminal device 4 sends the instruction input information corresponding to the selection to the abnormal data processing system 1. In addition, the user can also input and register user information such as gender and age on the initial screen. In this case, the terminal device 4 sends the input user information to the abnormal data processing system 1. The user information management unit 11 of the abnormal data processing system 1 registers the user information in the user information 41.

[0095] (S2) Based on the instruction input information from S1 and the finger tapping task data 42A, the task processing unit 12 of the abnormal data processing system sends task data corresponding to the user to the terminal device 4. This task data includes information on one or more tasks related to finger movement, such as free movement of one hand, simultaneous free movement of both hands, and alternating free movement of both hands. Based on the received task data, the terminal device 4 displays the finger movement task information on a display screen. The user performs the finger movement task according to the task information displayed on the screen. The measuring device 3 measures this task and sends the measurement data to the abnormal data processing system 1. The abnormal data processing system 1 stores this measurement data in measurement data 42B.

[0096] (S3) The analysis and evaluation unit 13 of the abnormal data processing system performs analysis and evaluation processing on the user's motion function, etc., based on the measurement data 42B of S2, generates user analysis and evaluation data 43, and saves it in DB40. In the analysis and evaluation processing, the analysis and evaluation unit 13 extracts feature quantities based on the waveform signal of the user's measurement data 42B. These feature quantities include those recorded in the feature quantity list 50A, calculated based on the distance waveform (described later), and those calculated based on the speed waveform, etc. The analysis and evaluation unit 13 can also modify the extracted feature quantities based on attribute values ​​such as the user's age. The modified feature quantities can also be used for evaluation.

[0097] (S4) The result output unit 17 of the abnormal data processing system 1 outputs analysis and evaluation result information to the display screen of the terminal device 4 based on the analysis and evaluation data 43 of S3. The user can use the display screen to confirm the analysis and evaluation result information indicating the status of their own motor function, etc. Step S4 can also be omitted.

[0098] (S5) The anomaly detection unit 14A of the anomaly data processing system 11, which does not use the database, detects anomaly data based on the analysis and evaluation data 43 of S3 without using the database. That is, it detects anomalies that can be detected solely based on the measurement data without referring to the previously accumulated single-subject database 45A and multi-subject database 45B. The detailed detection method will be described later. The anomaly detection reason and processing correspondence table 50B records a list of detected anomaly detection reasons. Anomaly data is detected based on these anomaly detection reasons. The result is stored in the anomaly data detection result 44 of the storage unit 40. Alternatively, step S5 can be omitted, and only step S6 can be performed.

[0099] (S6) The anomaly detection unit 14B of the anomaly data processing system 1 uses the analysis and evaluation data 43 from S3, the single-subject DB 45A, and the multi-subject DB 45B to perform anomaly detection using the DB. That is, anomalies that can only be detected by comparing the measured data with past single-subject or multi-subject DBs are detected. The detailed detection method is described later. The anomaly detection reason and processing correspondence table 50B records a list of detected anomaly detection items. Anomaly data is detected based on these anomaly detection items. The result is stored in the anomaly data detection result 44 of the storage unit 40. Alternatively, step S6 can be omitted, and only step S5 can be performed.

[0100] (S7) The anomaly data processing decision unit 15 of the anomaly data processing system 1 determines the anomaly data processing content 46 based on the anomaly data detection result 44 generated in steps S5 and S6. The detailed decision-making method will be described later. The anomaly detection reason and processing correspondence table 50B records the correspondence between the anomaly data detection result 44 and the anomaly data processing content 46. The anomaly data processing decision unit 15 determines the processing content of the anomaly data based on this correspondence table and saves it in the anomaly data processing content 46.

[0101] (S8) The anomaly data processing system 1 uses the analysis and evaluation unit 13 to display the anomaly data detection results 44 generated in S5 and S6 and the anomaly data processing content 46 generated in S7 on the display screen. The user can use the display screen to confirm the detection results and processing content of the anomaly data.

[0102] (S9) The abnormal data processing system 1 performs abnormal data processing based on the abnormal data processing content 46. As an example, processing could be considered where the data is stored in the single-subject DB45A and the multi-subject DB45B if no abnormality is detected, but not stored if an abnormality is detected. Another example could be processing where the abnormal data processing content 46 is output to the terminal device 4 via the communication unit 105 and a request for re-measurement is made. Detailed execution methods will be described later.

[0103] (S10) If the abnormal data processing system 1 requests the terminal device 4 to remeasure in S9, it returns to S2 and repeats the process in the same way. If no remeasurement is requested, the process ends.

[0104] [Characteristics]

[0105] Figure 8 Examples of waveform signals representing characteristic quantities. Figure 8(a) represents the waveform signal of the distance D between the two fingers, (b) represents the waveform signal of the velocity of the two fingers, and (c) represents the waveform signal of the acceleration of the two fingers. The velocity of (b) can be obtained by time differentiation of the waveform signal of the distance in (a). The acceleration of (c) can be obtained by time differentiation of the waveform signal of the velocity in (b). The analysis and evaluation unit 13 obtains the waveform signal of the specified characteristic quantity as shown in this example based on the waveform signal of the measured data 42B, using operations such as differentiation and integration. In addition, the analysis and evaluation unit 13 obtains the value calculated according to the specified method based on the characteristic quantity.

[0106] Figure 8 (d) is an enlargement of (a), representing an example of the characteristic quantity. It shows the maximum value Dmax of the finger-tapping distance D and the tapping interval TI, etc. The horizontal dashed line represents the average value Dav of the distance D over the entire measurement time. The maximum value Dmax represents the maximum value of the distance D over the entire measurement time. The tapping interval TI is the time corresponding to the period TC of one finger tap, specifically representing the time from the minimum point Pmin to the next minimum point Pmin. In addition, it shows the maximum point Pmax and minimum point Pmin within one period of the distance D, the time T1 of the opening action, and the time T2 of the closing action, described later.

[0107] The following provides further detailed examples of the feature quantities. In Embodiment 1, multiple feature quantities obtained from the waveforms of distance, velocity, and acceleration described above are used. Furthermore, in other embodiments, only some of the multiple feature quantities may be used, or other feature quantities may be used; the details of the definition of the feature quantities are not limited.

[0108] Figure 9 This refers to the [distance] portion of the finger tapping feature recorded in feature list 50A. This association setting is an example and can be changed. Figure 9 In the feature quantity list 50A, the columns include feature quantity classification, identification number, and feature quantity parameters. The feature quantity classification includes [distance], [velocity], [acceleration], [knock interval], [phase difference], and [marker tracking]. For example, the feature quantity [distance] has multiple feature quantity parameters identified by identification numbers (1) to (7). The units are indicated in parentheses [].

[0109] (1) "Maximum amplitude of distance" [mm] is the distance waveform ( Figure 8 (a) The difference between the maximum and minimum amplitude values ​​in (a) of the amplitude. (2) The total distance traveled [mm] is the sum of the absolute values ​​of the distance change over the total measurement time of one measurement. (3) The average of the maximum distance points [mm] is the average value of the maximum amplitude points of each period. (4) The standard deviation of the maximum distance points [mm] is the standard deviation of the above values.

[0110] (5) "Slope (attenuation rate) of the approximate curve of the maximum distance" [mm / sec] is the slope of the curve approximating the maximum amplitude. This parameter mainly represents the amplitude change caused by fatigue during the measurement time. (6) "Coefficient of variation of the maximum distance" is the coefficient of variation of the maximum amplitude, and the unit is dimensionless (represented by "-"). This parameter is obtained by normalizing the standard deviation with the mean, thus eliminating individual differences in finger length. (7) "Standard deviation of the local maximum distance" [mm] is the standard deviation of the amplitude maximum at three adjacent locations. This parameter is used to evaluate the degree of error in the amplitude over a short local time.

[0111] The following description, without illustrations, explains each characteristic parameter. Regarding the characteristic quantity [velocity], the following characteristic parameters are indicated by identification numbers (8) to (22). (8) "Maximum amplitude of velocity" [m / s] is the waveform of the velocity ( Figure 8 The difference between the maximum and minimum speeds in (b). (9) "The average of the maximum opening speeds" [m / s] is the average of the maximum speeds during the opening action of each finger's tapping waveform. The opening action refers to the action that brings the two fingers from a closed state to their maximum open state. Figure 8 (d) (10) The “average of the maximum points of the closing speed” [m / s] is the average of the maximum speeds during the closing action. The closing action refers to the action that moves the two fingers from the widest open state to the closed state. (11) The “standard deviation of the maximum points of the opening speed” [m / s] is the standard deviation of the maximum speeds during the opening action. (12) The “average of the maximum points of the closing speed” [m / s] is the standard deviation of the maximum speeds during the closing action.

