JP7579600B2 - Autonomic nerve evaluation device, autonomic nerve evaluation method, and program - Google Patents

Description

The present invention relates to an autonomic nerve evaluation device, an autonomic nerve evaluation method, and a program.

Functional analysis of the autonomic nervous system has been attracting attention as an index for objectively evaluating fatigue. The autonomic nervous system consists of the sympathetic nervous system, which mainly works when we are awake or tense, and the parasympathetic nervous system, which mainly works when we are asleep or relaxed, and it is known that the two work in balance to regulate physiological functions. For example, Patent Document 1 discloses a device for determining the autonomic nervous state of a vehicle occupant.

JP 2022-23702 A

When assessing a subject's autonomic nervous function, it is common to measure the subject's heart rate for 120 seconds and then perform frequency analysis of the heart rate intervals obtained. However, depending on how the evaluation of autonomic nervous function is used, such as when the evaluation is performed as a pre-employment test, it has been pointed out that the measurement takes too long, which can be a problem.

Therefore, the present invention aims to provide a technology that enables evaluation of autonomic nerve function more quickly.

An autonomic nerve evaluation device according to one aspect of the present invention includes an acquisition unit that acquires heart rate data of a subject to be evaluated, and a calculation unit that calculates total power, which is an index showing the overall function of the autonomic nerves of the subject, from the heart rate data based on the Lorenz plot method.

The present invention provides a technology that enables the evaluation of autonomic nerve function to be performed more quickly.

FIG. 2 is a diagram illustrating an example of a hardware configuration of an evaluation device according to the present embodiment. FIG. 1 shows the relationship between ccvTP and age. FIG. 13 is a diagram showing the results of investigating the relationship between age and LF value and the relationship between age and HF value. FIG. 2 is a diagram illustrating an example of a functional block configuration of an evaluation device. 13 is a flowchart showing an example of a processing procedure by which the evaluation device quickly evaluates autonomic nerve function. FIG. 13 is a diagram for explaining a method for calculating a subject's total power by the Lorenz plot method. 1 is a diagram showing the results of multiple regression analysis of LP.S, LP.m, age, and TP. FIG. 1 shows the correlation between LP.S, LP.m, age and TP. FIG. 1 shows the correlation between TP and LP.S. FIG. 1 shows the correlation between TP and LP.S. 1 is a diagram showing the results of multiple regression analysis of LP.S, LP.m, age, and TP. FIG. 1 shows the correlation between LP.S, LP.m, age and TP. FIG. 1 shows the correlation between TP and LP.S. FIG. 1 shows the correlation between TP and LP.S. FIG. 1 shows the correlation of LP.S. 13 is a flowchart showing an example of a processing procedure for the evaluation device to evaluate the parasympathetic nerves. 13 is a flowchart showing an example of a processing procedure for the evaluation device to evaluate the sympathetic nerves. FIG. 1 shows the correlation between HFs. FIG. 1 shows the correlation between HFs. 1 shows the results of multiple regression analysis of LP.S, LP.m, age, and HF. FIG. 1 shows the correlation between LP.S, LP.m, age and HF. FIG. 1 shows the correlation between HF and LP.S. FIG. 1 shows the correlation between HF and LP.S. FIG. 13 is a diagram showing the results of a comparison between the proportion of HF in TP and the fatigue level of subjects. FIG. 13 is a diagram showing an example of a screen for outputting the measurement results of a subject.

The embodiment of the present invention will be described with reference to the attached drawings. In each drawing, the same reference numerals are used to denote the same or similar configurations.

<System Configuration>
1 is a diagram showing an example of a hardware configuration of an evaluation device 10 according to this embodiment. The evaluation device 10 (autonomic nerve evaluation device) includes a processor 11 such as a central processing unit (CPU) or a graphical processing unit (GPU), a storage device 12 such as a memory, a hard disk drive (HDD) and/or a solid state drive (SSD), an input device 13, an output device 14, and a biological information acquisition device 15. The evaluation device 10 may be dedicated hardware or general-purpose hardware such as a personal computer.

The input device 13 is a device for inputting data about the subject, and is composed of, for example, a keyboard, operation buttons, or an input interface on a touch panel. In this embodiment, at least the subject's age is input to the input device 13 as data about the subject. In addition to the subject's age, the subject's name, address, sex, etc. may also be input.

The output device 14 is a device for outputting information, such as a display, an LED, a touch panel, a printer, and/or a speaker.

The bioinformation acquisition device 15 is a device for collecting bioinformation data of the subject. The bioinformation data of the subject is either heart rate data or pulse data. The bioinformation acquisition device 15 may be an electrocardiogram or a pulse meter. The evaluation device 10 may also accept, via the input device 13, input of heart rate data or pulse data measured by an external electrocardiogram or pulse meter. In this case, the evaluation device 10 may not be equipped with the bioinformation acquisition device 15. In the following description, the bioinformation data is described as heart rate data.

The LF (Low Frequency) value, which is an index showing the activity of the sympathetic nerves, and the HF (High Frequency) value, which is an index showing the activity of the parasympathetic nerves, can be obtained by frequency analysis of bioinformation data. The frequency analysis (time-frequency analysis) method can be a known analysis method, such as the maximum entropy model (MEM) method, the fast Fourier transform (FFT) method, and the wavelet method. Among these, the maximum entropy method is preferably used. The maximum entropy method allows for analysis with high time resolution. For example, the R-R interval of the heartbeat data can be separated into low frequency components (LF: 0.04-0.15 Hz) and high frequency components (HF: 0.15-0.40 Hz) in the frequency domain using the maximum entropy method, and the sum of the power of the low frequency components and the high frequency components can be calculated as the LF value and the HF value, respectively.

ここで、ＬＦ（ｔ）はＬＦ値、ＨＦ（ｔ）はＨＦ値、Ｐ（ｆ）はパワースペクトル関数、Ｃ（ｔ）は心拍（心電図）のＲ－Ｒ間隔の自己相関関数、ｔは時間、ｆは周波数を示す。数式３に示すとおり、数式上Ｐ（ｆ）は全時間領域で積分するものとしているが、実際は、観測領域で積分すれば足りる。 Specifically, the LF value and the HF value can be calculated, for example, by the following Equations 1, 2, and 3.

Here, LF(t) is the LF value, HF(t) is the HF value, P(f) is the power spectrum function, C(t) is the autocorrelation function of the R-R interval of the heartbeat (electrocardiogram), t is time, and f is frequency. As shown in Equation 3, P(f) is assumed to be integrated over the entire time domain, but in reality, it is sufficient to integrate it over the observation domain.

The sum of the LF and HF values (LF + HF) is an index showing the overall function of the autonomic nervous system, and is called TP (Total Power). TP is also called "autonomic nervous activity" or "total autonomic nervous activity."

ここで、ＲＲは、被験者における心拍のＲ－Ｒ間隔（秒）を示す。なお、Ｒ－Ｒ間隔（秒）は、心電図におけるＱＲＳ波から次のＱＲＳ波までの間隔を意味する。Ｒ－Ｒ間隔は、６０÷心拍数（回／分）でも算出可能である。 The ccvTP (Coefficient of Component Variance TP) is a value calculated by correcting the TP with the heart rate (more specifically, the average of the R-R interval) during the time used to measure the LF value and the HF value. The ccvTP, like the TP, is also called the "autonomic nerve activity" or the "total autonomic nerve activity." Specifically, the ccvTP can be calculated by the following formula 4.

Here, RR indicates the R-R interval (seconds) of the subject's heartbeat. The R-R interval (seconds) refers to the interval from one QRS wave to the next QRS wave in an electrocardiogram. The R-R interval can also be calculated by dividing 60 by the heart rate (beats/min).

It is known from experiments that when the heart rate is high, the TP value is low. Therefore, by using a ccvTP corrected according to the heart rate, the amount of autonomic nerve activity can be appropriately expressed regardless of the subject's heart rate. The inventors have also demonstrated that the {LF+HF} value decreases with age, and the same is true for ccvTP.

