JP2023160425A - Autonomic nerve function evaluation system and autonomic nerve function evaluation method - Google Patents

Abstract

To achieve autonomic nerve function evaluation taking into account wearability, readiness and reliability.SOLUTION: An autonomic nerve function evaluation system includes a processor and a storage device. The storage device holds biological data measured for a predetermined length of time and an autonomic nerve function index calculated on the basis of biological data measured before the predetermined length of time. The processor determines propriety of the biological data measured for the predetermined length of time on the basis of the biological data measured for the predetermined length of time and the autonomic nerve function index calculated on the basis of the biological data measured before the predetermined length of time, and calculates the autonomic nerve function index on the basis of the data determined to be proper of the biological data measured for the predetermined length of time.SELECTED DRAWING: Figure 2

Description

The present invention relates to a quick-response (real-time) biometric data evaluation technology used to improve the reliability of biometric data measured in a short period of time when promoting safe operation of transportation facilities.

In recent years, quantitative evaluation of the biological condition of drivers in the transportation industry has been carried out to prevent health-related accidents. Among biological conditions, autonomic nervous function evaluation is performed based on the measurement of beat-to-beat interval (BBI) data, which is the interval between heartbeats, using various types of heartbeat sensors that are easy to measure. Immediate response of analysis and reliability of data are important for autonomic nerve function evaluation during work.

For example, in Japanese Unexamined Patent Application Publication No. 2020-130335 (Patent Document 1), the subject is time-series data of biological signals that have periodicity and have missing sections caused by measurement abnormalities, etc., and that have periodicity measured within an arbitrary time. Accurately calculate temporal features focusing on the change tendency of biological signals. Specifically, the instantaneous heart rate abnormal value is recognized from the measurement status of two adjacent R waves, excluded from the time series data, and the fluctuation trend of the R-wave interval (R-R Interval, RRI) measured within a certain period of time is calculated. After performing interpolation using a linear function approximating Calculate the parasympathetic nerve activity index obtained from

Japanese Patent Application Publication No. 2020-130335

It is important to be able to manage autonomic nervous function evaluations during work in real time, and a measurement and evaluation system that comprehensively considers the wearability of sensors, the quick response of analysis, and the reliability of results is necessary.

During work that involves driving, the person to be evaluated (for example, the driver) is not necessarily in a resting state. Furthermore, when measuring heartbeat interval data using a wearable sensor that emphasizes wearability, measurement errors are likely to occur. In the method described in Patent Document 1, it may not be possible to appropriately determine whether the measured instantaneous heartbeat interval is abnormal using data within a short measurement time. If the measurement time is set for a long time, the responsiveness of the evaluation will deteriorate. Furthermore, on the wearable sensor side, there may be a storage capacity problem in order to hold a huge amount of heartbeat interval data for a long time.

In order to solve at least one of the above-mentioned problems, the present invention is an autonomic nerve function evaluation system, which includes a processor and a storage device, and the storage device is configured to perform measurement over a predetermined length of time. and an autonomic nerve function index calculated based on the biological data measured before the predetermined length of time, and the processor stores biological data measured before the predetermined length of time, of the biological data measured during the predetermined length of time, and an autonomic nerve function index calculated based on the biological data measured before the predetermined length of time. The method is characterized in that the validity is determined and the autonomic nerve function index is calculated based on the data determined to be valid among the biological data measured during the predetermined length of time.

Therefore, according to one aspect of the present invention, the validity of the data can be appropriately determined even for heartbeat interval data acquired in a short period of time, and autonomic nerve function evaluation can be performed in consideration of wearability, quick response, and reliability. It becomes possible.

The details of at least one implementation of the subject matter disclosed herein are set forth in the accompanying drawings and the description below. Other features, aspects, and advantages of the disclosed subject matter will be apparent from the following disclosure, drawings, and claims.

1 is a block diagram showing an example of the main configuration of a biometric data evaluation system according to an embodiment of the present invention. FIG. 1 is a functional block diagram showing an example of the configuration of a quick response autonomic nerve function evaluation system according to an embodiment of the present invention. 7 is a flowchart illustrating an example of processing by a correction processing section according to an embodiment of the present invention. It is an explanatory view showing an example of a heartbeat interval measured in an example of the present invention, and an autonomic nerve function index based on it. It is an explanatory view showing an example of a heartbeat interval measured in an example of the present invention, and an autonomic nerve function index based on it. It is an explanatory view showing an example of a heartbeat interval measured in an example of the present invention, and an autonomic nerve function index based on it. It is an explanatory view showing an example of a heartbeat interval measured in an example of the present invention, and an autonomic nerve function index based on it. FIG. 2 is an explanatory diagram showing an example of a list of multivariate autonomic nerve function indices measured in an example of the present invention. FIG. 2 is an explanatory diagram showing an example of heartbeat interval data held by the quick response autonomic nerve function evaluation system according to the embodiment of the present invention. FIG. 2 is an explanatory diagram showing an example of autonomic nerve function index data including multivariate autonomic nerve functions held by the quick-response autonomic nerve function evaluation system according to the embodiment of the present invention. FIG. 2 is an explanatory diagram showing an example of the internal configuration of an ANF index integrated DB held by the quick-response autonomic nerve function evaluation system according to the embodiment of the present invention. FIG. 6 is an explanatory diagram showing an example of validity determination based on LP within a predetermined analysis window in an embodiment of the present invention. FIG. 6 is an explanatory diagram showing an example of validity determination based on LP within a predetermined analysis window in an embodiment of the present invention. FIG. 6 is an explanatory diagram showing an example of validity determination based on LP within a predetermined analysis window in an embodiment of the present invention. FIG. 6 is an explanatory diagram showing an example of a method for calculating a valid region when a reference ANF index can be extracted from an ANF index integrated DB in an embodiment of the present invention. FIG. 6 is an explanatory diagram showing an example of a method for calculating a valid region when a reference ANF index can be extracted from an ANF index integrated DB in an embodiment of the present invention. FIG. 6 is an explanatory diagram showing an example of a process in which the confidence interval extraction unit according to the embodiment of the present invention performs reliability evaluation of analysis time data and extracts a confidence interval. It is an explanatory diagram showing an example of the display of the evaluation result of the present time by the quick-response type autonomic nerve function evaluation system concerning the example of the present invention. FIG. 2 is an explanatory diagram showing an example of display of time-series evaluation results by the immediate autonomic nerve function evaluation system according to the embodiment of the present invention.

Embodiments of the present invention will be described below with reference to the drawings.

<System configuration>
FIG. 1 is a block diagram showing an example of the main configuration of a biological data evaluation system according to an embodiment of the present invention.

The biometric data evaluation system of this embodiment includes a quick-response autonomic nerve function evaluation system 1 that processes biometric data received from one or more vehicles 7 via a network.

The vehicle 7 includes a GNSS (Global Navigation Satellite System) 8A that measures the position of the vehicle 7, an inter-vehicle distance sensor 8B that measures the distance between the vehicle 7 and vehicles traveling in front of and behind it, and a speedometer 8C that measures the speed of the vehicle 7. , an acceleration sensor 8D that measures the acceleration of the vehicle 7, a camera 8E that photographs the surroundings of the vehicle 7, and data on the measurement results and data on the photography results are collected from these and transmitted to the quick response autonomic nervous function evaluation system 1. It has an in-vehicle information collection device 10.