[0112] (13) "Energy balance" [-] is the ratio of the sum of squares of the velocities during the opening motion to the sum of squares of the velocities during the closing motion. (14) "Total energy" [m 2 / Second 2 ] is the sum of squares of the speeds over the total measurement time. (15) "Coefficient of variation of the maximum opening speed" [-] is the coefficient of variation of the maximum speed during the opening action, which is the value obtained by normalizing the standard deviation with the mean. (16) "Average of the maximum closing speed" [m / s] is the coefficient of variation of the minimum speed during the closing action.

[0113] (17) "Number of vibrations" [-] is the number obtained by subtracting the number of finger taps during larger opening and closing from the number of round trips in the waveform of the speed change from positive to negative. (18) "Average of distance ratio at peak opening speed" [-] is the average of the ratio of distance at the maximum speed in the opening action, with the amplitude of the finger taps taken as 1.0. (19) "Average of distance ratio at peak closing speed" [-] is the average of the ratio of distance at the minimum speed in the closing action, with the same ratio. (20) "Ratio of distance ratio at peak speed" [-] is the ratio of the value of (18) to the value of (19). (21) "Standard deviation of distance ratio at peak opening speed" [-] is the standard deviation of the ratio of distance at the maximum speed in the opening action, with the amplitude of the finger taps taken as 1.0. (22) "Standard deviation of distance ratio at peak closing speed" [-] is the standard deviation of the ratio of distance at the minimum speed in the closing action, with the same ratio.

[0114] Regarding the characteristic quantity [acceleration], the following characteristic quantity parameters are shown by identification numbers (23) to (32). (23) "Maximum amplitude of acceleration" [m / s] 2 ] is the waveform of acceleration ( Figure 8 The difference between the maximum and minimum values ​​of acceleration in (c) of the acceleration. (24) "The average of the maximum points of acceleration" [m / s] 2 [ ] is the average of the maximum values ​​of acceleration during the opening action, and is the first of the four extreme values ​​that occur in one cycle of finger tapping. (25) "Average of the minimum points of opening acceleration" [m / s 2 [ ] is the average of the minimum values ​​of acceleration during the opening action, and is the second of the four extreme values. (26) "Average of the maximum points of closing acceleration" [m / s 2 [ ] is the average of the maximum values ​​of acceleration during the closed motion, and is the third of the four extreme values. (27) "Average of the minimum points of the closed acceleration" [m / s 2 [ ] is the average of the minimum values ​​of acceleration during the closing motion, and is the fourth of the four extreme values.

[0115] (28) "Average contact time" [seconds] is the average contact time when the two fingers are closed. (29) "Standard deviation of contact time" [seconds] is the standard deviation of the contact time. (30) "Coefficient of variation of contact time" [-] is the coefficient of variation of the contact time. (31) "Number of zero-crossings of acceleration" [-] is the average number of positive and negative changes in acceleration during one cycle of finger tapping. Ideally, this value is 2. (32) "Number of stiffnesses" [-] is the value obtained by subtracting the number of finger taps with larger opening and closing from the number of round trips of positive and negative changes in acceleration during one cycle of finger tapping.

[0116] Regarding the characteristic quantity "tapping interval", the following characteristic parameters are shown by identification numbers (33) to (41). (33) "Number of taps" [-] is the number of finger taps in the total measurement time of one measurement. (34) "Average tapping interval" [seconds] is the above tapping interval in the waveform of distance ( Figure 8 (35) The “knock frequency” [Hz] is the frequency with the largest frequency spectrum when the waveform of the distance is subjected to a Fourier transform. (36) The “knock interval standard deviation” [seconds] is the standard deviation of the knock interval.

[0117] (37) "Tapping interval variation coefficient" [-] is the variation coefficient of the tapping interval, which is the value obtained by normalizing the standard deviation to the mean. (38) "Tapping interval variation" [mm] 2 [ ] is the cumulative value of the frequency range of 0.2 to 2.0 Hz under the condition of spectral analysis of the striking interval. (39) "Skewness of striking interval distribution" [-] is the skewness of the frequency distribution of the striking interval, which indicates the degree of skewness of the frequency distribution compared with the normal distribution. (40) "Standard deviation of local striking interval" [seconds] is the standard deviation of the striking interval over three adjacent locations. (41) "Slope (attenuation rate) of the approximate curve of striking interval" [-] is the slope of the curve obtained by approximating the striking interval. This slope mainly indicates the change in striking interval caused by fatigue during the measurement time.

[0118] Regarding the characteristic quantity [phase difference], the following characteristic quantity parameters are shown by identification numbers (42) to (45). (42) "Average phase difference" [degrees] is the average phase difference in the waveforms of both hands. The phase difference is an index value that expresses the deviation of the left hand's fingers from the right hand's fingers as an angle, taking one cycle of the right hand's fingers striking as 360 degrees. The case of no deviation is taken as 0 degrees. The larger the values ​​of (42) and (43), the greater and less stable the deviation of the two hands. (43) "Standard deviation of phase difference" [degrees] is the standard deviation of the phase difference mentioned above. (44) "Similarity of the two hands" [-] is the value of the correlation when the time difference is 0, taking the cross-correlation function applied to the waveforms of the left and right hands. (45) "Time difference of maximum similarity of the two hands" [seconds] is the value of the time difference of maximum correlation of (44).

[0119] Regarding the feature quantity [mark tracking], the following feature quantity parameters are shown with identification numbers (46) to (47). (46) "Average delay time from the mark" [seconds] is the average delay time of the finger tap relative to the time indicated by the periodic mark. The mark corresponds to stimuli such as visual stimuli, auditory stimuli, and tactile stimuli. The value of this parameter is based on the moment when the two fingers are in the closed state. (47) "Standard deviation of delay time from the mark" [seconds] is the standard deviation of the above delay time.

[0120] [Abnormal data detection without using DB]

[0121] The anomaly detection unit 14A, which does not use a database, within the anomaly data processing system 11, will be described. In the anomaly detection unit 14A, anomaly detection without using a database is performed solely based on the measured data, without referring to the single-subject database 45A or the multi-subject database 45B. Specifically, the following anomaly detection items can be cited. In the anomaly detection unit 14, if there is a discrepancy between the characteristics of the expected data and the characteristics of the acquired data, the acquired data is detected as anomaly data. During detection, the acquired measured data itself or the aforementioned characteristic quantities obtained from the measured data can be used. The anomaly detection unit 14, which detects anomaly data corresponding to the anomaly detection items, performs processing corresponding to the detected anomaly data.

[0122] Figure 10 This shows an example of the signal waveform obtained when abnormal data is detected.

[0123] (E1) The user mistakenly measures the same task consecutively.

[0124] If multiple data points corresponding to the same task are input, they are detected as abnormal data. As a corresponding action, the user is prompted on the terminal device 4 screen to choose between the first and second data points. Alternatively, it can be assumed that the second data point exhibits a practice effect due to repeated testing, and the decision is made to always use the first data point, which does not have a practice effect. Alternatively, it can be assumed that the first data point may fail due to unfamiliarity with the testing method, and the decision is made to always use the second data point. Furthermore, the better-performing data point can be selected by referring to the feature values ​​stored in the analysis and evaluation data 43.

[0125] (E2) The measurement of single-handed finger tapping was incorrectly performed using the two-handed finger tapping method.

[0126] This is a case where, although the user intended to measure the tapping motion of a single hand's fingers, the system actually selected tapping with both hands and recorded the measurement data. For the measurement data of each hand, the time during which no movement was performed (hereinafter referred to as "move non-performance time") is calculated. The movement non-performance time can be obtained using the aforementioned feature quantities. For example, it is defined as the period when the "total distance traveled" per unit time is below the specified value TDc. In addition to this definition, it can also be changed to the period when the "total energy" or "number of taps" is below the specified value (14) or (33). Then, if the movement non-performance time of only one hand is above the specified value Tc, it is judged as an abnormal data, indicating that although it is a task performed by both hands, it was incorrectly performed by only one hand. Tc can be predetermined to be two-thirds of the measurement time, for example. In this case, if the measurement time is 15 seconds, it is detected as abnormal data if the movement non-performance time is more than 10 seconds. In the case of abnormal data, the data of the hand with movement non-performance time is ignored, and it is regarded as the measurement data of a single-handed task consisting only of the data of the other hand. Alternatively, instead of processing automatically, the system may ask the user or administrator for confirmation on the screen of terminal device 4.