As described later, the evaluation device 10 calculates ccvHF by correcting the HF value with the heart rate (more specifically, the average of the R-R intervals) within the time used to measure the HF value. The HF value and ccvHF may be called an index related to the parasympathetic nerves, an index indicating the function of the parasympathetic nerves, a parasympathetic nerve activity amount, a parasympathetic nerve activity index, and a healing level, etc. The HF value and ccvHF may also be called a first index related to the parasympathetic nerves and a second index related to the parasympathetic nerves, etc., respectively. The evaluation device 10 calculates ccvLF by correcting the LF value with the heart rate (more specifically, the average of the R-R intervals) within the time used to measure the LF value. The LF value and ccvLF may also be called an index related to the sympathetic nerves, an index indicating the function of the sympathetic nerves, a sympathetic nerve activity index, and a sympathetic nerve activity amount, etc. Furthermore, the LF value and ccvLF may be referred to as the first sympathetic nerve index and the second sympathetic nerve index, respectively.

Figure 2 shows the relationship between ccvTP and age. Curve G1 indicates the boundary line between the top 25% of subjects with high ccvTP among subjects of the same age. Curve G2 indicates the median ccvTP among subjects of the same age. Curve G3 indicates the boundary line between the bottom 75% of subjects with low ccvTP among subjects of the same age. Figure 2 shows that ccvTP decreases with age.

(Rapid assessment of autonomic nervous function)
Currently, when evaluating a subject's autonomic nervous function, in order to obtain reliable evaluation results, it is common to perform frequency analysis using the maximum entropy method on the heart rate data obtained by continuously measuring the subject for 120 seconds.

It is also known that when TP is analyzed using heart rate data obtained by continuously measuring a subject for 60 seconds, the correlation coefficient with TP obtained from 120 seconds of heart rate data is approximately 0.7. Therefore, when the same subject undergoes repeated measurements, the state of the subject's autonomic nervous system is grasped by analyzing TP using 60 seconds of heart rate data from the second measurement onwards.

In the maximum entropy method, heart rate data measured for at least 25 seconds is required to calculate LF. This is because LF values are low-frequency components between 0.04 and 0.15 Hz, so measurement for at least 25 seconds is necessary to obtain data for one wavelength. Furthermore, measurement accuracy is insufficient with data from only one wavelength, so as mentioned above, it is considered preferable to measure for at least 120 seconds.

Here, when evaluating autonomic nervous function in a hospital or testing center, there is no particular problem with measuring the subject for 120 seconds. However, evaluation of autonomic nervous function is used in a variety of places, not just hospitals and testing centers, such as daily pre-employment tests for taxi drivers and truck drivers. Repeated daily measurements of heart rate data while remaining still for 120 or 60 seconds can be stressful for subjects, so a method that allows measurement in a shorter time is desirable.

The inventors considered a method for measuring autonomic nervous function in a shorter time and came to the conclusion that by using the Lorenz plot method instead of the maximum entropy method, TP can be calculated with high accuracy with only about 7 seconds of heart rate data. Therefore, the evaluation device 10 according to this embodiment calculates TP using the Lorenz plot method instead of the maximum entropy method, making it possible to evaluate the subject's autonomic nervous function in a shorter time than conventional methods.

(Evaluation of parasympathetic and sympathetic nervous systems)
TP, which is an index showing the activity of the entire autonomic nerves, changes with age. Therefore, in order to evaluate subjects of various ages in the same way, a common index that is not affected by the age of the subject is necessary. Therefore, the inventors decided to use the deviation value of TP as a common index regarding the activity of the entire autonomic nerves. Here, as described above, autonomic nerve activity is evaluated from two components, LF (0.04-0.15 Hz component) which mainly represents the activity of the sympathetic nerves, and HF (0.15-0.40 Hz component) which mainly represents the activity of the parasympathetic nerves. However, it has not been clear until now how each component changes with age, and there have been no reports of evaluating subjects by converting HF into a deviation value.

Figure 3 shows the results of the inventors' investigation into the relationship between age and LF values and the relationship between age and HF values for 264 people aged 18 to 73. As shown in Figure 3, logLF shows a significant negative correlation with age (r=-0.524, p<0.001), and logHF shows a significant negative correlation with age (r=-0.423, p<0.001). In other words, it has been confirmed that health assessment cannot be performed using LF and HF values directly when subjects are of different ages.

In this embodiment, the changes in HF and LF with age are clarified, and HF and LF are converted into standard deviations. By using the standard deviations of HF and LF, which are objective health indicators that exclude the age factor, it is possible to commonly evaluate the function of the parasympathetic nervous system or the sympathetic nervous system across subjects.

Here, when calculating the HF value using the maximum entropy method, since the HF value is a high-frequency component of 0.15 to 0.40 Hz, one wavelength of data can be obtained with about 6.6 seconds of heart rate data. In other words, the HF value can provide highly reliable measurement results in a shorter measurement time than the LF value, which requires 25 seconds to measure one wavelength. Taking this into consideration, the inventors verified the correlation between the HF value obtained from 90 seconds of heart rate data and the HF value obtained from about 10 seconds of heart rate data, and found that these HF values are highly correlated. Therefore, the evaluation device 10 according to this embodiment calculates the HF in a shorter measurement time than conventional methods, making it possible to evaluate the parasympathetic nervous activity of the subject in a shorter time.

The evaluation device 10 according to this embodiment may have only one of the functions for realizing "rapid evaluation of autonomic nerve function" and the function for realizing "evaluation of parasympathetic and sympathetic nerves", or may have both.

<Function block configuration>
FIG. 4 is a diagram showing an example of a functional block configuration of the evaluation device 10. The evaluation device 10 includes a storage unit 100, an acquisition unit 101, a calculation unit 102, and an output unit 103. The storage unit 100 can be realized using a storage device 12 provided in the evaluation device 10. The acquisition unit 101, the calculation unit 102, and the output unit 103 can be realized by the processor 11 of the evaluation device 10 executing a program stored in the storage device 12. The program can be stored in a storage medium. The storage medium storing the program may be a non-transitory computer readable storage medium. The non-transitory storage medium is not particularly limited, and may be, for example, a storage medium such as a USB memory or a CD-ROM.

The memory unit 100 stores distribution data (ccvTP) 100a showing the distribution of ccvTP (autonomic nerve activity) by age of the subject, distribution data (ccvHF) 100b showing the distribution of ccvHF values, which are an index of parasympathetic nerve activity, by age, and distribution data (ccvLF) 100c showing the distribution of ccvLF, which is an index of parasympathetic nerve activity, by age.

The distribution data (ccvTP) 100a is obtained by collecting ccvTP from many subjects of different ages and statistically analyzing it. For example, in this embodiment, the average value of ccvTP for each age of the subject and the standard deviation (σ) of ccvTP for each age of the subject may be calculated from the ccvTP of many subjects of different ages collected in advance, and these may be compiled into a database and stored in the memory unit 100 as distribution data (ccvTP) 100a.

The distribution data (ccvHF) 100b is obtained by collecting ccvHF from many subjects of different ages and statistically analyzing it. For example, in this embodiment, the average value of ccvHF for each age of the subject and the standard deviation (σ) of ccvHF for each age of the subject may be calculated from the ccvHF of many subjects of different ages collected in advance, and these may be compiled into a database and stored in the storage unit 100 as the distribution data (ccvHF) 100b.

The distribution data (ccvLF) 100c is obtained by collecting ccvLF from many subjects of different ages and statistically analyzing it. For example, in this embodiment, the average value of ccvLF for each age of the subject and the standard deviation (σ) of ccvLF for each age of the subject may be calculated from the ccvLF of many subjects of different ages collected in advance, and these may be compiled into a database and stored in the storage unit 100 as the distribution data (ccvLF) 100c.

The acquisition unit 101 acquires the heart rate data and/or the age of the subject to be evaluated.

The calculation unit 102 uses the Lorenz plot method to calculate the standard deviation of the subject's ccvTP (autonomic nerve activity) from the subject's heart rate data. More specifically, the calculation unit 102 calculates TP, which is an index showing the overall function of the subject's autonomic nerves, from the subject's heart rate data based on the Lorenz plot method. The calculation unit 102 also calculates the subject's ccvTP (autonomic nerve activity) using the subject's TP and the subject's heart rate data, and calculates the standard deviation of the subject's ccvTP using the distribution data (ccvTP) 100a and the subject's ccvTP.