Furthermore, the vehicle 7 includes a biological sensor 11, a biological data collection device 12, a sensor information input section 13, a sensor information collection device 14, a user information setting section 15, a user information collection device 16, a work content setting section 17, and a work information collection section. A device 18 is installed. These may be attached to the vehicle 7, or may be carried by a user (for example, a driver of the vehicle 7) and brought into the vehicle 7 when driving the vehicle 7.

The biosensor 11 includes a sensor that detects heartbeat (RRI or PPI) based on electrocardiography, pulse waves, heart sounds, or the like.

In addition to a heartbeat sensor, the biosensor 11 can employ a sensor that detects sweat amount, body temperature, blinking, eye movement, myoelectricity, brain waves, or the like. The biosensor 11 may include a wearable device that the driver can wear, a sensing device attached to any part of the vehicle 7 such as the steering wheel, seat, or seatbelt, or a sensor that captures an image of the driver's expression or behavior. An image recognition system or the like that analyzes the information can be used.

The biological data collection device 12 collects information measured by the biological sensor 11 and transmits it to the immediate response autonomic nerve function evaluation system 1 .

The sensor information input unit 13 has a function of receiving input of information regarding the biosensor 11. The information regarding the biosensor 11 may include, for example, information regarding the type, model, version, specifications, setting values, etc. of the biosensor 11. The sensor information collection device 14 collects the information input to the sensor information input unit 13 and transmits it to the immediate response autonomic nerve function evaluation system 1 .

The user information setting section 15 has a function of accepting settings of information regarding a user. The information regarding the user may include information such as the user's name, identification information, age, gender, affiliation, and work in charge, for example. The user information collection device 16 collects the information set in the user information setting section 15 and transmits it to the immediate response autonomic nerve function evaluation system 1 .

The business content setting section 17 has a function of accepting settings of information regarding the user's business. The information regarding the user's work may include, for example, information such as the content of the work performed by the user (for example, any one of driving, loading, unloading, and resting) and the time period during which the user performed the work. The work information collection device 18 collects the information set in the work content setting section 17 and transmits it to the immediate response autonomic nervous function evaluation system 1 .

The biosensor 11, the biodata collection device 12, the sensor information input unit 13, the sensor information collection device 14, the user information setting unit 15, the user information collection device 16, the work content setting unit 17, and the work information collection device 18 are each Although they may be independent devices, at least some of them may be realized by one device. To give a typical example, the biosensor 11 is a sensor device worn by the user, and the biometric data collection device 12 is a terminal device carried by the user (user terminal) or a terminal device attached to the vehicle 7 (in-vehicle terminal). There may be. Further, the sensor information input unit 13, the sensor information collection device 14, the user information setting unit 15, the user information collection device 16, the business content setting unit 17, and the business information collection device 18 are realized as applications of a user terminal or a vehicle-mounted terminal. Good too.

Although only one vehicle 7 and a set of biosensor 11 to business information collection device 18 are shown in FIG. 1, an actual biometric data evaluation system includes multiple vehicles 7 and multiple sets of biometric sensors corresponding to multiple users. The sensor 11 to the business information collection device 18 may also be included. The type of biosensor 11 used may differ depending on the user, and even for the same user, the type of biosensor 11 may differ depending on the day.

The immediate response autonomic nerve function evaluation system 1 is a computer including a processor 2, a memory 3, a storage device 4, an input/output device 5, and a communication device 6. The memory 3 includes a heartbeat interval acquisition unit 21 that acquires heartbeat intervals within a unit time according to sensor information, a correction processing unit 22 that performs correction processing to improve the reliability of heartbeat interval data, and a heartbeat interval acquisition unit 22 that performs correction processing to improve the reliability of heartbeat interval data. It includes an ANF index calculation unit 23 that calculates an autonomic nervous function (ANF) index for interval data, and a data integration unit 24 that integrates ANF index data, user information, sensor information business information, and the like. The integrated data is accumulated in the ANF index integration DB 48. The memory 3 further includes an evaluation result display section 25 that comprehensively evaluates the integrated ANF index and displays the evaluation results. The memory 3 further includes a setting section 26 (see FIG. 2) for setting display items and evaluation results.

The correction processing unit 22 includes a reference ANF index extraction unit 31 that extracts a reference ANF index from the ANF index integrated DB 48, and a validity determination unit 32 that determines the validity of heartbeat interval data based on a Lorentz plot (LP). 2, the same applies hereafter), a noise removal/interpolation processing unit 33 that removes noise and interpolates the heartbeat interval data based on the determination result, and evaluates the reliability of data within a predetermined time using the processed data. It has a confidence interval extraction unit 34 that extracts an interval.

Note that the functions of the heartbeat interval acquisition section 21, correction processing section 22, ANF index calculation section 23, data integration section 24, evaluation result display section 25, and setting section 26 in this embodiment are actually performed by the processor 2 in the memory 3. This is accomplished by executing instructions written in a stored program. That is, in the following explanation, the processes executed by the above-mentioned units are actually executed by the processor 2.

The storage device 4 stores heart rate sensor measurement data 41, user information 42, sensor information 43, business information 44, in-vehicle information 45, short-time BBI data 46, ANF index data 47, and ANF index integrated DB 48. The heartbeat sensor measurement data 41 includes data measured by the biological sensor 11 and transmitted by the biological data collection device 12. The user information 42 includes data set by the user information setting unit 15 and transmitted by the user information collection device 16. The sensor information 43 includes data input to the sensor information input section 13 and transmitted by the sensor information collection device 14. The business information 44 includes data set by the business content setting unit 17 and transmitted by the business information collection device 18. The in-vehicle information 45 includes information collected and transmitted by the in-vehicle information collection device 10. This information includes, for example, information acquired from the GNSS 8A, inter-vehicle distance sensor 8B, speedometer 8C, acceleration sensor 8D, and camera 8E.

Short-term BBI data 46 includes BBI data for a short period of time (for example, 2 minutes). This data includes, for example, 2-minute heartbeat interval (BBI) data acquired by the heartbeat interval acquisition unit 21 from data transmitted from the biological data collection device 12. The ANF index data 47 includes data resulting from correction processing performed by the correction processing section 22 and calculation of the ANF index by the ANF index calculation section 23 for the BBI data. The ANF index integration DB 48 includes data that integrates ANF indexes, information regarding users, information regarding operations, and the like.

<Processing details>
FIG. 2 is a functional block diagram showing an example of the configuration of a quick response autonomic nerve function evaluation system 1 according to an embodiment of the present invention.