[0127] (E3) The measurement of finger tapping with both hands was incorrectly performed by selecting the case of single-handed finger tapping.

[0128] This is a case where, although the user intended to measure the tapping motion of both hands' fingers, the system actually selected tapping with only one hand and recorded the measurement data. As will be described later... Figure 12 In the task measurement screen, when tapping with one hand, only the waveform of that hand is displayed to the user. However, since the measurement device 3 has acquired measurement data for both hands, the measurement data for both hands is saved in the background. For each hand in the measurement data for both hands, the time without movement is calculated as described above. If the time without movement in both hands is less than the specified value Tc, it can be determined that the user has performed a two-handed task. Then, the data of the hand that was not displayed to the user is also used as the measurement data for the two-handed task. Alternatively, the process can be automated, and the user or administrator can be prompted for confirmation on the screen.

[0129] (E4) The measurement of simultaneous free movement of both hands was incorrectly selected based on the case of alternating free movement of both hands.

[0130] This is a situation where the user wants to measure simultaneous free movement of both hands, but the measurement system 2 instructs the user to perform alternating free movement of both hands. To detect this anomaly, the feature quantity that evaluates the coordination of both hands from the feature quantity stored in the analysis and evaluation data 43 is used. For example, (42) the "average phase difference" is 0° when both hands move simultaneously and freely in an ideal state of perfect synchronization, and 180° when both hands move alternately in an ideal state of perfect alternating free movement. Therefore, if (42) the "average phase difference" is less than the specified value (e.g., 90°), even if alternating free movement of both hands is selected, it can be considered that the user has performed simultaneous free movement of both hands. That is, it is considered that alternating free movement of both hands was mistakenly selected during the measurement of simultaneous free movement of both hands, and the task data 42A is changed to the measurement data of simultaneous free movement of both hands after the measurement. In this case, it is also possible not to process it automatically, but to ask the user on the screen for confirmation. As an example of using other feature quantities, it can also be considered that both hands are moving freely at the same time if (44) "similarity of both hands" is above a specified value (e.g., 0), and that the absolute value of "the time difference of the largest similarity of both hands" in (45) is less than a specified value (e.g., (34) "average tapping interval" × 0.25). In addition, there are cases where the user wants to move both hands freely at the same time, but cannot make the hands move synchronously, and the measurement data is close to the alternating free movement of both hands. In this case, it is also possible not to automatically treat it as abnormal data, but to ask the user or the manager on the screen for confirmation.

[0131] (E5) The measurement of alternating free movement of both hands was incorrectly performed by selecting the case of simultaneous free movement of both hands.

[0132] This is the opposite of the previous one. Although the user wants to measure the alternating free movement of both hands, the measurement system 2 indicates to the user that both hands are moving freely at the same time. As with the previous one, the judgment can be made based on the characteristic quantity of finger tapping. If the "average phase difference" in (42) is above the specified value (e.g., the midpoint between the ideal value of 0° when both hands are moving freely at the same time and the ideal value of 180° when both hands are moving freely at alternating times, there is a possibility that the user wants to move freely at alternating times even if both hands are moving freely at the same time. That is, it is considered that the simultaneous free movement of both hands was mistakenly selected when measuring the alternating free movement of both hands, and the task data 42A is changed to simultaneous free movement of both hands after the measurement. As an example of using other characteristic quantities, it can also be considered that both hands are moving freely at the same time if the "similarity of both hands" in (44) is less than the specified value (e.g., 0), and that both hands are moving freely at the same time if the "time difference of the maximum similarity of both hands" in (45) is above the specified value (e.g., "average tapping interval" × 0.25 in (34)). Additionally, there are instances where a user intends to perform alternating free hand movements, but instead moves both hands synchronously, resulting in measurement data that approximates synchronous free hand movements. In such cases, instead of automatically treating it as abnormal data, the system may prompt the user or administrator on the screen for confirmation.

[0133] (E6) The situation of two fingers crossing during measurement

[0134] This occurs when the thumb and index finger cross during finger tapping motion measurement, resulting in an outlier distance between the two fingers. When using a magnetic sensor as the motion sensor 20, its properties, such as... Figure 10 (a) shows that the distance between the two fingers may be extrapolated to a very large value during the time period when the fingers are crossed. In order to detect this anomaly, a feature quantity representing the waveform amplitude is used among the feature quantities stored in the analysis and evaluation data 43. For example, if the "maximum amplitude of the distance" in (1) is greater than the specified value (e.g., a value 20 cm greater than the distance between the two fingers of most people) or the maximum value of the distance between the two fingers measured before the measurement, it can be considered that an anomaly of crossing the fingers has occurred. In addition, by extracting the time period greater than the maximum value of the distance between the two fingers from the measurement data 42B (original waveform data) before the feature quantity is calculated, the time period of the anomaly can be determined. In addition, since the phenomenon of crossing the fingers is also part of the nature of the movement of finger tapping, there is also a case where it is better not to consider it as abnormal data and not exclude it. In this case, it is also possible not to automatically process it as abnormal data, but to ask the user or manager on the screen for confirmation.

[0135] (E7) Case where the motion sensor detaches from the finger during measurement.

[0136] This occurs when the motion sensor detaches from a finger during measurement, causing the distance between the two fingers to become an abnormal value. When the motion sensor detaches, as... Figure 10 (b) shows that the distance between the two fingers is calculated to be a very large value. Similar to the previous item, (1) if the characteristic quantity of the waveform amplitude, such as "maximum amplitude of the distance," is greater than the specified value or the maximum value of the distance between the two fingers, and this situation continues until the end of the measurement, it can be considered that the motion sensor has fallen off the finger. If this state continues until the end of the measurement, it can be determined by extracting a period of time greater than the maximum value of the distance between the two fingers from the measurement data, just as in the previous item. In addition, even if the specified measurement time has not ended, when it is determined that the motion sensor has fallen off the finger, the measurement can be ended immediately in real time, and a prompt to remeasure can be made.

[0137] (E8) The case where the movement begins midway through the measurement time.

[0138] This occurs when the user does not correctly interpret the start signal of the measurement and begins movement midway through the measurement period. For example... Figure 10 As shown in (c), if there is a time when the movement is not performed at the beginning of the measurement, it can be determined that the anomaly has occurred. The determination of whether there is a time when the movement is not performed can be achieved by the above method. In addition, the measurement time is divided into a specified number of segments (e.g., 5), and the characteristic quantity of finger tapping is calculated in each segment. When comparing the characteristic quantity between segments, if there is a segment where the value of the characteristic quantity is significantly different from that of other segments, it can be determined that the anomaly has occurred. For example, when the 15-second measurement time is divided into 5 parts, each segment is 3 seconds. When calculating (33) "number of taps" for these segments, it is assumed that the values ​​are 0, 5, 5, 5, and 4 in sequence from the initial segment. In this case, based on the average of these 5 characteristic quantities, it is investigated whether there is a segment that deviates from the direction of decreasing movement by more than N standard deviations (e.g., N=2), and the 0 times of the first segment is consistent. In this way, it is also possible to determine whether there is a time when the movement is not performed.

[0139] (E9) The situation where the motion ends midway through the measurement time.

[0140] This occurs when the user mistakenly ends the exercise before the measurement is complete. For example... Figure 10 As shown in (d), if there is a time period during which movement was not performed at the end of the measurement, the anomaly can be detected. The determination of whether there is a time period during which movement was not performed can be achieved using the method described above. Alternatively, it can also be achieved by dividing the measurement time into N segments and calculating the time period during which movement was not performed, as described above. Furthermore, even if the specified measurement time has not ended, if it is determined that the movement ended midway through the measurement time, the measurement can be immediately terminated in real time, and a prompt to remeasure can be issued.

[0141] (E10) The motion was temporarily interrupted during the measurement period.