The calculation unit 102 also calculates the subject's HF value based on the subject's heart rate data. The calculation unit 102 also calculates the subject's ccvHF using the subject's HF value and the subject's heart rate data, and calculates the standard deviation of the subject's ccvHF value using the calculated ccvHF value of the subject, the subject's age, and distribution data (ccvHF) 100b.

The calculation unit 102 may also calculate the subject's HF value by frequency analyzing the subject's heart rate data, and calculate the subject's ccvHF and a deviation value of the subject's ccvHF from the calculated subject's HF value. More specifically, the calculation unit 102 may calculate the subject's HF value by frequency analyzing the subject's heart rate data, calculate the subject's ccvHF using the calculated HF value and the subject's heart rate data, and calculate the deviation value of the subject's ccvHF value using the calculated subject's ccvHF value, the subject's age, and distribution data (ccvHF) 100b.

The calculation unit 102 may also use the Lorenz plot method to calculate the subject's HF value from the subject's heart rate data, and calculate the subject's ccvHF and the deviation value of the subject's ccvHF from the calculated subject's HF value. More specifically, the calculation unit 102 may use the Lorenz plot method to calculate the subject's HF value from the subject's heart rate data, calculate the subject's ccvHF using the calculated HF value and the subject's heart rate data, and calculate the deviation value of the subject's ccvHF value using the calculated subject's ccvHF value, the subject's age, and distribution data (ccvHF) 100b.

The calculation unit 102 also calculates the subject's LF value based on the subject's heart rate data. The calculation unit 102 also calculates the subject's LF value by performing frequency analysis on the subject's heart rate data. The calculation unit 102 also calculates the subject's ccvLF using the subject's LF value and the subject's heart rate data, and calculates the standard deviation of the subject's ccvLF value using the calculated ccvLF value of the subject, the subject's age, and distribution data (ccvLF) 100c.

The calculation unit 102 may also use the Lorenz plot method to calculate the subject's TP (autonomic nerve activity) from the subject's heart rate data, and calculate the ratio of the subject's HF value to the subject's TP (i.e., HF/TP). Alternatively, the calculation unit 102 may use the Lorenz plot method to calculate the subject's ccvTP (autonomic nerve activity) from the subject's heart rate data, and calculate the ratio of the subject's ccvHF to the subject's ccvTP (i.e., ccvHF/ccvTP).

The output unit 103 outputs various data related to the subject's autonomic nerves calculated by the calculation unit 102 to a screen or paper, etc. For example, the output unit 103 outputs the standard deviation of the subject's TP and/or ccvTP. The output unit 103 also outputs the standard deviation of the subject's ccvHF. The output unit 103 may also output the standard deviation of the subject's ccvHF and the standard deviation of the subject's ccvTP (autonomic nerve activity) side by side. The output unit 103 also outputs the standard deviation of the subject's ccvLF. The output unit 103 may also output the standard deviation of the subject's ccvLF and the standard deviation of the subject's ccvTP side by side. The output unit 103 may also output the standard deviation of the subject's ccvLF, the standard deviation of the subject's ccvHF, and the standard deviation of the subject's ccvTP side by side. Furthermore, the output unit 103 may output the ratio of the subject's HF value to the subject's TP calculated by the calculation unit 102. Alternatively, the output unit 103 may output the ratio of the subject's ccvHF to the subject's ccvTP calculated by the calculation unit 102.

<Processing Procedure>
(Rapid assessment of autonomic nervous function)
FIG. 5 is a flowchart showing an example of a processing procedure for the evaluation device 10 to quickly evaluate an autonomic nerve function.

In step S11, the acquisition unit 101 acquires the subject's age and heart rate data.

In step S12, the calculation unit 102 calculates the subject's TP using the Lorenz plot method.

Figure 6 is a diagram for explaining a method for calculating the subject's TP using the Lorenz plot method. The calculation unit 102 calculates the subject's TP according to the following steps 1 to 4.

"Step 1": The calculation unit 102 plots the subject's heart rate data on a graph with the nth R-R interval (RRI (RR Interval)) of the heart rate data (hereinafter referred to as (RRI _n )) on the horizontal axis and the n+1th interval (hereinafter referred to as (RRI _n+1 )) consecutive to the nth on the vertical axis. That is, the calculation unit 102 plots on the graph a point (x, y) = (RRI _n , RRI _n+1 ) consisting of two consecutive R-R intervals among the R-R intervals of the heart rate data. For example, in 10 seconds of heart rate data, if the first R-R interval is 1000 ms, the second R-R interval is 900 ms, the third R-R interval is 950 ms, and the fourth R-R interval is 1050 ms, the calculation unit 102 plots, for example, points (1000, 900), (900, 950), and (950, 1050) on the graph.

"Step 2": The calculation unit 102 projects all points plotted on the graph onto the y = x axis and y = -x axis. For example, when point (1000, 900) is projected onto the y = x axis, the point is projected to a position (any position on line A) where a perpendicular line is drawn from coordinates (1000, 900) to the y = x axis. Similarly, when point (1000, 900) is projected onto the y = -x axis, the point is projected to a position (any position on line B) where a perpendicular line is drawn from coordinates (1000, 900) to the y = -x axis.

"Step 3": The calculation unit 102 calculates the average distance (LP.m) from the origin (0,0) to each point projected onto the y = x axis. Next, the calculation unit 102 calculates the standard deviation σx of the distance from the origin (0,0) to each point projected onto the y = x axis, and the standard deviation σ-x of the distance from the origin (0,0) to each point projected onto the y = -x axis.

「手順５」：算出部１０２は、以下の数式６を用いて、被験者のＴＰを推定（算出）する。

なお、数式６において、４つの係数（ａ、ｂ、ｃ、ｄ）は、心拍データの測定時間ごとに、ＴＰ９０を従属変数とし、ＬＰ．Ｓ、ＬＰ．ｍ及び年齢を独立変数として重回帰分析を行うことで算出された係数であってもよい。すなわち、記憶部１００に、心拍データの測定時間ごとに４つの係数（ａ、ｂ、ｃ、ｄ）のセットを予め記憶しておき、算出部１０２は、心拍データの測定時間に対応する係数のセットを記憶部１００から取得し、取得した係数を数式６に適用することで、被験者のＴＰを算出（推定）するようにしてもよい。例えば、心拍データの測定時間が連続１０秒である場合、算出部１０２は、係数としてａ＝０．０２９６８、ｂ＝０．６９９６５、ｃ＝－１２．９６６及びｄ＝１１０．８２６を利用するようにしてもよい。なお、当該４つの係数は、心拍データの測定時間に関わらず固定値であってもよい。この場合、４つの係数は、ａ＝０．０２９６８、ｂ＝０．６９９６５、ｃ＝－１２．９６６及びｄ＝１１０．８２６であってもよい。 "Step 4": The calculation unit 102 calculates the area (LP.S) of an ellipse M whose major axis is a predetermined coefficient (D) x σx and whose minor axis is a predetermined coefficient (D) x σ-x, using the following formula 5. The smaller the area (LP.S) of the ellipse M, the smaller the fluctuation of the R-R interval. Note that the predetermined coefficient (D) may be any value as long as it is a positive integer. For example, it may be 2 or 3. In the example of FIG. 6, an ellipse is illustrated when the predetermined coefficient (D) = 2.

“Step 5”: The calculation unit 102 estimates (calculates) the TP of the subject using the following Equation 6.

In addition, in the formula 6, the four coefficients (a, b, c, d) may be coefficients calculated by performing a multiple regression analysis with TP90 as a dependent variable and LP.S, LP.m, and age as independent variables for each measurement time of the heartbeat data. That is, a set of four coefficients (a, b, c, d) may be stored in advance in the storage unit 100 for each measurement time of the heartbeat data, and the calculation unit 102 may obtain a set of coefficients corresponding to the measurement time of the heartbeat data from the storage unit 100 and apply the obtained coefficients to the formula 6 to calculate (estimate) the TP of the subject. For example, when the measurement time of the heartbeat data is 10 consecutive seconds, the calculation unit 102 may use a=0.02968, b=0.69965, c=-12.966, and d=110.826 as coefficients. In addition, the four coefficients may be fixed values regardless of the measurement time of the heartbeat data. In this case the four coefficients may be a=0.02968, b=0.69965, c=-12.966 and d=110.826.