Heart rate sensor measurement data, sensor information, user information, business information, and in-vehicle information acquired on the user terminal side and the vehicle 7 side are transmitted to the immediate response autonomic nerve function evaluation system 1 via the network 19. In order to evaluate autonomic nervous function in real time during work, the acquisition time of sensor measurement data is shortened to ensure immediate response. As an example, every time sensor measurement data for two minutes is acquired, it is transmitted from the biological data collection device 12 to the immediate autonomic nerve function evaluation system 1 and stored in the storage device 4 as heartbeat sensor measurement data 41. Similarly, data transmitted from the sensor information collection device 14, user information collection device 16, business information collection device 18, and in-vehicle information collection device 10 are sensor information 43, user information 42, business information 44, and in-vehicle information 45, respectively. It is stored in the storage device 4 as .

The heartbeat interval acquisition unit 21 identifies the type of heartbeat sensor that measured the heartbeat sensor measurement data 41 by referring to the sensor information 42 corresponding to the data, and performs shaping according to the identified heartbeat sensor type. Processing is performed and converted into a common format of BBI data of unit analysis time. The correction processing unit 22 performs data validity determination, noise removal/interpolation processing, and highly reliable interval extraction processing on the acquired heartbeat interval data.

In the correction processing unit 22, a reference ANF index extraction unit extracts an ANF index to be referenced from the ANF index integrated DB 48 accumulated up to the time when the newly acquired short-term heart rate sensor measurement data 41 is analyzed. 31 is a feature of this embodiment. By using the accumulated reference ANF index to determine the validity area of the measurement data (that is, the area of data determined to be valid) for heartbeat interval data acquired over a short period of time that is likely to contain noise, short-time data improve reliability.

Specifically, the correction processing unit 22 includes a validity determination unit 32. The validity determining unit 32 determines the validity of the data area on the LP, and returns the data determined to be valid to the time series, thereby determining the validity of each piece of measured heartbeat interval data.

The noise removal/interpolation processing unit 33 of the correction processing unit 22 performs interpolation processing on data determined to be valid, and performs interpolation processing on data that can be interpolated based on the tendency of the acquired data, and removes data that cannot be interpolated with noise. Remove as. Further, the correction processing section 22 includes a confidence interval extraction section 34. The confidence interval extraction unit 34 checks the position of occurrence of defects in the time series for the data subjected to noise removal and interpolation processing, checks the continuity of good intervals with no (or few) defects, and determines the continuity and A reliability score is calculated taking into account the time length of usable data, and a confidence interval suitable for calculating the ANF index is extracted.

Further, the ANF index calculation unit 23 calculates the ANF index for the corresponding analysis time based on the corrected heartbeat interval data. The data integration unit 24 integrates the ANF index, user information, business information, and in-vehicle information and stores them in the ANF index integration DB 48.

The user or administrator can comprehensively evaluate the analyzed ANF index data via the evaluation result display section 25 and check the evaluation result at the current time or the change trend over time. Furthermore, the evaluation result display section 25 is connected to the setting section 26, and display items, basic information changes, analysis conditions, etc. may be changed by user input or the like.

FIG. 3 is a flowchart showing an example of the processing of the correction processing section 22 according to the embodiment of the present invention.

The heartbeat interval acquisition unit 21 performs a data shaping process (step 303), obtain standard time heartbeat interval data (BBI) for a predetermined short period (for example, 2 minutes). It is difficult to analyze heart rate variability based only on one BBI (ie, the interval between one beat and the next beat), and an analysis time window of a predetermined length is required. When emphasis is placed on quick response, it is preferable that the predetermined time is short, while when emphasis is placed on reliability of analysis, it is preferable that the predetermined time is longer than the time width required for heart rate variability analysis.

The validity determining unit 32 of the correction processing unit 22 determines whether advanced noise removal is necessary according to the heart rate sensor information input for a predetermined period of time and the quality of the acquired data (step 304). If it is determined that advanced removal is not necessary, a basic outlier determination is performed (step 305). On the other hand, if it is determined that advanced noise removal is necessary, the reference ANF index extraction unit 31 accesses the ANF index integrated DB 48 accumulated up to the analysis time (step 306), and extracts the reference target ANF index that satisfies certain extraction conditions. Extract. As an example of the extraction condition, data for the most recent predetermined time (for example, 30 minutes) in the same work state as the current analysis time of the same user may be extracted as the ANF index to be referenced. For example, if the predetermined short time is 2 minutes and the most recent predetermined time is 30 minutes, 15 sets of ANF indicators are extracted.

Note that the above-mentioned "most recent predetermined time in the same business state as the current analysis time of the same user" is an example of an extraction condition, and other extraction conditions can also be set. For example, data on users close in age to the user to be analyzed or data on users of the same gender may be extracted. Alternatively, resting state data of the same user may be extracted, or the biosensor 11 with high measurement accuracy may be used for the same user (or other users, or users with similar conditions such as age and gender). You may also extract data measured using

Furthermore, examples of the reference ANF index include the average value of RRI, the standard deviation SD1 in the Y=-X direction related to LP, and the standard deviation SD2 in the Y=X direction.

Regarding the time series data T[t], the LP displays the time series data T[t] of the preceding time t (t=0, 1, 2, ...) on the x-axis, and the period dt (dt) from time t on the y-axis. This is a two-dimensional plot for performing chaos analysis drawn by plotting time-series data T[t+dt] at time t+dt after which time (=1, 2, 3, . . . ) has elapsed. In this embodiment, an LP in which T[t] is plotted on the x-axis and T[t+dt] is plotted on the y-axis is shown as an example.

By analyzing the geometric figures drawn on the LP, the characteristics of the time series data T[t] can be analyzed. In particular, it is known that when an LP is drawn for RRI data in a continuous period, it is possible to evaluate measurement characteristics such as the extent to which the data measured in a predetermined period contains arrhythmia or poor measurement based on its geometric characteristics. ing. In this embodiment, an example of LP when dt=1 time will be shown. Furthermore, a case is shown in which the original series RRI[t] of the measured RRI data is used as time series data.

If the reference ANF index can be extracted (step 307: Yes), the validity determination unit 32 determines the validity of the acquired BBI based on the reference index and the LP of the data to be analyzed (step 308). Specific validity determination will be explained later using FIGS. 7A to 8B. On the other hand, an example of a case where it is not possible to extract a reference target ANF index that satisfies the predetermined extraction conditions is a case where data of the corresponding user cannot be extracted from the ANF index integrated DB 48 when the user initially wears a sensor. If the reference index cannot be extracted (step 307: No), the validity determination unit 32 refers only to the BBI data in the predetermined analysis window, checks the geometric area on the LP, and determines the validity of the data. (step 309). A specific calculation method will be explained later with reference to FIGS. 7A to 7C.

Next, the noise removal/interpolation processing unit 33 performs interpolation processing on the data whose validity has been determined, and removes the remaining abnormal data as noise (step 310). Specifically, the noise removal/interpolation processing unit 33 returns the data whose validity has been determined on the LP to the RRI data on the time series, and removes the abnormal data on the time series (i.e., the data that is not valid on the LP). Check the occurrence status and occurrence section of data determined to be If the time of the abnormal data occurrence section is short and it is determined that interpolation is possible based on the RRI data of several beats in the normal range from the change trend of RRI data before and after the analysis section, the noise removal/interpolation processing unit 33 Perform interpolation processing. The noise removal/interpolation processing unit 33 removes the data in the section that remains as an abnormal value after the interpolation process as noise, and sets it as a missing section in the time series.