[0142] This refers to a situation where the measurement process is temporarily interrupted due to reasons such as cable tangling. For example... Figure 10 As shown in (e), if there is a period of inactivity during the measurement, this anomaly can be detected. The determination of whether there is a period of inactivity can be achieved using the method described above. Alternatively, it can be achieved by dividing the measurement time into N segments and calculating the period of inactivity as described above. Furthermore, even if the specified measurement time has not ended, if it is determined that the movement was interrupted during the measurement time, the measurement can be immediately terminated in real time, and a prompt to re-measure can be issued.

[0143] [Detecting abnormal data using DB]

[0144] The DB-based anomaly detection unit 14B in the anomaly detection unit 14 of the anomaly data processing system 1 will be described. In the DB-based anomaly detection unit 14B, for example, when new data for user X is obtained, it determines whether the new data is abnormal by referring to the single-subject DB45A and the multiple-subject DB45B, which store past data for user X. Specifically, the following anomaly detection items can be cited.

[0145] (E11) The nature of motion changes due to the user's intention

[0146] This refers to situations where the user intentionally doesn't exercise diligently or misunderstands the task instructions, resulting in poor exercise outcomes. This can be detected by comparing the user's past DB (single-subject DB) or the past DB of multiple subjects (multi-subject DB). A large deviation from this comparison will be detected. The method for calculating the deviation will be described later.

[0147] (E12) How the nature of the movement changes due to changes in the user's physical condition

[0148] This includes situations where exercise results worsen due to decreased brain or motor function or extreme fatigue, and situations where exercise results improve due to medication or rehabilitation. This can be detected by comparing the results to the user's own past DB (single-subject DB). The method for calculating the deviation is described later.

[0149] (E13) Cases where someone else impersonates the user

[0150] This refers to situations where someone impersonates the user, thus manifesting changes in the nature of the movement. In such cases, it is also possible to compare the data with the user's own past DB (single subject DB), and detect cases where the deviation is large. The method for calculating the deviation will be described later.

[0151] Calculation of Deviation

[0152] The calculation method for the above deviation is explained. The deviation is calculated by the single-subject DB deviation calculation unit 14Ba, the multi-subject DB deviation calculation unit 14Bb, and the composite deviation calculation unit 14Bc.

[0153] First, the method for calculating the deviation used in both the single-subject DB deviation calculation unit 14Ba and the multi-subject DB deviation calculation unit 14Bb will be explained. N (N≥1) feature quantities of the finger tapping motion are selected, and a data distribution in an N-dimensional space is generated for the DB (single-subject DB45A or multi-subject DB45B). Let the mean of this data distribution be M (= [m1, m2, ..., mN]), and the standard deviation be Σ (= [σ1, σ2, ..., σN]). Then, let the data to be detected as abnormal data be A (= [a1, a2, ..., aN]). The deviation is then calculated as d = |(AM) / Σ|. Here, || represents the absolute value of the vector (the square root of the sum of squares). If d is greater than the specified value dc, it is judged to be a sufficient deviation from the DB. For example, if dc = 1, it can be considered that the measured data is not among the 68.3% of the data close to the average within the DB. Similarly, it can be assumed that if dc=2, the data is not in the 95.5%; and if dc=3, the data is not in the 99.7%. That is, for strict anomaly detection, dc should be set to a smaller value; for lenient anomaly detection, dc should be set to a larger value. Furthermore, the deviation can be defined using methods other than those described above, as long as it represents an indicator of deviation from the database.

[0154] Furthermore, while the above calculation of deviation directly used the feature quantity of finger tapping motion, the feature quantity can be processed to generate new indicators. For example, principal component analysis can be applied to all feature quantities, using the N principal components with the highest contribution rates.

[0155] By observing changes in the characteristic quantities of finger tapping motion, it is possible to identify whether the outcome of the exercise is worsening or improving. For example, if (2) "total distance traveled" and (14) "total energy" and (33) "number of taps" are smaller than the average DB of a single subject, it indicates that the outcome of the exercise is declining. Conversely, if these characteristic quantities are larger than the average DB of a single subject, it indicates that the outcome of the exercise is improving.

[0156] Additionally, to determine the cause of changes in exercise results, a screen can be set up for user input before measurement regarding whether brain or motor function has decreased, whether fatigue has occurred, or whether medication or rehabilitation treatment has been administered. In this case, abnormal detection reasons can be used when the exercise results deteriorate or improve compared to a single subject's DB45A (in... Figure 15 (As explained later in the instructions), or you can refer to its input content. This increases the persuasiveness of the reasons for detecting abnormal data to the user.

[0157] Synthesis of Deviation

[0158] Using the above method, the deviation of the measurement data from the single-subject DB45A or the multi-subject DB45B can be calculated in the single-subject DB45A or the multi-subject DB45B calculation unit 14Ba and 14Bb, respectively. In this regard, since the single-subject DB45A can perform anomaly detection that reflects the individuality of the user's data, it is preferable to use the single-subject DB45A for accurate anomaly detection. Specifically, as follows... Figure 11 As shown in (a), data 1 and data 2, indicated by ★, have the same deviation from user A's single-subject DB45A, and both are correctly detected as anomalous data. However, although single-subject DB45A can be used when the user has performed multiple measurements, it has not been sufficiently accumulated in the initial or limited number of measurements, so multi-subject DB45B is required. Furthermore, Figures 11-13 In this paper, a two-dimensional space consisting of two features is used to schematically illustrate DB and outlier data. However, there can also be one or more features, and a multi-dimensional space corresponding to the number of features used is considered.

[0159] Using the multi-subject DB45B has the advantage of being easy to prepare because data from users other than the current user only needs to be pre-stored. However, the multi-subject DB45B is a collection of data from multiple users and does not reflect the characteristics of that user's data. Therefore, compared to using the single-subject DB45A, it also has the disadvantage of potentially failing to detect outliers correctly. Specifically, for example... Figure 11 As shown in (b), data 1 and data 2, indicated by ★, deviate from the multi-subject DB45B at different degrees. Data 2 is correctly detected as anomalous data, but data 1 is not detected as anomalous data. Based on this awareness, it is believed that a technique is needed that combines the strengths and weaknesses of single-subject DB45A and multi-subject DB45B, utilizing both technologies simultaneously.

[0160] Therefore, instead of choosing to use either the deviation of single-subject DB45A or the deviation of multiple-subject DB45B, a new deviation (composite deviation) is calculated by combining the two. It is believed that by using the composite deviation, it is possible to achieve anomaly detection by compensating for the strengths and weaknesses of using either single-subject DB45A or multiple-subject DB45B.

[0161] The composite deviation is calculated by the composite deviation calculation unit 14Bc. The composite deviation ds is as follows: Figure 12 As shown in (a), the DB confidence coefficient c for a single subject, the DB deviation d1 for a single subject, and the DB deviation d2 for multiple subjects are calculated. The DB confidence coefficient c for a single subject is an indicator of the degree to which a single subject's DB45A can be trusted, as shown in (a). Figure 12 (b) shows values ​​from 0.0 to 1.0. The value is 0.0 when the number of data points for a single subject's DB45A is 0, and gradually approaches 1.0 as the number of data points increases. The formula expressing the relationship between the single subject's DB confidence coefficient c and the number of data points k can be arbitrary as long as c increases with the number of data points. For example, as... Figure 12 (b) shows that the Sigmoid function can be used, set as c = 1 / (1 + exp(-α(k-β)) + γ) (e.g., α = 0.1, β = 50). Here, α is set to a larger value when trusting the single subject DB in stages with fewer data points, and a smaller value when trusting it in stages with more data points. β and γ are adjusted so that c = 0 when k = 0. Using c defined in this way, the composite deviation ds is defined as ds = d1 × c + d2 × (1.0 - c). Then, if ds is larger than the specified value dc, it is judged as a sufficient deviation from the DB. That is, the single subject DB confidence coefficient c is the weight of the single subject DB deviation d1, which increases as the number of single subject DB45A data points increases. In addition, the time-decayed deviation described below can also be used as the single subject DB deviation d1 and the multi-subject DB deviation d2.

[0162] Calculation of Deviation Considering Timing

[0163] In the single-subject DB deviation calculation unit 14Ba, a method is explained for calculating the deviation from a single subject's DB45A, taking into account the temporal relationship of data within the DB (hereinafter referred to as time-decayed deviation). The single-subject DB45A data is accumulated over time through regular measurements taken by the user. However, the user's health status changes daily due to aging, decreased cognitive function, and decreased motor function. Therefore, when detecting abnormal data, more recent data is considered to have higher reliability, while data from further back is considered less reliable. Specifically, as... Figure 13As shown in (a), when the data in the single subject DB45A changes over time as 1→2→3, the average deviation degrees of 4a and 4b from the DB are the same. However, since 4a is farther from the nearest data, which is 3, it should be judged as abnormal. On the other hand, since 4b is closer to the nearest data, which is 3, it should not be judged as abnormal.