Regarding the above-described steps 1 to 5, the calculation unit 102 may consider data outside the ellipse M in FIG. 6 to be noise, remove the points that are noise, and then calculate the area (LP.S) of the ellipse M again. Specifically, the calculation unit 102 may calculate the area (LP.S) of the ellipse M by executing steps 1 to 4, remove the points that are outside the ellipse M among the points plotted on the graph, and calculate the area (LP.S) of the ellipse M after noise removal by executing steps 1 to 4 again. The calculation unit 102 may also calculate the total power of the subject by using the area (LP.S) of the ellipse M after noise removal by step 5. In other words, the calculation unit 102 may perform the following processing.
1. Two consecutive R-R intervals of the heart rate data are plotted as points x and y on a graph with one axis as the x-axis and the other as the y-axis.
2. A first ellipse is calculated with the major axis being the first standard deviation obtained by multiplying the standard deviation of the distance between the origin (0,0) and the projection points obtained by projecting the points plotted on the graph onto the x = y axis by a specified coefficient (D), and the minor axis being the second standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points obtained by projecting the points plotted on the graph onto the x = -y axis by a specified coefficient (D).
3. For points that are plotted on the graph and exist inside ellipse M, a second ellipse is calculated whose major axis is the third standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points of the plotted points on the graph onto the x = y axis by a predetermined coefficient (D), and whose minor axis is the fourth standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points of the plotted points on the graph onto the x = -y axis by a predetermined coefficient (D).
4. The area of the second ellipse is substituted into a predetermined formula (Formula 6) to estimate (calculate) the TP of the subject.

5, in step S13, the calculation unit 102 calculates the ccvTP of the subject using the TP of the subject calculated in step S12 and the subject's heart rate data. The calculation unit 102 also calculates the standard deviation of the ccvTP of the subject using the ccvTP of the subject and distribution data (ccvTP) 100a. For example, the calculation unit 102 can calculate the standard deviation of the ccvTP of the subject by the following procedure.
1. Calculate the subject's ccvTP by correcting the TP calculated in step S12 with the subject's heart rate according to Equation 4.
2. From the distribution data (ccvTP) 100a, the mean value and standard deviation (σ) of ccvTP corresponding to the subject's age are obtained.
3. Calculate the deviation value of the subject's ccvTP using the formula: "Deviation value of ccvTP = 10 x (subject's ccvTP - mean value of ccvTP corresponding to subject's age) / standard deviation of ccvTP corresponding to subject's age (σ) + 50."

(Variation 1 of rapid assessment of autonomic function)
In the above-described method for evaluating autonomic nerve function, logarithmic transformed LP.S (hereinafter referred to as "logLP.S") and LP.m (hereinafter referred to as "logLP.m") may be used to estimate logarithmic transformed TP (hereinafter referred to as "logTP").

数式７において、４つの係数（ｅ、ｆ、ｇ、ｈ）は、心拍データの測定時間ごとに、ｌｏｇＴＰ９０を従属変数とし、ｌｏｇＬＰ．Ｓ、ｌｏｇＬＰ．ｍ及び年齢を独立変数として重回帰分析を行うことで算出された係数であってもよい。すなわち、記憶部１００に、心拍データの測定時間ごとに４つの係数（ｅ、ｆ、ｇ、ｈ）のセットを予め記憶しておき、算出部１０２は、心拍データの測定時間に対応する係数のセットを記憶部１００から取得し、取得した係数を数式７に適用することで、被験者のｌｏｇＴＰを算出（推定）するようにしてもよい。例えば、心拍データの測定時間が連続１０秒である場合、算出部１０２は、係数としてｅ＝０．５１３３３、ｆ＝１．４２４４６、ｇ＝－０．００８１及びｈ＝－３．３０１６を利用するようにしてもよい。なお、当該４つの係数は、心拍データの測定時間に関わらず固定値であってもよい。この場合、４つの係数は、ｅ＝０．５１３３３、ｆ＝１．４２４４６、ｇ＝－０．００８１及びｈ＝－３．３０１６であってもよい。 Specifically, in "Step 5" of the above-described processing procedure, logTP may be estimated by using the following Equation 7 instead of Equation 6. Note that the logarithm may be a common logarithm.

In the formula 7, the four coefficients (e, f, g, h) may be coefficients calculated by performing a multiple regression analysis with logTP90 as a dependent variable and logLP.S, logLP.m, and age as independent variables for each measurement time of the heartbeat data. That is, a set of four coefficients (e, f, g, h) for each measurement time of the heartbeat data may be stored in advance in the storage unit 100, and the calculation unit 102 may obtain a set of coefficients corresponding to the measurement time of the heartbeat data from the storage unit 100 and apply the obtained coefficients to the formula 7 to calculate (estimate) the logTP of the subject. For example, when the measurement time of the heartbeat data is 10 consecutive seconds, the calculation unit 102 may use e=0.51333, f=1.42446, g=-0.0081, and h=-3.3016 as coefficients. The four coefficients may be fixed values regardless of the measurement time of the heartbeat data. In this case the four coefficients may be e=0.51333, f=1.42446, g=-0.0081 and h=-3.3016.

The calculation unit 102 may also calculate an estimated TP (i.e., TP before logarithmic transformation) from the estimated logTP, and execute the processing procedure of step S13 using the calculated estimated TP.

The noise removal process described in "(Rapid evaluation of autonomic nerve function)" may also be applied to this modification 1. That is, the calculation unit 102 may perform the following process.
1. Two consecutive intervals of heart rate data, designated x and y, are plotted on a graph with one axis as the x-axis and the other as the y-axis.
2. A first ellipse is calculated with the major axis being the first standard deviation obtained by multiplying the standard deviation of the distance between the origin (0,0) and the projection points obtained by projecting the points plotted on the graph onto the x = y axis by a specified coefficient (D), and the minor axis being the second standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points obtained by projecting the points plotted on the graph onto the x = -y axis by a specified coefficient (D).
3. For points that are plotted on the graph and exist inside ellipse M, a second ellipse is calculated whose major axis is the third standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points of the plotted points on the graph onto the x = y axis by a predetermined coefficient (D), and whose minor axis is the fourth standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points of the plotted points on the graph onto the x = -y axis by a predetermined coefficient (D).
4. The area of the second ellipse is substituted into a predetermined calculation formula (Formula 7) to estimate (calculate) the logTF of the subject.

(Variation 2 of rapid assessment of autonomic function)
In the above-described "Step 1", the calculation unit 102 plots on a graph a point (x, y) = (RRI _n , RRI _n+1 ) consisting of two consecutive R-R intervals among the R-R intervals of the heartbeat data, but a point consisting of two non-consecutive R-R intervals may be plotted on a graph. In other words, the calculation unit 102 may plot a plurality of points, where x and y are two non-consecutive R-R intervals among the R-R intervals of the heartbeat data, on a graph with one axis being the x-axis and the other axis being the y-axis.

Specifically, in "Step 1", the calculation unit 102 may plot, from the heart rate data of the subject, a graph with the nth R-R interval of the heart rate data on the horizontal axis and the n+αth R-R interval (hereinafter, referred to as (RRI _n+ α), where α is an integer of 2 or more) on the vertical axis. That is, the calculation unit 102 may plot, on the graph, a point (x, y) = (RRI _n , RRI _n+ α) consisting of two non-consecutive R-R intervals among the R-R intervals of the heart rate data. The value of α may be an integer of 2 or more, and may be, for example, α = 4, 5, or 6 or more.

Here, the nth R-R interval is expressed as RRI(n). When α = 2, the calculation unit 102 may, for example, plot points (RRI(1), RRI(3)), (RRI(2), RRI(4)), and (RRI(3), RRI(5)) on the graph. When α = 3, the calculation unit 102 may, for example, plot points (RRI(1), RRI(4)), (RRI(2), RRI(5)), and (RRI(3), RRI(6)) on the graph.

The value of α may be varied by any value equal to or greater than 1. The any value may be, for example, a value determined to skip heart rate data for periods in which measurement was not performed correctly. In other words, the calculation unit 102 may plot a plurality of points, each of which represents two non-consecutive R-R intervals as x and y, and in which the number of R-R intervals present between the two non-consecutive R-R intervals is different, on a graph with one axis being the x-axis and the other axis being the y-axis. The value of α may be varied randomly.