Thereafter, the confidence interval extraction unit 34 extracts a confidence interval for calculating the ANF index depending on the occurrence status of the missing interval (step 311). An example of a specific method for calculating the confidence interval will be described later with reference to FIG.

Correction processing was performed through a series of processes: reference ANF index extraction (step 306), validity determination by LP (step 308 or 309), noise removal/interpolation processing (step 310), and confidence interval extraction (step 311). Using the BBI data, the ANF index calculation unit 23 calculates the ANF index at the relevant analysis time and holds it as ANF index data 47 (step 312). The data integration unit 24 integrates the calculated ANF index with the user information 42, sensor information 43, business information 44, in-vehicle information 45, etc., and stores it in the ANF index integration DB 48 (step 313). The quick-response autonomic nerve function evaluation system 1 performs a comprehensive evaluation based on the integrated ANF index integrated DB 48, and displays the evaluation results of the analysis at real time or in chronological order via the evaluation result display section 25. and display it on the input/output device 5.

FIGS. 4A to 4D are explanatory diagrams showing examples of heartbeat intervals and autonomic nerve function indicators based on the heartbeat intervals measured in the embodiment of the present invention.

FIG. 4A shows an example of RRI (R-R Interval) that measures the interval between R waves of an electrocardiogram. FIG. 4B shows an example of PPI (Peak-to-Peak Interval) that measures the interval between peaks of pulse waves. The heartbeat interval BBI used to calculate the ANF index in this embodiment may be either the RRI or PPI described above. However, although RRI is generally highly accurate because it measures the heartbeat itself, it imposes a heavy burden on the user, such as the need to wear a belt or underwear equipped with electrodes in order to perform the measurement. On the other hand, PPI can also be measured using, for example, a wristwatch-type pulse sensor, which reduces the burden on the user, but tends to be less accurate, with many measurement defects such as noise.

FIG. 4C shows an example of a change in heartbeat interval BBI (RRI or PPI) on the time axis. FIG. 4D shows frequency analysis based on the BBI data in a predetermined analysis time window (e.g. 2 minutes) in FIG. shows. Here, LF is the power spectrum of a predetermined relatively low frequency band (for example, 0.04 Hz to 0.15 Hz), and HF is the power spectrum of a predetermined relatively high frequency band (for example, 0.15 Hz to 0.4 Hz). , and the total power TP is the sum of LF and HF.

FIG. 5 is an explanatory diagram showing an example of a list of multivariate autonomic nerve function indices measured in the example of the present invention.

Based on the BBI data within the analysis time window, many indicators in the frequency domain, time domain, or nonlinear domain can be calculated to comprehensively evaluate the autonomic nervous function. Here, we will explain some representative indicators.

Among the frequency domain indices, LF, HF, and TP are as described with reference to FIG. 4D, and LF/HF is the ratio of LF and HF. LF and HF are used as indicators related to sympathetic and parasympathetic nerves, respectively. Among the time-domain indices, RRI is the interval between R waves of electrocardiogram as described above. Among the nonlinear region indexes, SD1 is the standard deviation of the LP in the Y=-X direction, SD2 is the standard deviation of the LP in the Y=X direction, and SD1/SD2 is the ratio thereof.

FIG. 6A is an explanatory diagram showing an example of heartbeat interval data held by the immediate autonomic nerve function evaluation system 1 according to the embodiment of the present invention.

Specifically, the heartbeat interval data shown in FIG. 6A is an example of data included in the short-time BBI data 46. The RRI data measured at the measurement time includes a list of combinations of user ID 601, vehicle ID 602, heart rate sensor 603, measurement time 604, and RRI 605. Although an example in which the heartbeat interval data includes RRI 605 is shown here, this is an example of BBI data, and PPI may be held instead of RRI.

The user ID 601 and the vehicle ID 602 are information that identifies the user to be measured and information that identifies the vehicle 7 in which the user rides. The heartbeat sensor 603 is information that identifies the heartbeat sensor used for measurement (for example, information that identifies at least one of the type, model, or individual of the heartbeat sensor). One line (one record) of heartbeat interval data corresponds to one heartbeat, measurement time 604 is the appearance time of each heartbeat (for example, R wave or pulse wave peak, etc.), and RRI 605 is the time from the previous heartbeat to the current heartbeat. Indicates the time interval until. Here, the RRI 605 is data on a scale of about several hundred milliseconds.

For example, the first row of Figure 6A shows that a certain R wave was measured at 10:00:00.000 on April 1, 2020, and the interval between that R wave and the previous R wave was 750 ms. It shows that In the second row, the next R wave after the R wave in the first row was measured at 10:00:00.755 on April 1, 2020, and that R wave and the previous R wave (i.e., in the first row) R wave) was 755 milliseconds.

FIG. 6B is an explanatory diagram showing an example of autonomic nerve function index data including multivariate autonomic nerve functions held by the quick response autonomic nerve function evaluation system 1 according to the embodiment of the present invention.

Specifically, the autonomic nervous function (ANF) index data within a unit time shown in FIG. 6B is an example of data included in the ANF index data 47. One row (one record) of ANF index data within a unit time shown in FIG. 6B includes an ANF index calculated from BBI data for one predetermined short time (2 minutes in the example of FIG. 6B). User ID 611, vehicle ID 612, and heart rate sensor 613 correspond to user ID 601, vehicle ID 602, and heart rate sensor 603 shown in FIG. 6A, respectively. The date and time 614 indicates a time representative of each predetermined short time period (for example, the time of the starting point).

The autonomic nerve function index 615 is an ANF index calculated from BBI data for each predetermined short time period, and may include, for example, at least one of the ANF indexes shown in FIG. 5. The reliability score 616 indicates the value of a reliability score that is an index for evaluating the reliability of the ANF index calculated from BBI data for each predetermined short time period. A method for calculating the reliability score will be described later with reference to FIG. 9. Reference time 617 indicates the time length of past ANF index data referred to for determining the validity of BBI data. For example, if past ANF index data is not referenced, the value of the reference time 617 is 0. Further, when referring to ANF index data for the past 30 minutes, the value of the reference time 617 is 30.

Noise removal method 618 indicates the method used to determine the validity of BBI data. For example, when validity determination is performed based on LP as described later, the value of the noise removal method 618 becomes LP. The interpolation method 619 indicates the method (eg, moving average, etc.) used to interpolate the BBI data that was determined to be invalid (ie, noise) and removed.

FIG. 6C is an explanatory diagram showing an example of the internal configuration of the ANF index integrated DB 48 held by the quick response autonomic nerve function evaluation system 1 according to the embodiment of the present invention.

The ANF index integrated DB 48 is data that is an integrated list of user information 621, vehicle information 622, heart rate sensor information 623, time information 624, work information 625, autonomic nerve function index 626, reliability score 627, and application method 628. Show the configuration.