[0164] Therefore, as shown in Figure 13 (b), the time-dependent attenuation deviation degree is calculated by increasing the weight for the nearest data and decreasing the weight for the data further back in the past. Specifically, let the data in the single subject DB45A be Bi (= [bi1, bi2,..., biN], i = 1~k (the number of data in the single subject DB45A)). Corresponding to the time ti since the acquisition of the new data, the credibility qi of the past data is defined as qi = p ti (0.0 < p < 1.0). Then, the average of the data distribution in the single subject DB45A is defined as M = q1B1 + q2B2 +... + qkBk. When M is defined in this way, the more recent the data, the higher the trust, and the further back in the past, the lower the trust. Furthermore, using this M, the standard deviation Σ of the data distribution in the DB can be defined as Σ = ((q1B1 - M) 2 + (q2B2 - M) 2 +... + (qkBk - M) 2 ) / k. The single subject DB deviation degree d1 is obtained as described above, using this M and Σ as d = |(A - M) / Σ|.

[0165] The above deviation degree considering the time sequence is explained with respect to the single subject DB45A, but the same calculation can also be performed for the multi-subject DB45B. That is, for each subject in the multi-subject DB45B, the above deviation degree considering the time sequence can be calculated, and their average can be calculated, thereby calculating the multi-subject DB deviation degree d2.

[0166] [Abnormal Data Processing Decision]

[0167] The abnormal data processing decision unit 15 processes the abnormal data detected by the abnormal data detection unit 14A that does not use the DB and the abnormal data detection unit 14B that uses the DB in the abnormal data processing system 1. Figure 14The anomaly detection reason and processing correspondence table 50B within the management table 50 shows the anomaly items detected by the anomaly detection unit 14A (without using DB) and the anomaly detection unit 14B (using DB) in the anomaly data processing system 1. This can be provided in advance when constructing the anomaly data processing system 1, or it can be set by the administrator of the anomaly data processing system 1. Multiple processing options are listed in the processing column, but only one is actually selected for setting. The anomaly data processing decision unit 15 performs processing corresponding to the anomaly detection items detected by the anomaly detection unit 14A (without using DB) and the anomaly detection unit 14B (using DB) based on the anomaly detection reason and processing correspondence table 50B.

[0168] [Execution of abnormal data handling]

[0169] The abnormal data processing execution unit 16 executes the abnormal data processing content 46 determined by the abnormal data processing decision unit 15. If it is decided not to use the measurement data, data is not registered in the single-subject DB45A and multi-subject DB45B. If remeasurement is required, the abnormal data processing content 46 of remeasurement is notified to the terminal device 4 via the communication unit 105. The terminal device 4 accepts the notification and performs remeasurement in cooperation with the measurement device 3. In the case of processing user query data, the query content is notified to the terminal device 4 via the communication unit 105. The terminal device 4 displays the query content on the screen, and the user views the screen and answers. The user's answer is sent to the abnormal data processing system via the communication unit 105. The abnormal data processing execution unit 16 performs processing based on the user's answer.

[0170] [Display screen (1) - Menu]

[0171] Figure 15 This example shows the initial service screen, or menu screen, as an example of the display screen of terminal device 4. This menu screen includes a user information bar 1501, an operation menu bar 1502, and a settings bar 1503, etc.

[0172] User information can be entered and registered by the user in user information field 1501. Additionally, if user information already exists in electronic medical records, it can be used in conjunction with that information. Examples of user information that can be entered include user ID, name, date of birth or age, gender, dominant hand, illness / symptoms, and remarks. Dominant hand can be selected from right hand, left hand, both hands, unknown, etc. Illness / symptoms can be selected from a list box or entered as any text. When this system is used in hospitals or similar settings, it can also be entered by a physician or other professional on behalf of the user. This abnormal data processing system can also be applied to situations where user information is not registered.

[0173] The operation menu bar 1502 displays the operation items for the functions provided by the service. These include "Calibration," "Finger Motion Measurement," "Abnormal Data Detection / Processing," and "End." Selecting "Calibration" performs the aforementioned calibration, i.e., the adjustment of the motion sensor 20 relative to the user's finger. It also displays whether the adjustment has been completed. Selecting "Finger Motion Measurement" navigates to the task measurement screen for measuring finger movements such as tapping. Selecting "Abnormal Data Detection / Processing" navigates to the screen that detects anomalies in the measured data, displays the anomaly detection results, and processes the detected anomalies. Selecting "End" terminates the service.

[0174] User settings can be made in the settings panel 1503. For example, if there are types of anomaly detection items that a user, meter reader, or manager wishes to detect, they can select and set those anomaly detection items from the options. Additionally, the processing corresponding to each anomaly detection item can be selected. Furthermore, thresholds for anomaly data detection can also be set. These settings are sent to the anomaly data processing system 1 via the communication unit 105, and the anomaly data processing system 1 detects and processes anomaly data according to the settings specified here.

[0175] [Display screen (2) - Task measurement]

[0176] Figure 16 This section shows a task measurement screen as an example. The screen displays task information. For example, for the left and right hands, graph 1600 is displayed with time as the horizontal axis and the distance between the two fingers as the vertical axis. Other guidance information explaining the task content can also be output on the screen. For example, a video area explaining the task content with images and sound can be set. The screen has operation buttons such as "Start Measurement," "Re-measure," "End Measurement," and "Save (Register)," which can be selected by the user. The user selects "Start Measurement" according to the task information on the screen and performs the task movement. The measurement device 3 measures the task movement and obtains waveform signals. The terminal device 4 displays the measurement waveform 1602 corresponding to the measured waveform signals on graph 1600 in real time. After the movement, the user selects "End Measurement," and if confirmed, selects "Save (Register)." The measurement device 3 sends the measurement data to the abnormal data processing system 1.

[0177] [Display screen (3) - Evaluation results]

[0178] Figure 17This screen displays the evaluation results as an example. It shows the analysis and evaluation results of the task. This screen is automatically displayed after the task analysis and evaluation. In this example, a radar chart is shown showing the characteristic quantities of the five finger tapping movements (A to E). The solid border 1701 indicates the analysis and evaluation results after the task measurement. The display of characteristic quantities is not limited to a radar chart; it can also be displayed using a prescribed format of graphs. The characteristic quantities can also be converted into a score (e.g., a perfect score of 100). In addition to the characteristic quantity graph, evaluation comments regarding the analysis and evaluation results can also be displayed. The analysis and evaluation unit 13 generates these comments. For example, it may display a message such as "(B), (E) Good". The screen includes operation buttons such as "Overlay past results", "Proceed to anomaly detection / processing", and "End". When "Proceed to anomaly detection / processing" is selected, the anomaly data processing system moves to the anomaly data detection / processing screen; when "End" is selected, it moves back to the initial screen.

[0179] [Display screen (3) - Abnormal data detection / processing]

[0180] Figure 18 This screen displays an anomaly detection / processing screen as another example. It shows the anomaly detection result 44 sent from the anomaly data processing system 1 and its anomaly data processing content 46. (The text then abruptly shifts to a seemingly unrelated topic: "Press...") Figure 11 The "Abnormal Data Detection / Processing" button or Figure 18This screen appears after clicking the "Proceed to Anomaly Data Detection / Processing" button. This screen displays basic information such as user information and measurement data, and also shows the anomaly data detection results. If an anomaly is detected in the measurement data, "Anomaly" is displayed; otherwise, "No Anomaly" is displayed. Furthermore, if an anomaly is detected, the reason for the anomaly data detection is displayed. The reason for the anomaly data detection is included in the anomaly data detection result 44 sent from the anomaly data processing system 1. In this figure, as an example of an anomaly data detection unit 14A that does not use a database, an example is shown: "Because no movement occurred during the measurement time of 0-3 seconds." Below this, a waveform of finger tapping motion is displayed, visually explaining the reason for the anomaly data detection in an easy-to-understand way. Below this, the recommended processing corresponding to the anomaly data detection reason is shown. In this figure, "Remeasurement" is shown as an example. The correspondence table between the anomaly data detection reason and the processing is recorded in the anomaly detection reason and processing correspondence table 50B within the management table 50. Users, measurement personnel, or managers can select the "Execute Recommended Processing" button if they wish to follow the recommended processing shown in this diagram, and select the "Do Not Execute Recommended Processing" button if they do not wish to follow the recommended processing. Alternatively, the exception data processing system 1 can automatically perform the processing and notify the user of the results afterward, instead of allowing the user, measurement personnel, or manager to choose whether to perform the processing.