Here, when the value of α is arbitrarily changed, the calculation unit 102 may plot, for example, points (RRI(1), RRI(2)), (RRI(2), RRI(4)), (RRI(3), RRI(6)), (RRI(4), RRI(5)), (RRI(5), RRI(7)), and (RRI(6), RRI(7)) on the graph.

The value of n does not necessarily have to be increased by 1, but may be increased by any value equal to or greater than 1. The any value may be, for example, a value determined to skip heart rate data for periods in which measurement was not performed correctly. The value of n may also be increased by a random value. For example, the calculation unit 102 may plot points (RRI(1), RRI(2)), (RRI(2), RRI(4)), (RRI(4), RRI(8)), (RRI(5), RRI(6)), (RRI(8), RRI(9)), and (RRI(10), RRI(13)) on a graph.

Note that the method of estimating logarithmically transformed TP using logarithmically transformed LP.S described in "Variation 1 of rapid evaluation of autonomic nervous function" may be applied to this variation 2. Also, the noise removal process described in "(Rapid evaluation of autonomic nervous function)" may be applied to this variation 2. In this case, the process of "1. Plot points in which two consecutive intervals of the heartbeat data are x and y, respectively, on a graph with one axis being the x-axis and the other being the y-axis" in the noise removal process can be replaced with "1. Plot points in which two non-consecutive intervals of the heartbeat data are x and y, respectively, on a graph with one axis being the x-axis and the other being the y-axis."

Conventionally, when heart rate data is acquired at a hospital or testing center, it is possible to acquire relatively stable data. On the other hand, when heart rate data is acquired outside a hospital or testing center, such as daily pre-employment testing for taxi drivers or truck drivers, it is not always possible to acquire stable data, and it is expected that heart rate data containing a lot of noise will have to be used. For example, it is expected that the heart rate data may contain noise, making it impossible to use the heart rate data corresponding to the noise. According to this modification 2, it is possible to evaluate autonomic nervous function even in a situation where it is difficult to use continuous heart rate data. This also makes it possible to measure autonomic nervous function using a simpler heart rate interval measuring device in various environments (during exercise, working, studying, etc.) that are expected in the future.

(TP Experimental Results Part 1)
In the explanation of "TP Experimental Results 1" and "TP Experimental Results 2" described later, "TPxx" means TP calculated by the maximum entropy method using heart rate data measured continuously for xx seconds. For example, "TP90" means TP calculated by the maximum entropy method using heart rate data measured continuously for 90 seconds. Also, "LP.Sxx" means LP.S (area of an ellipse) calculated by the above-mentioned Lorentz plot method using heart rate data measured continuously for xx seconds. For example, "LP.S10" means LP.S calculated by the Lorentz plot method using heart rate data measured continuously for 10 seconds. Also, "LP.mxx" means LP.m (average distance from the origin) calculated by the above-mentioned Lorentz plot method using heart rate data measured continuously for xx seconds. For example, "LP.m10" means LP.m calculated using the Lorenz plot method using heart rate data measured continuously for 10 seconds.

Figure 7 shows the results of multiple regression analysis of LP.S, LP.m, age, and TP. More specifically, Figure 7 shows the results of multiple regression analysis of 500 subjects with TP90 as the dependent variable and LP.S, LP.m, and age as independent variables. The number of seconds in Figure 7 indicates the measurement time of the heartbeat data used in the multiple regression analysis. For example, the multiple correlation coefficient in the column where the measurement time of the heartbeat data is 15 seconds is 0.81955, and the degree of freedom adjusted coefficient of determination is 0.66967. This shows that the result of multiple regression analysis of the heartbeat data obtained by measuring 500 subjects for 15 seconds for TP90, LP.S15, LP.m15, and the age of the subjects was 0.81955, and the degree of freedom adjusted coefficient of determination was 0.66967.

According to FIG. 7, the multiple correlation coefficient obtained from heart rate data with a measurement time of 3 to 11 seconds is 0.49 or more, and furthermore, when the measurement time is 12 seconds or more, the multiple correlation coefficient is 0.8 or more. Furthermore, the coefficient of determination adjusted for the degrees of freedom exceeds 0.5 when the measurement time is 8 seconds or more. Therefore, when the measurement time is 8 seconds or more, it can be said that Equation 6 is at a level that can be used as an estimation model for TP. Furthermore, when the measurement time is 4 seconds or more, the multiple correlation coefficient exceeds 0.6, so it can be said that even heart rate data with a measurement time of 4 seconds or more and less than 8 seconds can be used as a simple measurement of the autonomic nerves.

Figure 8 shows the correlation between LP.S, LP.m, age, and TP. Figures 9 and 10 show the correlation between TP and LP.S. Figures 8 to 10 show the results of measurements taken on 500 subjects. In Figures 8 to 10, the numbers in the upper row are the correlation coefficients (r values), and the numbers in the lower row are the significance probabilities (p values). Figure 8 shows that LP.S10 and LP.m10 have a significant correlation with TP90. It also shows that there is a negative correlation between the subject's age and TP90.

9 and 10, it can be seen that LP.S4 to LP.S13 have a correlation coefficient with TP90 of 0.49 or more, and further, LP.S14 to LP.S90 have a correlation coefficient with TP90 of 0.8 or more. In other words, it can be seen that the area of the ellipse calculated by the Lorenz plot method using heart rate data of 4 seconds or more is correlated with TP90 calculated by the maximum entropy method. From these results, it can be seen that by estimating TP using Equation 6 using the Lorenz plot method, it is possible to estimate the subject's TP even if the measurement time of the heart rate data is short.

(TP Experimental Results No. 2)
Fig. 11 is a diagram showing the results of multiple regression analysis of LP.S, LP.m, age and TP. More specifically, Fig. 11 shows the results of multiple regression analysis of 500 subjects, with TP90 as the dependent variable and LP.S, LP.m and age as independent variables. In Fig. 11, logarithmic transformation values were used for LP.S, LP.m and TP.

The number of seconds in Figure 11 indicates the measurement time of the heart rate data used in the multiple regression analysis. For example, in the column where the measurement time of the heart rate data is 15 seconds, the multiple correlation coefficient is 0.83456 and the degree of freedom adjusted coefficient of determination is 0.69466. This shows that the result of multiple regression analysis of the heart rate data obtained by measuring 500 subjects for 15 seconds for TP90, LP.S15, LP.m15, and the subject's age was that the multiple correlation coefficient was 0.83456 and the degree of freedom adjusted coefficient of determination was 0.69466.

According to FIG. 11, the multiple correlation coefficient obtained from heart rate data with a measurement time of 3 to 8 seconds is 0.67 or more, and furthermore, when the measurement time is 9 seconds or more, the multiple correlation coefficient is 0.8 or more. Furthermore, the coefficient of determination adjusted for the degrees of freedom exceeds 0.5 when the measurement time is 4 seconds or more. Therefore, when the measurement time is 4 seconds or more, it can be said that Equation 7 is at a level that can be used as an estimation model for TP. Furthermore, when the measurement time is 3 seconds or more, the multiple correlation coefficient exceeds 0.6, so it can be said that even heart rate data with a measurement time of 3 seconds or more and less than 4 seconds can be used as a simple measurement of the autonomic nerves.

Figure 12 shows the correlation between LP.S, LP.m, age, and TP. Figures 13 and 14 show the correlation between TP and LP.S. Figures 12 to 14 show the results of measurements taken on 500 subjects. In Figures 12 to 14, the numbers in the upper rows are correlation coefficients (r values), and the numbers in the lower rows are significance probabilities (p values). Note that logarithmic transformation values were used for LP.S, LP.m, and TP in Figures 12 to 14.

Figure 12 shows that LP.S10 and LP.m10 have a significant correlation with TP90. It also shows that there is a negative correlation between the subject's age and TP90.