Here, the user information 621, vehicle information 622, heartbeat sensor information 623, and time information 624 correspond to the user ID 611, vehicle ID 612, heartbeat sensor 613, and date and time 614 shown in FIG. 6B, respectively. The business information 625 is information related to the business that each user was doing at each time, extracted from the business information 44. The autonomic nerve function index 626 and reliability score 627 correspond to the autonomic nerve function index 615 and reliability score 616 shown in FIG. 6B, respectively. Application method 628 corresponds to noise removal method 618 and interpolation method 619 shown in FIG. 6B.

7A to 7C are explanatory diagrams showing an example of validity determination based on LP within a predetermined analysis window in an embodiment of the present invention.

Specifically, FIGS. 7A to 7C illustrate the processing performed in step 309 of FIG. 3. In this case, the validity determination unit 32 generates an LP by pairing the BBI data before and after the BBI data within a predetermined time, performs basic outlier removal, and confirms the data distribution within the initial normal range. As an example of the initial normal range, the median value of BBI data within the relevant time period may be calculated, and the range from median x 0.75 to median x 1.75 may be set as the initial normal range. In that case, BBI data outside the initial normal range is removed as an outlier.

In addition, as an example of a valid data area on an LP, an ellipse shape along the Y=X direction where data is intensively distributed is virtually formed on the LP, and its internal area (ellipse area) is It may be set as a reasonable area. In order to define an ellipse region based on the data concentration state, the ellipse center, the ellipse's vertical axis, and the ellipse's horizontal axis are specified. Here, the axis of the ellipse in the y=-x direction is referred to as the vertical axis, and the axis in the y=x direction is referred to as the horizontal axis. Project all points within the initial normal range onto the y=x and y=-x axes. Let m be the average distance between the projection point of the y=x axis and the origin, and let the position on the y=x axis at a distance m from the origin be the center position of the ellipse. The average BBI

Then, the center m in the y=x direction is

corresponds to

Further, as an example of the horizontal axis of the ellipse (y=x direction) a, the standard deviation in the y=x direction is assumed to be twice the standard deviation std2. In the case of data that follows a normal distribution, this corresponds to a range that contains approximately 96% of the data. As an example of the ellipse vertical axis (y=-x direction) b, the standard deviation in the y=-x direction is assumed to be twice std1. A on the horizontal axis of the ellipse (y=x direction) is not necessarily larger than b on the vertical axis (y=−x direction).

The standard deviation in the y=x direction refers to the standard deviation of data when the target data group is projected onto the straight line of y=x.

The standard deviation in the y=-x direction refers to the standard deviation of data when the target data group is projected onto the straight line of y=-x.

FIG. 7B is an example of LP of BBI data, and each point on the plane of FIG. 7B is a plot of BBI data. For example, when the heartbeat interval data shown in FIG. 6A is obtained, the value of the horizontal axis PPI _t is 750 milliseconds and the value of the vertical axis PPI _t+1 is 755 milliseconds, based on the data in the first and second rows. Points are plotted. Ellipse 701 has a center, a horizontal axis length, and a vertical axis length determined by the method described above. The inside of this ellipse 701 becomes a valid region. FIG. 7C shows the state after removing BBI data outside the ellipse 701 as invalid.

FIGS. 8A and 8B are explanatory diagrams showing an example of a method for calculating a valid region when a reference ANF index can be extracted from the ANF index integrated DB 48 in the embodiment of the present invention.

Specifically, FIGS. 8A and 8B illustrate the processing performed in step 308 of FIG. 3. The ANF index that satisfies the extraction conditions in the past is referred to, for example, the ANF index of the unit time of n points (for example, n=15) of the same user in the same work state. The shape of the valid region is an ellipse as in FIGS. 7A to 7C. As an example of the reference ANF index, MeanNN, Poincare SD1, and Poincare SD2 representing the average value of RRI in the ANF index list of FIG. 5 are referred to.

The validity determination unit 32 determines a new validity region by fusing the distribution of BBI data on the LP within the analysis target time and the extracted ANF index. As an example, the center of the ellipse is calculated as follows. In other words, calculate the weighted average of n points of MeanNN, multiply it by a coefficient, calculate the average with the average value of BBI within the analysis target time (i.e., the above value (1)), and add √2 to the calculated average. Multiplying, the center position of y=x is calculated (Equation (3)).

Regarding the horizontal axis of the ellipse (y=x direction), a weighted average of SD2 of n points is calculated, multiplied by a coefficient, and the average with std2 within the analysis target time is calculated (Formula (4)).

Regarding the vertical axis of the ellipse (y=-x direction), a weighted average of SD1 of n points is calculated, and the average is calculated with std1 within the analysis target time by multiplying by a coefficient (Formula (5)).

The weight w of the reference object may be set to the same value for all n points, or a point closer to the current time of the analysis object (in other words, newer measurement data) may be given a larger weight. The former corresponds to a moving average, and the latter corresponds to a weighted moving average. Further, the weight α of the current time to be analyzed is set to a value of 0.0 to 1.0, and the smaller this value is, the more important the trend of the past reference ANF index is.

For the center, horizontal axis, and vertical axis, a weight set with the same tendency may be assigned, or a different weight set may be assigned for each item.

After determining the range of the appropriate region, the validity determining unit 32 determines data within the region as good data and data outside the region as abnormal values.

FIG. 8A shows an example of the result of determining validity using only BBI data within a predetermined short time without using the reference ANF index extracted from the ANF index integrated DB 48. On the other hand, FIG. 8B shows an example of the results of determining validity using the reference ANF index extracted from the ANF index integrated DB 48 in addition to the BBI data within the same predetermined time period as in the case of FIG. 8A. shows.

If only BBI data within a predetermined period of time is used, the number of samples obtained is small, so the overall distribution is greatly biased due to a small number of noises, as shown by the ellipse 801 in FIG. 8A, and as a result, the calculation of the ANF index is BBI data that should be used for ANF index calculation may be removed as an abnormal value, or BBI data that should not originally be used for ANF index calculation may be used as valid data for ANF index calculation. On the other hand, when the reference ANF index is used, the value of the ANF index is based on BBI data acquired over a certain length of time in the past. It is possible to set a reasonable area where the influence of noise is small. As a result, BBI data suitable for calculating the ANF index can be obtained.

Note that in the above example, the center of the ellipse of the valid region is calculated based on the average value of the BBI data, but the median value may be used instead of the average value.

The noise removal/interpolation processing unit 33 returns the good data whose validity has been determined on the LP to the RRI data on the time series, and confirms the occurrence status and occurrence section of the abnormal data on the time series. When the time of the abnormal value generation section is short and the noise removal/interpolation processing unit 33 can perform interpolation using RRI data for several beats in the normal range based on the change tendency of RRI data before and after the analysis section, Interpolation processing is performed (step 310 in FIG. 3). The noise removal/interpolation processing unit 33 removes data in the section remaining as an abnormal value as noise, and sets the section as a missing section in the time series.

FIG. 9 is an explanatory diagram showing an example of a process in which the confidence interval extraction unit 34 according to the embodiment of the present invention performs reliability evaluation of analysis time data and extracts a confidence interval.