[0181] Figure 19 This is another example of an anomaly detection / processing screen. In this figure, as an example of an anomaly detection unit 14B using a database, the reason for anomaly detection is set as "because of deviation from the past personal database." Then, as a recommended process corresponding to this anomaly detection reason, "do not register the data in the database" is shown as an example.

[0182] [Effects, etc.]

[0183] According to the anomaly data processing system 1 of embodiment 1, high-precision anomaly data processing can be achieved by using both single-subject DB45A and multi-subject DB45B. This is because the decrease in precision when the number of data in single-subject DB45A is insufficient can be compensated for by increasing the weight of multi-subject DB45B, and the decrease in precision caused by the inability to reflect individual differences when using multi-subject DB45B can be compensated for by increasing the weight of single-subject DB45A.

[0184] Implementation Method 2

[0185] use Figures 20-26 The abnormal data processing system of Embodiment 2 of the present invention will be described below. The basic structure of Embodiment 2 is the same as that of Embodiment 1. The parts of the structure of Embodiment 2 that are different from those of Embodiment 1 will be described below.

[0186] [System (2)]

[0187] Figure 20 This describes a human data measurement system that includes the abnormal data processing system of Embodiment 2. This human data measurement system is installed in hospitals, elderly care facilities, or users' homes. The abnormal data processing system of Embodiment 2 uses a measurement system that is a tablet-type finger tapping motion measurement system. This measurement system consists of a terminal device 5, which is a tablet terminal. In Embodiment 2, motion measurement and information display are performed using the touch panel of the terminal device 5. Embodiment 2 is equivalent to integrating the measurement function of the measurement device 3 of Embodiment 1 and the display function of the terminal device 4 into a single terminal device 5. The terminal device 5 can be a device installed in a facility or a device held by the user.

[0188] Terminal device 5 includes a control unit 501, a storage unit 502, a communication unit 505, and a touch panel 510, which are connected via a bus. Touch panel 510 includes a display unit 511 and a touch sensor 512. The display unit 511 is, for example, a liquid crystal display unit or an organic EL display unit, and has a display screen. The touch sensor 512 is, for example, a capacitive touch sensor, and is disposed in an area corresponding to the display screen. The touch sensor 512 detects changes in capacitive capacitance corresponding to the state of a finger approaching or touching the display screen, and outputs this detection signal to touch detection unit 521.

[0189] The control unit 501 controls the entire terminal device 5 and consists of a CPU, ROM, RAM, etc. Based on software program processing, it implements a data processing unit 500 for handling abnormal data. The structure of the data processing unit 500 is largely the same as in Embodiment 1. The control unit 501 also includes a touch detection unit 521 and a measurement processing unit 522. The control unit 501 performs functions such as obtaining measurement data through the touch panel 510, processing and analyzing the measurement data, and outputting information to the display screen 511 of the touch panel 510. The touch detection unit 521 processes the state of the user's finger approaching or touching the display screen and the state of finger movement based on the detection signal from the touch sensor 512, converting it into touch position coordinates and timing signals. The measurement processing unit 522 uses the detection information from the touch detection unit 521 to measure the position and movement of the finger on the display screen as waveform signals, obtaining measurement data. This measurement data is equivalent to measurement data 42B. The data processing unit 500 performs abnormal data detection and abnormal data processing decisions based on the measurement data through the same processing as in Embodiment 1, and displays the results on the display screen of the display unit 511. In addition, the data processing unit 500 generates analysis and evaluation data, etc., and displays evaluation screens, etc., on the display screen of the display unit 511. The data processing unit 500 includes a user information management unit 11, an abnormal data detection unit 14 consisting of an abnormal data detection unit 14A that does not use a database and an abnormal data detection unit 14B that uses a database, etc. Figure 2 The data processing unit has the same functions as the storage unit. The storage unit 502 has user information 41, task data 42A, measurement data 42B, analysis and evaluation data 43, abnormal data detection results 44, single subject DB45A, multiple subject DB45B, management table 50, abnormal data processing content 46, etc. Figure 2 It has the same function as the storage unit 102.

[0190] [Example of motion and display screen (1)]

[0191] Figure 21This describes a method of finger tapping on the display screen 210 of the terminal device 5. The terminal device 5 can also provide a task using this method. In this method, the control unit 501 displays an area 211 on the background area of ​​the display screen 210 for the two fingers of an object with both hands positioned. For example, the first finger is the thumb and the second finger is the index finger. The user positions the two fingers of each hand in this area 211 in a state of contact or proximity. Although it depends on the touch sensor 512, in this example, the state of the fingers contacting the area 211 of the display screen is basically maintained during this movement. The user taps the area 211 by opening and closing the two fingers. The terminal device 5 measures the finger tapping movement by the touch sensor 512, etc., and obtains measurement data such as waveform signals in the same way as in Embodiment 1. The movement 212 of the first finger and the movement 213 of the second finger in the area 211 are shown by arrows. The distance L between the fingertips of the two fingers is shown as the distance L1 on the left side and the distance L2 on the right side.

[0192] Figure 22 Indicates as and Figure 21 The waveform signal of the distance L between two fingers is an example of the measurement data corresponding to the movement of the fingers tapping. The horizontal axis represents the elapsed time t [seconds], and the vertical axis represents the distance L(t) [mm] at each elapsed time t. In addition, part 221 of the waveform represents the part when the fingers leave the region 211 to a certain extent. In the case of such waveform interruption, a continuous waveform can be obtained by interpolating the waveform when the fingers are in the region 211. In the above manner, the terminal device 5 extracts feature quantities based on the measurement data in the same way as in embodiment 1, performs abnormal data detection, abnormal data processing decision, abnormal data processing execution, and displays the result.

[0193] [Examples of motion and display screen (2)]

[0194] Figure 23 This illustrates the extension method as an example of other finger tapping movements and screen display. Terminal device 5 can also provide tasks using the extension method. Figure 23(a) Indicates a cross-shaped extension. On the display screen 210 of the terminal device 5, the initial position graphic 231 is first displayed. Measurement begins when the target finger, such as the index finger, is placed on the initial position graphic 231. After the start, the target graphic 232, such as a cross, corresponding to the mark is displayed on the display screen 210. The control unit 501 displays the graphic 232 at different positions at a predetermined period, for example. The user taps the finger by extending the finger in a manner that tracks the position of the graphic 232. In this example, the state of tapping the finger at a position 233 with a deviation from the center position of the graphic 232 is shown. There is a distance E between the center position of the target graphic 232 and the tapping or touching position 233 corresponding to the deviation. Based on the measurement data, the terminal device 5 calculates the distance E and the delay time TD, etc., as one of the characteristic quantities. The delay time TD is the time from the moment when the target graphic 232 is displayed in the standby state when the finger is placed on the initial position graphic 231 until the moment when the finger comes into contact with the target graphic 232.

[0195] Figure 23 (b) Represents a circular extension. A circular area is displayed as the target area for graphic 234. The user taps the circular area of ​​graphic 234 with their finger. As a feature, for example, the distance between the center of graphic 234 and the tapping location is extracted.

[0196] [Examples of motion and display screens (3)]

[0197] Figure 24 This demonstrates continuous touch as an example of the movement of other fingers tapping and the display of an image. Terminal device 5 can also provide tasks and exercises using continuous touch. Figure 24 (a) Represents continuous single-handed touch. A graphic 241, such as a circular area, for contact with the left thumb is displayed on the display screen 210, for example, near the lower left. The user continuously touches the displayed graphic 241 with their finger. When the graphic 241 is not displayed, the user removes their finger from it. The control unit 501 controls the display of the graphic 241. For example, the display and non-display of the graphic 241 are switched at predetermined intervals, for a predetermined number of times. Alternatively, auditory stimulation or other guidance information can be applied along with the display of the graphic 241. Feature quantities include, for example, the number of touches to the graphic 241, the touch interval, and the touch delay time.