Also, from Figures 13 and 14, it can be seen that LP.S3 to LP.S12 have a correlation coefficient with TP90 of 0.49 or more, and furthermore, LP.S13 to LP.S90 have a correlation coefficient with TP90 of 0.8 or more. In other words, it can be seen that the area of the ellipse calculated by the Lorenz plot method using heart rate data of 3 seconds or more is correlated with TP90 calculated by the maximum entropy method. From these results, it can be seen that by using the Lorenz plot method to estimate TP using Equation 7, it is possible to estimate the subject's TP even if the measurement time of the heart rate data is short.

Based on the experimental results 1 and 2 described above, the acquisition unit 101 may acquire the subject's heart rate data by measuring the subject for 3 seconds or more and less than 60 seconds. The calculation unit 102 may calculate (estimate) the subject's TP by the Lorenz plot method using the heart rate data acquired by measuring the subject for 3 seconds or more and less than 60 seconds. This makes it possible to evaluate the subject's autonomic nervous function in a shorter time than before.

The acquisition unit 101 may acquire the subject's heart rate data by measuring the subject for 3 seconds or more and less than 25 seconds. The calculation unit 102 may calculate (estimate) the subject's TP by the Lorenz plot method using the heart rate data acquired by measuring the subject for 3 seconds or more and less than 25 seconds. As described above, when using the maximum entropy method, a minimum of 25 seconds of measurement is required to calculate LP. Therefore, by measuring the subject for less than 25 seconds, it becomes possible to evaluate the subject's autonomic nerve function in a short time that was previously impossible.

In addition, without being limited to the above, the acquisition unit 101 may acquire the subject's heart rate data by measuring the subject for N seconds or more and less than M seconds. Furthermore, the calculation unit 102 may calculate (estimate) the subject's TP by the Lorenz plot method using the heart rate data acquired by measuring the subject for N seconds or more and less than M seconds. Any time may be set for N and M in the range of 4 seconds to 60 seconds. For example, N and M may be 3 seconds and 15 seconds, or 8 seconds and 60 seconds, respectively. Furthermore, N and M may be 8 seconds and 15 seconds, respectively. Furthermore, N and M may be 4 seconds and 15 seconds, or 4 seconds and 60 seconds, or 4 seconds and 25 seconds, respectively.

(Experimental results on the area of an ellipse)
FIG. 15 is a diagram showing the correlation of LP.S. LP.S(n,n+1) indicates the area of the ellipse LP.S when points consisting of two consecutive R-R intervals are plotted. LP.S(n,n+2) indicates the area of the ellipse LP.S when points consisting of two non-consecutive R-R intervals (when α=2) are plotted. LP.S(n,n+3) indicates the area of the ellipse LP.S when points consisting of two non-consecutive R-R intervals (when α=3) are plotted. LP.S(n,random) indicates the area of the ellipse LP.S when points consisting of two non-consecutive R-R intervals (when α=random) are plotted.

As shown in FIG. 15A, there is a significant correlation between the ellipse area LP.S(n,n+1) and the ellipse area LP.S(n,n+2). As shown in FIG. 15B, there is a significant correlation between the ellipse area LP.S(n,n+1) and the ellipse area LP.S(n,n+3). As shown in FIG. 15C, there is a significant correlation between the ellipse area LP.S(n,n+1) and the ellipse area LP.S(n,random).

(Evaluation of the parasympathetic nervous system)
FIG. 16 is a flowchart showing an example of a processing procedure for the evaluation device 10 to evaluate the parasympathetic nerves.

In step S21, the acquisition unit 101 acquires the subject's age and heart rate data.

In step S22, the calculation unit 102 calculates the subject's HF using the maximum entropy method.

２．分布データ（ｃｃｖＨＦ）１００ｂから、被験者の年齢に対応するｃｃｖＨＦの平均値と標準偏差（σ）を取得する。
３．「ｃｃｖＨＦの偏差値＝１０×（被験者のｃｃｖＨＦ－被験者の年齢に対応するｃｃｖＨＦの平均値）÷被験者の年齢に対応するｃｃｖＨＦの標準偏差（σ）＋５０」の式を用いて、被験者のｃｃｖＨＦの偏差値を算出する。 In step S23, the calculation unit 102 calculates the ccvHF of the subject using the HF of the subject calculated in step S22 and the heart rate data of the subject. The calculation unit 102 also calculates the standard deviation of the ccvHF of the subject using the ccvHF of the subject and the distribution data (ccvHF) 100b. For example, the calculation unit 102 can calculate the standard deviation of the ccvHF of the subject by the following procedure.
1. Calculate the subject's ccvHF by correcting the HF calculated in step S22 with the subject's heart rate (more specifically, the average R-R interval) according to Equation 8.

2. From the distribution data (ccvHF) 100b, the mean value and standard deviation (σ) of ccvHF corresponding to the subject's age are obtained.
3. Calculate the deviation score of the subject's ccvHF using the formula: "deviation score of ccvHF = 10 x (subject's ccvHF - mean ccvHF for subject's age) / standard deviation of ccvHF for subject's age (σ) + 50."

In step S24, the calculation unit 102 calculates the ratio of the subject's HF calculated in the processing procedure of step S22 to the subject's TP calculated based on the Lorenz plot method (i.e., HF/TP). Note that the calculation unit 102 may calculate the ratio of the subject's ccvHF calculated in the processing procedure of step S23 to the subject's ccvTP calculated based on the Lorenz plot method (i.e., ccvHF/ccvTP). Also, the processing procedure of step S24 may be omitted.

(Modification 1 of Parasympathetic Nerve Evaluation)
In the processing procedure of step S22 of the parasympathetic nerve evaluation method described above, the calculation unit 102 may calculate the HF of the subject based on the Lorenz plot method.

数式９において、４つの係数（ｉ、ｊ、ｋ、ｌ）は、心拍データの測定時間ごとに、ｌｏｇＨＦ９０を従属変数とし、ｌｏｇＬＰ．Ｓ、ｌｏｇＬＰ．ｍ及び年齢を独立変数として重回帰分析を行うことで算出された係数であってもよい。すなわち、記憶部１００に、心拍データの測定時間ごとに４つの係数（ｉ、ｊ、ｋ、ｌ）のセットを予め記憶しておき、算出部１０２は、心拍データの測定時間に対応する係数のセットを記憶部１００から取得し、取得した係数を数式９に適用することで、被験者のｌｏｇＨＦを算出（推定）するようにしてもよい。例えば、心拍データの測定時間が連続１０秒である場合、算出部１０２は、係数としてｉ＝０．６５６６０、ｊ＝１．８１０７４、ｋ＝－０．００７２及びｌ＝－５．５８８０を利用するようにしてもよい。なお、当該４つの係数は、心拍データの測定時間に関わらず固定値であってもよい。この場合、４つの係数は、ｉ＝０．６５６６０、ｊ＝１．８１０７４、ｋ＝－０．００７２及びｌ＝－５．５８８０であってもよい。 また、算出部１０２は、推定したｌｏｇＨＦから推定ＨＦ（つまり対数変換前のＨＦ）を算出し、算出した推定ＨＦを用いてステップＳ２３の処理手順を実行するようにしてもよい。 Specifically, in "Step 5" of the process for calculating the subject's TP described in "(Rapid evaluation of autonomic nervous function)", logHF may be estimated by using Equation 9 instead of Equation 6. The logarithm may be a common logarithm.

In Equation 9, the four coefficients (i, j, k, l) may be coefficients calculated by performing multiple regression analysis with logHF90 as the dependent variable and logLP.S, logLP.m, and age as independent variables for each measurement time of the heartbeat data. That is, a set of four coefficients (i, j, k, l) for each measurement time of the heartbeat data may be stored in advance in the storage unit 100, and the calculation unit 102 may obtain a set of coefficients corresponding to the measurement time of the heartbeat data from the storage unit 100 and apply the obtained coefficients to Equation 9 to calculate (estimate) the logHF of the subject. For example, when the measurement time of the heartbeat data is 10 consecutive seconds, the calculation unit 102 may use i=0.65660, j=1.81074, k=-0.0072, and l=-5.5880 as coefficients. The four coefficients may be fixed values regardless of the measurement time of the heartbeat data. In this case, the four coefficients may be i = 0.65660, j = 1.81074, k = -0.0072, and l = -5.5880. The calculation unit 102 may also calculate an estimated HF (i.e., HF before logarithmic transformation) from the estimated logHF, and execute the processing procedure of step S23 using the calculated estimated HF.