Specifically, FIG. 9 explains the process executed in step 311 of FIG. 3. Through the validity determination and noise removal/interpolation processing using LP, the position where the defect occurs within the analysis target time becomes clear. The confidence interval extraction unit 34 searches for continuous intervals of good data (that is, data determined to be valid on the LP and data interpolated by interpolation processing) according to the defect situation, and determines the reliability. By evaluating the score, confidence interval data suitable for ANF index calculation is extracted, and the reliability of the ANF index is improved.

As a specific search method, an example in which the analysis window width is divided into two will be explained. It is most desirable that all of the 2-minute BBI data be obtained in good condition, but if there are abnormal values or missing values, generally if the continuous interval of good data is 60 seconds or more, it will be considered when calculating the ANF index. , has a correlation of more than 80% compared to the original good data. In this case, that is, when the 2-minute BBI data includes a continuous interval of good data for 60 seconds or more, a method of extracting the data of the continuous interval as confidence interval data is referred to as shortening method 1.

Next, if it is not possible to extract one long continuous section depending on the location of the defect, if you have two or more short continuous sections of 30 seconds or more, you can combine them to achieve an accuracy close to that of shortening method 1. It was confirmed through experiments that the ANF index of can be calculated. In this case, ie, a method of extracting data of an interval in which two or more continuous intervals of good data of 30 seconds or more and less than 60 seconds are connected as confidence interval data is referred to as shortening method 2.

Furthermore, if the missing points are scattered and it is difficult to extract a section that satisfies both the conditions of shortening method 1 and shortening method 2, it is possible to obtain data of 80 seconds or more by connecting intermittent normal RRI data. If possible, the accuracy will be close to that of shortening method 1. This method, that is, a method of extracting data of an interval of 80 seconds or more by connecting a plurality of consecutive intervals of good data of less than 30 seconds as confidence interval data is referred to as shortening method 3.

Even if the total time length of good data is the same, if shortening method 1 can be applied, if shortening method 1 cannot be applied and shortening method 2 is applied, and if both shortening methods 1 and 2 In cases where shortening method 3 is applied without being able to be applied, the reliability decreases little by little in that order (that is, the case where shortening method 1 can be applied has the highest reliability). The confidence interval extraction unit 34 takes into account the occurrence of defects and the amount of available data, multiplies the reliability weight according to the applied shortening method by the total time of the extracted confidence interval, and analyzes the value. By dividing by the time of the window width, a reliability score representing the reliability of the analysis target interval is calculated. The reliability weight value corresponding to shortening method 1 is the largest, and that corresponding to shortening method 3 is the smallest. For example, the reliability weights for shortening methods 1, 2, and 3 may be set to 1.0, 0.9, and 0.8, respectively, but other values may be used.

The ANF index calculation unit 23 calculates the ANF index using the confidence interval data extracted by the shortening method with the highest reliability score (step 312 in FIG. 3). The processing after step 312 is as described with reference to FIG.

FIG. 10 is an explanatory diagram showing an example of the display of the evaluation results at the current time by the quick response autonomic nerve function evaluation system 1 according to the embodiment of the present invention.

In this embodiment, current time and time series can be selected as display modes. FIG. 10 shows an example where the current time is selected.

In the basic information display section 1001, the user ID, vehicle ID, heart rate sensor, work status at the current time, and stress level comprehensively evaluated based on the multivariate ANF index are displayed as basic data of the displayed object. Additionally, the reliability score of the section of analysis data acquired at the current time is displayed. Further, as information regarding data correction processing, whether or not past ANF indexes are referenced, the reference time when referenced, whether LP is applied, and whether or not interpolation processing is performed are displayed.

In the graph display section 1002, the current time and the most recently comprehensively evaluated stress/relaxation level are displayed.

The setting change unit 1003 changes the basic settings by changing the user ID setting, changing the sensor, changing the business information, changing the ANF index reference, changing the reference time, changing the noise removal method, changing the interpolation method, etc. through input settings. can.

When ANF index list display 1004 is clicked, a list of multivariate ANF index values at the current time is displayed in a table format. Although omitted in FIG. 10, for example, information in the table shown in FIG. 6B may be displayed.

By clicking the graph display item specification 1005, it is possible to specify each ANF index in addition to the stress level of the comprehensive evaluation, and display the current time and recent change trend of the specified ANF index.

FIG. 11 is an explanatory diagram showing an example of display of time-series evaluation results by the quick-response autonomic nerve function evaluation system 1 according to the embodiment of the present invention.

This example shows a time series display on a daily basis. When time series is selected as the display mode, the information shown in FIG. 11, for example, is displayed.

The basic information display section 1101 is almost the same as the basic information display section 1001 in the current time display mode. However, in the basic information display section 1101, an average value over a predetermined relatively long period (for example, one day) is displayed as the stress level.

Setting change 1102, ANF index list display 1103, and graph display item designation 1104 are the same as the setting change section 1003, ANF index list display 1004, and graph display item designation 1005 in the current time display mode, respectively.

In the graph display section 1105, the transition of the autonomic nerve function index in units of a predetermined period (for example, one day) is displayed.

In addition, a chronological work status 1106 such as operation, loading, unloading, rest, rest, etc. is also displayed. Further, based on the reliability score of the data acquired in a predetermined short period of time, the reliability level 1107 of the acquired data in time series and the ANF index calculated based on the acquired data is displayed.

By aligning the cursor 1108 with a display point on the graph, detailed display of the time corresponding to that display point can also be confirmed. Although detailed display is omitted in FIG. 11, for example, information on the table shown in FIG. 6B at the relevant time may be displayed.

Further, the system according to the embodiment of the present invention may be configured as follows.

(1) An autonomic nervous function evaluation system, which includes a processor (for example, processor 2) and a storage device (for example, at least one of memory 3 and storage device 4), and the storage device has a predetermined length. Autonomic nerve function index (for example, ANF index included in ANF index integrated DB 48) calculated based on biological data measured at time (e.g. BBI data) and biological data measured before a predetermined length of time and the processor stores biological data measured during a predetermined length of time and an autonomic nervous function index calculated based on biological data measured before the predetermined length of time. Based on this, the validity of the biometric data measured during a predetermined length of time is determined (for example, step 308), and the data determined to be valid among the biometric data measured during a predetermined length of time is determined. An autonomic nerve function index is calculated based on (for example, step 312).

With this, for example, ANF indicators dynamically calculated for biological data such as heart rate interval data measured from a wearable sensor worn during work are stored in the ANF indicator integrated DB, and a reference ANF indicator is extracted. It is possible to appropriately determine the validity of biological data acquired in a short period of time, and it is possible to evaluate autonomic nerve function in consideration of wearability, quick response, and reliability.

(2) In (1) above, the biological data is heartbeat interval data (for example, BBI data), and the processor calculates the heartbeat interval data based on the Lorentz plot of the heartbeat intervals measured over a predetermined length of time. Determine the validity of the heartbeat interval data measured at the time.

Thereby, the validity of the heartbeat interval data can be appropriately determined.