[0198] Figure 24(b) Simultaneous continuous touch with both hands. Graphics 242 representing the touch positions of the fingers of the left and right hands are displayed at two locations on the display screen 210. The user touches these displayed graphics 242 simultaneously with both hands at the same time. Similarly, alternating continuous touch with both hands is also possible. In this case, the control unit 501 switches between displaying the left and right graphics 242 alternately. The user touches these graphics 242 with both hands at alternating times. For example, the phase difference between the touches of the left and right graphics 242 is extracted as a feature quantity.

[0199] As an example of other sports, instead of displaying graphics, auditory stimuli or other forms of guidance can be output. For instance, two sounds can be output at predetermined intervals for when it is appropriate to touch and when it is inappropriate to touch.

[0200] [Examples of motion and display screens (4)]

[0201] Figure 25 This describes a tapping motion that, as an example, displays an image and is accompanied by light. Terminal device 5 can also provide tasks that utilize this method. Figure 25 (a) Represents single-handed tapping. On display screen 210, a tapping graphic 251 of the left-hand finger and a light graphic 252, serving as a visual stimulus, indicate the timing of the tapping of graphic 251. The control unit 501 flashes the graphic 252, switching it between display and non-display. The user taps the tapping graphic 251 at the timing of the display of graphic 252. As an example of other movements, auditory stimulation, i.e., sound output, may be used instead of the visual stimulus graphic 252, or continuous touch may be employed. A characteristic quantity is, for example, the time difference between the moment of tapping or touching and the moment of occurrence of the periodic stimulus. This time difference corresponds to the delay time from the moment graphic 252 is displayed until the moment graphic 251 is tapped. Figure 25 (b) Similarly, the case of simultaneous tapping with both hands is shown. Two tapping graphics 251 are set on the left and right sides, and the two visual stimulus graphics 252 are flashed at the same time on both sides. Similarly, when tapping with both hands alternately, the control unit 501 causes the two graphics 252 on the left and right sides to flash at the alternate time.

[0202] [Examples of motion and display screens (5)]

[0203] Figure 26This describes a five-finger tapping method, which serves as an example of the movement and display of other fingers. The terminal device 5 can also provide tasks using this five-finger tapping method. In this method, the five fingers of the user's hand are used. The terminal device 5 displays graphics 261 on the background area of ​​the display screen 210, indicating how to tap with the five fingers of each hand, totaling ten fingers. The user initially touches the display screen 210 with their five fingers. Based on the detection of this touch position, the terminal device 5 automatically adjusts the display position of the graphics 261. The terminal device 5 controls the display of the graphics 261 at each position. The terminal device 5 sets the graphics 261 at the position where they should be tapped to a specific display state (e.g., indicated by a black circle), and sets other graphics 261 at positions where they should not be tapped to other display states. The terminal device 5 controls the switching of the display states of the graphics 261. The user taps the graphics 261 with their fingers in accordance with the display of the graphics 261 that should be tapped.

[0204] [Characteristics]

[0205] Examples of the characteristic quantities specific to Implementation Method 2 are described below.

[0206] As characteristic parameters relating to the extension method, the following are provided: (2-1) "Average delay time from target display" [seconds] is the average of the aforementioned delay time. (2-2) "Standard deviation of delay time from target display" [seconds] is the standard deviation of the aforementioned delay time. (2-3) "Average position error relative to the target" [mm] is the average of the aforementioned distance E. (2-4) "Standard deviation of position error relative to the target" [mm] is the standard deviation of the aforementioned distance E.

[0207] The following are characteristic parameters related to continuous single-handed touch: (2-5) Number of taps [-], (2-6) Average tap interval [seconds], (2-7) Tap frequency [Hz], (2-8) Standard deviation of tap interval [seconds], (2-9) Coefficient of variation of tap interval [-], (2-10) Variation of tap interval [mm] 2 (2-11) "Skewness of the tapping interval distribution" [-], (2-12) "Standard deviation of the local tapping interval" [seconds], (2-13) "Tapping interval attenuation rate", etc. The definitions of each characteristic quantity are the same as in Implementation Method 1.

[0208] As characteristic parameters concerning the continuous touch method of both hands, the following are described. (2-14) "Average phase difference" [degrees] is the average phase difference of the touch of both hands, etc. (2-15) "Standard deviation of phase difference" [degrees] is the standard deviation of the aforementioned phase difference.

[0209] As characteristic parameters relating to the manner of touching or tapping in conjunction with light or sound stimuli, the following are provided: (2-16) "The average value of the time difference relative to the stimulus" [seconds] is the average value of the time difference mentioned above. (2-17) "The standard deviation of the time difference relative to the stimulus" [degrees] is the standard deviation of the time difference mentioned above.

[0210] [Abnormal data detection without using DB]

[0211] An anomaly detection unit 14A, which does not use a database (DB), will be described within the anomaly detection unit 14 of the data processing unit 500. Similar to Embodiment 1, in the anomaly detection unit 14A, anomalies are determined solely based on measured data without referring to a DB. Basically, the same anomaly detection items as in Embodiment 1 can be cited; only items unique to this embodiment will be described below. Furthermore, regarding the anomalies described in Embodiment 1, since in this embodiment, when using both hands, the location to be touched by both hands is visually indicated, (E2) is unlikely to occur. Additionally, since (E6) is a phenomenon unique to magnetic sensors, it can be disregarded in this embodiment of touchscreens.

[0212] (E4) The measurement of simultaneous tapping with both hands was incorrectly selected as alternating tapping with both hands.

[0213] This is a situation where the user intended to measure the movement of tapping with both hands simultaneously, but actually selected alternating tapping and recorded the measurement data in the system. To detect this anomaly, the feature quantity that evaluates the coordination of both hands from the feature quantity stored in the analysis and evaluation data 43 is used. For example, (2-14) "Average phase difference" is 0° when both hands move in perfect synchronization during ideal simultaneous tapping, and 180° when both hands move in perfect alternating tapping. Therefore, if the "Average phase difference" in (2-14) is less than the specified value (e.g., 90°), even if alternating tapping is selected, it can be considered that the user intended to tap with both hands simultaneously. That is, it is considered that alternating tapping was incorrectly selected during the measurement of simultaneous tapping, and the task data 42A is changed to the measurement data of simultaneous tapping after the measurement.

[0214] (E5) The measurement of alternating hand tapping was incorrectly performed by selecting the case of simultaneous hand tapping.

[0215] This is the opposite of the previous point. Although the user intended to measure the movement of alternating hand tapping, the system actually selected simultaneous hand tapping and recorded the measurement data. Similar to the previous point, the judgment can be based on the characteristic quantity of finger tapping. If the "average phase difference" in (2-14) is above a specified value (e.g., 90°), even if simultaneous hand tapping is selected, it can be considered that the user intended to perform alternating hand tapping. That is, it is considered that simultaneous hand tapping was incorrectly selected during the measurement of alternating hand tapping, and the task data 42A is changed to the measurement data of simultaneous hand tapping after the measurement.

[0216] (E7) The situation where the finger leaves the designated area during the measurement.

[0217] This refers to the situation where a finger touches a location that leaves a designated area on the screen during the measurement. If there is a period of time during which no movement is performed, this can be considered an anomaly. Specifically, for example, the period of time during which no movement is performed can be evaluated as a time when (2-5) [number of taps] is 0. Alternatively, the measurement can be stopped in real time before the end of the specified measurement period, prompting a re-measurement. Alternatively, instead of detecting it as an anomaly, touches around the designated area can also be detected, thereby visually or audibly guiding the user back to the correct location if the touched area leaves the designated area. The periods of time during which no movement is performed in (E8), (E9), and (E10) can be evaluated as a time when (2-5) [number of taps] is 0, similar to this anomaly detection item.

[0218] [Effects, etc.]

[0219] The abnormal data processing system according to Embodiment 2, similar to Embodiment 1, enables high-precision abnormal data detection by simultaneously using both the single-subject DB45A and the multi-subject DB45B. In particular, Embodiment 2 eliminates the need for motion sensors 20, saving users the trouble of measurement.

[0220]

Implementation Method 3

[0221] use Figures 27-29 The abnormal data processing system of Embodiment 3 of the present invention will be described below. The basic structure of Embodiment 3 is the same as that of Embodiment 1. The parts of the structure of Embodiment 3 that are different from those of Embodiment 1 will be described below.