The noise removal process described in "(Rapid evaluation of autonomic nerve function)" may also be applied to the method of estimating the HF of the subject based on the Lorenz plot method. That is, the calculation unit 102 may perform the following process.
1. Two consecutive intervals of heart rate data, designated x and y, are plotted on a graph with one axis as the x-axis and the other as the y-axis.
2. A first ellipse is calculated with the major axis being the first standard deviation obtained by multiplying the standard deviation of the distance between the origin (0,0) and the projection points obtained by projecting the points plotted on the graph onto the x = y axis by a specified coefficient (D), and the minor axis being the second standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points obtained by projecting the points plotted on the graph onto the x = -y axis by a specified coefficient (D).
3. For points that are plotted on the graph and exist inside ellipse M, a second ellipse is calculated whose major axis is the third standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points of the plotted points on the graph onto the x = y axis by a predetermined coefficient (D), and whose minor axis is the fourth standard deviation obtained by multiplying the standard deviation of the distance between the origin and the projection points of the plotted points on the graph onto the x = -y axis by a predetermined coefficient (D).
4. The area of the second ellipse is substituted into a predetermined formula (Formula 9) to estimate (calculate) the log HF of the subject.

Furthermore, when calculating the HF of the subject based on the Lorenz plot method, the calculation unit 102 may calculate LP.S and LP.m by plotting on a graph a point (x, y) = (RRI _n , RRI _{n +} α) consisting of two non-consecutive R-R intervals among the R-R intervals of the heart rate data, as described in "Modification 2 of Rapid Evaluation of Autonomic Nervous Function". Anything not mentioned in particular may be the same as the description in "Modification 2 of Rapid Evaluation of Autonomic Nervous Function".

(Modification 2 of Parasympathetic Nerve Evaluation)
In the process procedure of step S23 described above, ccvHF is calculated using formula 8, but is not limited to this. In this embodiment, ccvHF may be calculated using any of formulas 10 to 13 below. Note that "average HF" is the average value of multiple HF values obtained by measuring a predetermined number of subjects. Also, "standard deviation of HF" is the standard deviation (σ) of multiple HF values obtained by measuring a predetermined number of subjects. Also, "HF _max " and "HF _min " respectively mean the maximum and minimum values of multiple HF values obtained by measuring a predetermined number of subjects.

(Modification 3 of parasympathetic nerve evaluation)
In step S23, the calculation unit 102 calculates the ccvHF standard deviation of the subject using the procedures 1 to 3. However, instead of this, the calculation unit 102 may calculate the HF standard deviation of the subject using the following procedure.

すなわち、算出部１０２は、被験者の心拍データのＲ－Ｒ間隔のうち２つのＲ－Ｒ間隔をそれぞれｘ及びｙとする複数の点を、一方の軸がｘ軸であり他方の軸がｙ軸であるグラフにプロットすることで、ローレンツプロット法における楕円面積（ＬＰ．Ｓ）及び平均距離（ＬＰ．ｍ）とを算出してもよい。また、算出部１０２は、算出した楕円面積（ＬＰ．Ｓ）及び平均距離（ＬＰ．ｍ）に基づいて、被験者のＬＰ指標を算出するようにしてもよい。 1. Calculate the Lorenz plot (LP) index using the following formula 14. The LP index may be called the first index related to the Lorenz plot method. When calculating LP.S and LP.m, the calculation unit 102 may calculate LP.S and LP.m by plotting on a graph a point (x, y) = (RRI _n , RRI _{n +} α) consisting of two non-consecutive R-R intervals among the R-R intervals of the heartbeat data, as described in "Modification 2 of the rapid evaluation of the autonomic nervous function". Points that are not particularly mentioned may be the same as those described in "Modification 2 of the rapid evaluation of the autonomic nervous function". Alternatively, the calculation unit 102 may calculate LP.S and LP.m by plotting on a graph a point (x, y) = (RRI _n , RRI _{n + 1} ) consisting of two consecutive R-R intervals.

That is, the calculation unit 102 may calculate the ellipse area (LP.S) and the average distance (LP.m) in the Lorenz plot method by plotting a plurality of points, each of which has two R-R intervals of the R-R intervals of the heart rate data of the subject as x and y, on a graph with one axis as the x-axis and the other axis as the y-axis. The calculation unit 102 may also calculate the LP index of the subject based on the calculated ellipse area (LP.S) and average distance (LP.m).

Note that, when LP.S = S and LP.m = M, the S/M part in Equation 14 may be replaced with a logarithmic or normalized value. Specifically, the value of "S/M" may be replaced with any of the values shown below. S/logM, S/rootM, logS/M, logS/logM, logS/rootM, rootS/M, rootS/logM, rootS/rootM, log(S/M), log(S/logM), log(S/rootM), log(logS/M), log(logS/logM), log(logS/rootM), log(rootS/M), log(rootS/logM), log(rootS/rootM), root(S/M), root(S/logM), root(S/rootM), root(logS/M), root(logS/logM), root(logS/rootM), root(rootS/M), root(rootS/logM), root(rootS/rootM).

2. Calculate the deviation value of the subject's LP index using the formula: "LP index deviation value = 10 x (subject's LP index - average value of LP index corresponding to subject's age) ÷ standard deviation of LP index corresponding to subject's age (σ) + 50."

To enable calculation of the standard deviation of the LP index, the storage unit 100 may store distribution data indicating the distribution of the LP index by the subject's age. Furthermore, the calculation unit 102 may calculate the standard deviation of the subject's LP index based on the distribution data, the subject's age, and the subject's LP index.

すなわち、算出部１０２は、被験者のＬＰ指標の偏差値を所定計算式（数式１５）に代入することで、被験者のＨＦ値の偏差値を算出するようにしてもよい。 3. Calculate the HF deviation value using the following Equation 15.

That is, the calculation unit 102 may calculate the standard deviation of the subject's HF value by substituting the standard deviation of the subject's LP index into a predetermined formula (Formula 15).

In formula 15, the two coefficients (m, n) may be coefficients calculated by performing regression analysis with the HF deviation value calculated by frequency analysis as the dependent variable and the deviation value of the LP index as the independent variable for each measurement time of the heartbeat data. That is, a set of two coefficients (m, n) for each measurement time of the heartbeat data may be stored in advance in the storage unit 100, and the calculation unit 102 may obtain from the storage unit 100 a set of coefficients corresponding to the measurement time of the heartbeat data, and apply the obtained coefficients to formula 15 to calculate (estimate) the HF deviation value of the subject.

In addition, in Equation 15, the two coefficients (m, n) can be coefficients calculated from the coefficient o of the major axis and the coefficient p of the minor axis used in calculating LP.S. For example, using the major and minor axes shown in FIG. 6, the coefficients o = 2σx and p = 2σ-x can be used.

According to the third modified example of the parasympathetic nerve evaluation, the LP index is calculated using the area and distance values themselves, which are actual measurements calculated by the Lorenz plot method, and the HF deviation value is calculated from the deviation value of the LP index. This makes it possible to reduce errors in the measurement values compared to when the HF value is first estimated and then converted into a deviation value.

数式１６において、２つの係数（ｑ、ｒ）は、心拍データの測定時間ごとに、周波数解析により算出したＴＰ偏差値を従属変数とし、ＬＰ指標の偏差値を独立変数として回帰分析を行うことで算出された係数であってもよい。すなわち、記憶部１００に、心拍データの測定時間ごとに２つの係数（ｑ、ｒ）のセットを予め記憶しておき、算出部１０２は、心拍データの測定時間に対応する係数のセットを記憶部１００から取得し、取得した係数を数式１６に適用することで、被験者のＴＰ偏差値を算出（推定）するようにしてもよい。 (Evaluation of autonomic nervous function using the LP index)
In "Modification 3 of Parasympathetic Evaluation", the HF deviation value is calculated from the LP index, but it is also possible to calculate the TP deviation value from the LP index. Specifically, the calculation unit 102 may calculate the deviation value of the subject's TP value by substituting the deviation value of the subject's LP index calculated in "Modification 3 of Parasympathetic Evaluation" into the following formula 16 instead of formula 15. That is, the calculation unit 102 may calculate the deviation value of the subject's TP value by substituting the deviation value of the subject's LP index into a predetermined calculation formula (formula 16).