(3) In (2) above, the processor determines that heartbeat interval data within an elliptical area (for example, ellipse 802) set on the Lorentz plot of heartbeat intervals measured over a predetermined length of time is valid. The center of the ellipse area is calculated based on the average or median value of the heartbeat intervals measured over a predetermined length of time, and the biological data measured before the predetermined length of time. Among the neurological function indexes, the diameter of the elliptical area is calculated based on the average value or median value of the heartbeat intervals (for example, formula (3)), and the Lorentz plot of the heartbeat intervals measured over a predetermined length of time is calculated. Calculated based on the above standard deviation and the standard deviation on the Lorentz plot of the heartbeat interval among the autonomic nervous function indexes calculated based on biological data measured before a predetermined length of time ( For example, formulas (4) and (5)).

That is, by virtually setting an elliptical area on the Lorentz plot, drawing the elliptical area based on the state of concentration of biological data, and determining that the heartbeat interval data within that range is appropriate, It is also possible to appropriately set the validity range of heartbeat interval data for biometric data acquired in .

(4) In (3) above, the processor performs calculations based on the average or median value of heartbeat intervals measured over a predetermined length of time and biological data measured before the predetermined length of time. Among the autonomic nerve function indexes, the weighted average of the average value or median value of the heartbeat intervals is calculated as the center of the ellipse area (for example, formula (3)), and the heartbeat intervals measured over a predetermined length of time are calculated. The weighted average of the standard deviation on the Lorentz plot of the heartbeat interval and the standard deviation on the Lorentz plot of the heartbeat interval among autonomic nervous function indicators calculated based on biological data measured before a predetermined length of time. A value obtained by multiplying by a predetermined coefficient (for example, 2) is calculated as the diameter of the ellipse area (for example, Equations (4) and (5)).

With this, it is possible to appropriately set the validity range of heartbeat interval data even for biological data acquired in a short period of time.

(5) In (4) above, greater weight is given to the value extracted from the autonomic nerve function index calculated based on newer biological data.

With this, it is possible to appropriately set the validity range of heartbeat interval data even for biological data acquired in a short period of time.

(6) In (2) above, if the interval of data determined to be invalid based on the Lorentz plot is shorter than a predetermined standard among heartbeat interval data measured over a predetermined length of time, Interpolate the heartbeat interval data for the section based on the heartbeat interval data before and after the section (for example, step 310), and determine the data determined to be valid based on the Lorentz plot and the interpolated data as good data. do.

Thereby, it is possible to effectively utilize the acquired data to perform reliable evaluation of autonomic nerve indicators.

(7) In (6) above, the processor extracts a confidence interval from one or more intervals of continuous good data among the heartbeat interval data measured over a predetermined length of time (for example, in step 311, 9) Calculate a reliability score for evaluating the reliability of the confidence interval based on at least one of the length of an interval where good data is continuous and the total length of multiple intervals where good data are continuous. (eg step 312).

Thereby, it is possible to effectively utilize the acquired data to perform reliable evaluation of autonomic nerve indicators.

(8) In (7) above, the processor extracts a confidence interval from the continuous interval of good data based on at least one of the first extraction method, the second extraction method, and the third extraction method. , the first extraction method is a method of extracting one interval longer than a predetermined first continuous interval length as a confidence interval, and the second extraction method is a method of extracting one interval longer than a predetermined first continuous interval length as a confidence interval. A third extraction method is a method of extracting two or more intervals that are longer than a predetermined continuous interval length, and the sum of their lengths is longer than a predetermined third continuous interval length, as a confidence interval. is a method of extracting two or more intervals whose total length is longer than the length of the third continuous interval from among intervals with continuous good data as confidence intervals, and the length of the second continuous interval is equal to the length of the first continuous interval. The third continuous section length is shorter than the continuous section length, and the third continuous section length is longer than the first continuous section length, and the processor extracts two or more of the first extraction method, the second extraction method, and the third extraction method. When confidence intervals are extracted by the method, an autonomic nerve function index is calculated based on the heartbeat interval data of the confidence interval with the highest reliability score among the two or more extracted confidence intervals.

Thereby, it is possible to effectively utilize the acquired data to perform reliable evaluation of autonomic nerve indicators.

(9) In (8) above, the processor calculates the reliability score by multiplying the ratio of the length of the confidence interval to the predetermined length of time by the reliability weight according to the confidence interval extraction method. , the reliability weight according to the first extraction method is the largest, and the reliability weight according to the third extraction method is the smallest.

This makes it possible to appropriately evaluate the reliability of the acquired data.

(10) In (2) above, the storage device holds information regarding the sensor that measured the biometric data (for example, sensor information 43), and the processor stores the sensor information for the biometric data measured over a predetermined length of time. A formatting process is performed according to the format (for example, step 303), and the validity of the biometric data subjected to the formatting process is determined.

Thereby, it is possible to appropriately generate data used for evaluation of autonomic nerve indicators depending on the type of sensor and the like.

(11) In (2) above, the processor applies the calculated autonomic nerve function index to the identification information of the person whose heartbeat interval data is to be measured, the age of the person, the gender of the person, and the vehicle in which the person rides. Autonomic nerve function index integrated data (for example, ANF index integrated DB 48 ), and the autonomic nerve function index extracted from the autonomic nerve function index integrated data based on the specified extraction conditions (for example, the index referred to in step 306) is stored in the storage device as a predetermined length of time. Used to determine the validity of measured biometric data.

This makes it possible to appropriately determine the validity of biometric data acquired in a short period of time.

(12) In (2) above, the heartbeat interval data includes R wave interval data (for example, FIG. 4A) based on electrocardiogram measurement values.

This allows highly accurate heartbeat interval data to be obtained.

(13) In (2) above, the heartbeat interval data includes data on peak intervals of pulse wave measurement values (for example, FIG. 4B).

This can reduce the burden on the person being measured during measurement.

Note that the present invention is not limited to the above-described embodiments, and includes various modifications. For example, the above-described embodiments have been described in detail for better understanding of the present invention, and the present invention is not necessarily limited to having all the configurations described. Furthermore, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Further, it is possible to add, delete, or replace a part of the configuration of each embodiment with other configurations.

Further, each of the above-mentioned configurations, functions, processing units, processing means, etc. may be partially or entirely realized in hardware by designing, for example, an integrated circuit. Further, each of the above configurations, functions, etc. may be realized by software by a processor interpreting and executing a program for realizing each function. Information such as programs, tables, and files that realize each function is stored in storage devices such as nonvolatile semiconductor memory, hard disk drives, and SSDs (Solid State Drives), or computer-readable non-volatile devices such as IC cards, SD cards, and DVDs. It may be stored on a temporary data storage medium.

Further, the control lines and information lines are shown to be necessary for explanation purposes, and not all control lines and information lines are necessarily shown in the product. In reality, almost all configurations may be considered to be interconnected.