[0222] [System (3)]

[0223] Figure 27This describes the anomaly data processing system of Implementation 3. The anomaly data processing system includes a server 6 serving an operator and a system 7 containing multiple facilities connected via a communication network 8. The communication network 8 and server 6 may also include a cloud computing system. The anomaly data processing system of Implementation 3 is configured by the terminal devices 4 of system 7 and server 6 sharing the workload. The sharing will be described later.

[0224] The facility can be a hospital or medical examination center, a public facility, an entertainment facility, or even a user's own home. A system 7 is installed within the facility. Examples of system 7 for a facility include system 7A for hospital H1 and system 7B for hospital H2. For instance, each hospital's system 7A and system 7B has a measuring device 3 and a terminal device 4 that constitute the same measuring system 2 as in embodiment 1. The structures of each system 7 can be the same or different. The facility's system 7 can also include a hospital's electronic medical record management system, etc. The measuring device for system 7 can also be a dedicated terminal.

[0225] Server 6 is a device managed by the service operator. Server 6 provides the same anomaly data processing service as the anomaly data processing system 1 in Embodiment 1 to facilities and users as an information processing service. Server 6 provides service processing to the measurement system in a client-server manner. In addition to this function, Server 6 also has user management functions, etc. The user management function is the function of registering and storing user information, measurement data, and analysis and evaluation data of user groups obtained through the system 7 of multiple facilities in a database for management. Furthermore, the terminal device 5 in Embodiment 3 does not require its own anomaly data processing function, but has measurement functions using a touch panel and display functions for displaying anomaly data detection results generated by Server 6.

[0226] [server]

[0227] Figure 28 This describes the structure of server 6. Server 6 has a control unit 601, a storage unit 602, an input unit 603, an output unit 604, and a communication unit 605, which are connected via a bus. The input unit 603 is used for inputting operations performed by the administrator of server 6. The output unit 604 is used for displaying images to the administrator of server 6. The communication unit 605 has a communication interface and is used for communication processing with the communication network 8. DB 640 is stored in the storage unit 602. DB 640 can also be managed by a different DB server than server 6.

[0228] The control unit 601 controls the entire server 6 and consists of a CPU, ROM, RAM, etc. Based on software program processing, it implements a data processing unit 600 that performs anomaly detection and anomaly processing decisions. The data processing unit 600 includes a user information management unit 11, a task processing unit 12, an analysis and evaluation unit 13, an anomaly detection unit 14, an anomaly processing decision unit 15, an anomaly processing execution unit 16, and a result output unit 17. The anomaly detection unit 14 differs from that in Embodiment 1; it does not include an anomaly detection unit 14A that does not use a database, but only includes an anomaly detection unit 14B that uses a database.

[0229] The User Information Management Department 11 manages user information for user groups of System 7 across multiple facilities by registering it as User Information 41 in DB640. User Information 41 includes each user's individual attribute values, usage history information, user settings information, etc. Usage history information includes information on each user's past usage of the abnormal data processing service.

[0230] [Server Management Information]

[0231] Figure 29 This is an example of the data structure for user information 41 managed by server 6 in DB640. The table for user information 41 contains user ID, facility ID, facility-internal user ID, gender, age, illness, severity score, symptoms, and historical information. The user ID is the unique identification information for each user in this system. The facility ID is the identification information for the facility where system 7 is located. Additionally, the communication addresses of the measuring devices in each system 7 are also managed separately. The facility-internal user ID is the user identification information if user identification information exists within that facility or system 7. That is, user IDs are managed in association with facility-internal user IDs. The illness and symptom items store values ​​representing illnesses and symptoms selected and entered by the user, or values ​​obtained from diagnoses by doctors or other diagnoses in hospitals. The severity score is a value representing the degree of illness.

[0232] The historical information item manages information about a user's past service usage and anomaly handling, saving information such as the date and time of each usage in chronological order. Additionally, the historical information item stores data from each practice session, including the aforementioned measurement data, analysis and evaluation data, anomaly detection results, and anomaly handling details. The historical information item can also store information about the storage addresses of these data.

[0233] [Distribution of workload for abnormal data detection between local and server]

[0234] In Embodiment 1, the abnormal data detection unit 14 executes both an abnormal data detection unit 14A (without using a database) and an abnormal data detection unit 14B (using a database). In contrast, in this embodiment, the abnormal data detection unit 14A (without using a database) is implemented in the terminal device 4 of the local system 7, while the abnormal data detection unit 14B (using a database) is implemented in the server 6. The reason for this division is that the server 6, based on data collected from multiple systems 7 (7A, 7B, ...), constitutes a single-subject database 45A and a multi-subject database 45B, making abnormal data detection using a database suitable. On the other hand, to perform abnormal data detection that does not require these databases as quickly as possible, it is preferable to execute it in the local terminal device 4. By performing abnormal data detection in the local terminal device 4, abnormal data detection can be performed in real time if an anomaly occurs during measurement, and a remeasurement instruction can be issued immediately. Furthermore, when the network connection with the server is not continuous, time loss from sending data to the server and waiting for abnormal data detection results can be prevented.

[0235] Furthermore, the method described above for sharing the abnormal data detection function between terminal device 4 and server 6 based on whether or not a database is used can also be used. For example, if there are many users visiting hospital H1, and a large-scale database can be built, then the abnormal data detection unit 14B using the database can also be executed in terminal device 4. In addition, if the system administrator is given the authority to change the settings of the abnormal detection reasons in management table 50 and the corresponding processing table 50B, the abnormal data detection unit 14A without using the database can also be executed in server 6.

[0236] [Effects, etc.]

[0237] The anomaly data processing system according to Embodiment 3, similar to Embodiment 1, achieves high-precision anomaly data detection by simultaneously using both the single-subject DB45A and the multi-subject DB45B databases. Furthermore, by managing both the single-subject DB45A and the multi-subject DB45B with a server, data from multiple facilities can be collected to construct a large-scale database, which is believed to enable even more accurate anomaly data detection. Additionally, by having the local terminal device 4 and the server 6 share the anomaly data detection function, anomaly data detection without time loss can be achieved.

[0238] The present invention has been specifically described above based on the embodiments, but the present invention is not limited to the above embodiments and various modifications can be made without departing from its spirit.

[0239] This invention is not limited to the embodiments described above, and includes various modifications. For example, a portion of the structure of one embodiment can be replaced with the structure of another embodiment, or the structure of another embodiment can be added to the structure of one embodiment. Furthermore, for a portion of the structure of each embodiment, other structures can be added, deleted, or replaced.

[0240] Industrial availability

[0241] It can be used for information processing service technologies.

[0242] Explanation of reference numerals in the attached figures

[0243] 1…Abnormal data processing system, 2…Measuring system, 3…Measuring device, 4…Terminal device.

Claims

1. An anomaly data processing system for detecting and processing new data for anomalies, characterized in that, include: The storage unit is used to store a multi-subject database that stores data from multiple subjects and a single-subject database that stores data from a single subject. A single-subject database deviation calculation unit is used to calculate the single-subject database deviation as the degree to which the new data deviates from the single-subject database; A multi-subject database deviation calculation unit is used to calculate the multi-subject database deviation degree as the extent to which the new data deviates from the multi-subject database; The single-subject database deviation calculation unit uses a single-subject database reliability coefficient that increases with the number of data in the single-subject database, and uses the single-subject database reliability coefficient to weight the single-subject database deviation to obtain the single-subject database deviation. The composite deviation calculation unit uses the number of data points in the single-subject database to calculate the composite deviation obtained by combining the deviation of the single-subject database with the deviation of the multi-subject database; and An anomaly detection unit that does not use a database does not rely on the single-subject database or the multi-subject database. Instead, it determines whether the new data is abnormal based on the new data itself or features derived from the new data. The new data is judged to be anomaly based on the synthesized deviation. The new data is finger movement data. The abnormal data detection unit that does not use a database will use at least one of the waveform amplitude, the time when the movement was not performed, and the coordination of both hands calculated from the finger movement data as the feature quantity, and detect whether the feature quantity deviates from the specified numerical range.

2. The abnormal data processing system as described in claim 1, characterized in that, include: An exception data processing decision unit is used to determine the handling when the new data is judged to be abnormal; and An exception data processing execution unit is used to perform the aforementioned processing.

3. The abnormal data processing system as described in claim 1, characterized in that: The single-subject database deviation calculation unit has the following functions: The function calculates the time decay deviation of the single subject database by calculating the difference between the measurement time of the new data and the measurement time of each data in the single subject database, and calculates the reliability of the past data as the larger the difference in measurement time is.

Images