1 Immediate response autonomic nerve function evaluation system 2 Processor 3 Memory 4 Storage device 5 Input/output device 6 Communication device 7 Vehicle 11 Biosensor 19 Network 21 Heartbeat interval acquisition section 22 Correction processing section 23 ANF index calculation section 24 Data integration section 25 Evaluation result Display unit 26 Setting unit 31 Reference ANF index extraction unit 32 Validity determination unit 33 Noise removal/interpolation processing unit 34 Confidence interval extraction unit 41 Heart rate sensor measurement data 42 User information 43 Sensor information 44 Business information 45 In-vehicle information 46 Short-time BBI data 47 ANF indicator data 48 ANF indicator integrated DB

Claims (14)

An autonomic nerve function evaluation system,
has a processor and a storage device,
The storage device holds biological data measured during a predetermined length of time and an autonomic nerve function index calculated based on biological data measured before the predetermined length of time,
The processor includes:
The predetermined length of time is determined based on the biological data measured during the predetermined length of time and the autonomic nerve function index calculated based on the biological data measured before the predetermined length of time. Determine the validity of the biometric data measured at that time,
An autonomic nerve function evaluation system, characterized in that an autonomic nerve function index is calculated based on data determined to be valid among biological data measured during the predetermined length of time. The autonomic nerve function evaluation system according to claim 1,
The biological data is heartbeat interval data,
The processor is autonomous, characterized in that the processor determines the validity of the heartbeat interval data measured during the predetermined length of time based on a Lorentz plot of the heartbeat intervals measured during the predetermined length of time. Neurological function evaluation system. The autonomic nerve function evaluation system according to claim 2,
The processor includes:
Determining that the heartbeat interval data within the range of an ellipse area set on the Lorentz plot of the heartbeat intervals measured during the predetermined length of time is valid;
The center of the elliptical area is determined based on the average value or median value of the heartbeat intervals measured during the predetermined length of time and the biological data measured before the predetermined length of time. Calculated based on the average value or median value of the heartbeat interval among the neurological function indicators,
The diameter of the elliptical region is calculated based on the standard deviation on a Lorentz plot of heartbeat intervals measured during the predetermined length of time and biological data measured before the predetermined length of time. An autonomic nervous function evaluation system characterized by calculating based on, among autonomic nervous function indicators, the standard deviation on a Lorentz plot of heartbeat intervals. The autonomic nerve function evaluation system according to claim 3,
The processor includes:
Among the autonomic nerve function indexes calculated based on the average value or median value of the heartbeat intervals measured during the predetermined length of time and the biological data measured before the predetermined length of time, the heartbeat Calculating the weighted average of the average value or median of the intervals as the center of the elliptical area,
Among the autonomic nervous function indexes calculated based on the standard deviation on the Lorentz plot of heartbeat intervals measured during the predetermined length of time and biological data measured before the predetermined length of time, An autonomic nervous function evaluation system characterized in that the diameter of the elliptical region is calculated by multiplying the weighted average of the standard deviation on the Lorentz plot of the heartbeat interval by a predetermined coefficient. The autonomic nerve function evaluation system according to claim 4,
An autonomic nerve function evaluation system characterized in that a greater weight is attached to a value extracted from an autonomic nerve function index calculated based on newer biological data. The autonomic nerve function evaluation system according to claim 2,
The processor includes:
Among the heartbeat interval data measured during the predetermined length of time, if a section of data that is determined to be invalid based on the Lorentz plot is shorter than a predetermined standard, based on the heartbeat interval data before and after the said section. interpolate the heartbeat interval data for the relevant section,
An autonomic nervous function evaluation system characterized in that data determined to be appropriate based on the Lorentz plot and the interpolated data are determined to be good data. The autonomic nerve function evaluation system according to claim 6,
The processor extracts a confidence interval from one or more intervals in which the good data is continuous among the heartbeat interval data measured during the predetermined length of time,
A reliability score for evaluating the reliability of the confidence interval is calculated based on at least one of the length of the interval in which the good data is continuous and the total length of a plurality of intervals in which the good data is continuous. An autonomic nerve function evaluation system characterized by: The autonomic nerve function evaluation system according to claim 7,
The processor extracts the confidence interval from the interval in which the good data is continuous based on at least one of a first extraction method, a second extraction method, and a third extraction method,
The first extraction method is a method of extracting one interval longer than a predetermined first continuous interval length as the confidence interval,
The second extraction method extracts two or more sections that are longer than a predetermined second continuous section length among the continuous sections of good data, and the sum of the lengths is a predetermined third continuous section. A method of extracting two or more intervals longer than the length as the confidence interval,
The third extraction method is a method of extracting two or more intervals whose total length is longer than the third continuous interval length from among the intervals where the good data is continuous, as the confidence interval,
The second continuous section length is shorter than the first continuous section length,
The third continuous section length is longer than the first continuous section length,
When the processor extracts the confidence intervals by two or more extraction methods among the first extraction method, the second extraction method, and the third extraction method, the processor extracts the confidence intervals from among the two or more extracted confidence intervals. The autonomic nervous function evaluation system is characterized in that the autonomic nervous function index is calculated based on the heartbeat interval data in the confidence interval with the highest reliability score. The autonomic nerve function evaluation system according to claim 8,
the processor calculates the reliability score by multiplying a ratio of the length of the confidence interval to the predetermined length of time by a reliability weight according to a method for extracting the confidence interval;
An autonomic nerve function evaluation system, wherein the reliability weight according to the first extraction method is the largest, and the reliability weight according to the third extraction method is the smallest. The autonomic nerve function evaluation system according to claim 2,
The storage device holds information regarding the sensor that measured the biological data,
The autonomic nerve function is characterized in that the processor performs a shaping process according to the sensor on the biometric data measured during the predetermined length of time, and determines the validity of the shaped biometric data. Rating system. The autonomic nerve function evaluation system according to claim 2,
The processor includes:
The calculated autonomic nerve function index is used to indicate the identification information of the person who is the measurement target of the heartbeat interval data, the age of the person, the gender of the person, the identification information of the vehicle in which the person rides, and the work content of the person. information, information regarding the sensor used to obtain the heartbeat interval data, and at least one of the measurement time, and storing it in the storage device as autonomic nerve function index integrated data;
The autonomic nerve function index extracted from the autonomic nerve function index integrated data based on specified extraction conditions is used to determine the validity of the biological data measured during the predetermined length of time. Autonomic nerve function evaluation system. The autonomic nerve function evaluation system according to claim 2,
The autonomic nervous function evaluation system is characterized in that the heartbeat interval data includes R wave interval data based on electrocardiogram measurements. The autonomic nerve function evaluation system according to claim 2,
The autonomic nervous function evaluation system is characterized in that the heartbeat interval data includes data on peak intervals of measured values of pulse waves. An autonomic nerve function evaluation method performed by an autonomic nerve function evaluation system,
The autonomic nerve function evaluation system includes a processor, a storage device,
The storage device holds biological data measured during a predetermined length of time and an autonomic nerve function index calculated based on biological data measured before the predetermined length of time,
The autonomic nerve function evaluation method includes:
The processor is based on the biological data measured during the predetermined length of time and the autonomic nerve function index calculated based on the biological data measured before the predetermined length of time, a step of determining the validity of the biometric data measured during the predetermined length of time;
The autonomous system includes a step in which the processor calculates an autonomic nervous function index based on data determined to be valid among biological data measured during the predetermined length of time. Neurological function evaluation method.

Images