In Equation 16, the two coefficients (q, r) may be coefficients calculated by performing regression analysis with the TP deviation value calculated by frequency analysis for each measurement time of the heartbeat data as the dependent variable and the deviation value of the LP index as the independent variable. That is, a set of two coefficients (q, r) for each measurement time of the heartbeat data may be stored in advance in the storage unit 100, and the calculation unit 102 may obtain a set of coefficients corresponding to the measurement time of the heartbeat data from the storage unit 100 and apply the obtained coefficients to Equation 16 to calculate (estimate) the TP deviation value of the subject.

In addition, in Equation 16, the two coefficients (q, r) can be coefficients calculated from the coefficient o of the major axis and the coefficient p of the minor axis used in calculating LP.S. For example, using the major and minor axes shown in FIG. 6, the coefficients q = 3σx and r = 3σ-x can be used.

According to the evaluation of autonomic nerve function using the LP index, the LP index is calculated using the area and distance values, which are the actual measured values calculated by the Lorenz plot method, and the TP deviation value is calculated from the deviation value of the LP index. This makes it possible to reduce the error of the measurement value compared to the case where the TP value is first estimated and then converted into a deviation value.

(Sympathetic nerve evaluation)
FIG. 17 is a flowchart showing an example of a processing procedure for the evaluation device 10 to evaluate the sympathetic nerves.

In step S31, the acquisition unit 101 acquires the subject's age and heart rate data.

In step S32, the calculation unit 102 calculates the subject's LF using the maximum entropy method.

２．分布データ（ｃｃｖＬＦ）１００ｃから、被験者の年齢に対応するｃｃｖＬＦの平均値と標準偏差（σ）を取得する。
３．「ｃｃｖＬＦの偏差値＝１０×（被験者のｃｃｖＬＦ－被験者の年齢に対応するｃｃｖＬＦの平均値）÷被験者の年齢に対応するｃｃｖＬＦの標準偏差（σ）＋５０」の式を用いて、被験者のｃｃｖＬＦの偏差値を算出する。 In step S33, the calculation unit 102 calculates the ccvLF of the subject using the LF of the subject calculated in step S32 and the heart rate data of the subject. The calculation unit 102 also calculates the standard deviation of the ccvLF of the subject using the ccvLF of the subject and the distribution data (ccvLF) 100c. For example, the calculation unit 102 can calculate the standard deviation of the ccvLF of the subject by the following procedure.
1. Calculate the subject's ccvLF by correcting the LF calculated in step S32 with the subject's heart rate (more specifically, the average R-R interval) according to Equation 10.

2. From the distribution data (ccvLF) 100c, the mean value and standard deviation (σ) of ccvLF corresponding to the subject's age are obtained.
3. Calculate the deviation value of the subject's ccvLF using the formula: "Deviation value of ccvLF = 10 x (subject's ccvLF - mean value of ccvLF corresponding to subject's age) / standard deviation (σ) of ccvLF corresponding to subject's age + 50."

(Sympathetic nerve evaluation variation 1)
In the processing procedure of step S33 described above, ccvLF is calculated using formula 10, but is not limited to this. In this embodiment, ccvLF may be calculated using any of formulas 18 to 21 below. Note that "average LF" is the average value of multiple LFs obtained by measuring a predetermined number of subjects. Furthermore, "standard deviation of LF" is the standard deviation (σ) of multiple LFs obtained by measuring a predetermined number of subjects. Furthermore, "LF _max " and "LF _min " respectively mean the maximum and minimum values of multiple LFs obtained by measuring a predetermined number of subjects.

(HF Experimental Results Part 1)
In the explanation of "HF Experimental Results 1" and "HF Experimental Results 2" described later, "HFxx" means HF calculated by the maximum entropy method using heart rate data measured continuously for xx seconds. For example, "HF90" means HF calculated by the maximum entropy method using heart rate data measured continuously for 90 seconds.

HF analyzes the frequency band of 0.15 to 0.4 Hz, so one wavelength is at least 1/0.15 (≒6.66) seconds. In other words, HF can be measured if there is heart rate data of at least 1/0.15 seconds. Figure 18 shows the results of a correlation analysis between HF and HF90 obtained from heart rate data of 1/0.15 seconds or more. Figure 18 shows the results of measurements taken on 500 subjects. In Figure 18, the numbers in the upper row are correlation coefficients (r values), and the numbers in the lower row are significance probabilities (p values). Figure 18 shows that the correlation coefficients with HF90 for HF6.66 to HF10 are generally 0.5 or more, and furthermore, the correlation coefficients with HF90 for HF11 to HF60 are generally 0.8 or more. In other words, it can be seen that HF calculated by the maximum entropy method using 6.66 seconds of heart rate data is correlated with HF90 calculated by the maximum entropy method. Furthermore, even when heart rate data is measured for the theoretically shortest time (i.e., 1/0.15 seconds), the correlation coefficient with HF90 is 0.5 or more, confirming that it can be used as a simple measurement of the parasympathetic nervous system.

In some literature, HF is defined as the frequency band of 0.2 to 0.5 Hz, not the frequency band of 0.15 to 0.4 Hz. In accordance with this definition, a correlation analysis was performed between HF obtained by analyzing the frequency band of 0.2 to 0.5 Hz and HF90 obtained by analyzing the frequency band of 0.2 to 0.5 Hz. The results are shown in FIG. 19. As shown in FIG. 19, even in a measurement of 5 seconds, which is the theoretically shortest time, the correlation coefficient with HF90 was found to be 0.5 or more. Therefore, even if the definition of HF is changed and heart rate data is measured in the theoretically shortest time (1/0.2 seconds, i.e. 5 seconds), the correlation coefficient with HF90 is found to be 0.5 or more, confirming that it can be used as a simple measurement of the parasympathetic nervous system.

Based on the experimental result 1 described above, the acquisition unit 101 may acquire the subject's heart rate data by measuring the subject for a time period that is equal to or longer than the shortest measurement time based on the definition of the frequency components when calculating HF by frequency analysis and less than 60 seconds. The calculation unit 102 may calculate the subject's HF by the maximum entropy method using the heart rate data acquired by measuring the subject for a time period that is equal to or longer than the shortest measurement time based on the definition of the frequency components when calculating HF by frequency analysis and less than 60 seconds. This makes it possible to evaluate the deviation value of the subject's HF in a shorter time than before.

If the HF frequency component is defined as 0.15 to 0.4 Hz, the above shortest measurement time is 1/0.15 seconds (≒6.66 seconds). If the HF frequency component is defined as 0.2 to 0.5 Hz, the above shortest measurement time is 1/0.2 seconds (=5 seconds).

In addition, without being limited to the above, the acquisition unit 101 may acquire the subject's heart rate data by measuring the subject for P seconds or more and less than Q seconds. Furthermore, the calculation unit 102 may calculate the subject's HF by the maximum entropy method using the heart rate data acquired by measuring the subject for P seconds or more and less than Q seconds. P and Q may be set to any time within the range of the shortest measurement time to 60 seconds. For example, P and Q may be 11 seconds and 60 seconds, respectively, or 11 seconds and 15 seconds, respectively.

(HF Experimental Results Part 2)
Fig. 20 is a diagram showing the results of multiple regression analysis of LP.S, LP.m, age and HF. More specifically, Fig. 20 shows the results of multiple regression analysis of 500 subjects, with HF90 as the dependent variable and LP.S, LP.m and age as independent variables. In Fig. 20, logarithmic transformation values were used for LP.S, LP.m and HF.

The number of seconds in Figure 20 indicates the measurement time of the heart rate data used in the multiple regression analysis. For example, in the column where the measurement time of the heart rate data is 15 seconds, the multiple correlation coefficient is 0.85625 and the degree of freedom adjusted coefficient of determination is 0.73154. This shows that the result of multiple regression analysis of the heart rate data obtained by measuring 500 subjects for 15 seconds for HF90, LP.S15, LP.m15, and the subject's age was that the multiple correlation coefficient was 0.85625 and the degree of freedom adjusted coefficient of determination was 0.73154.

According to FIG. 20, the multiple correlation coefficient obtained from heart rate data with a measurement time of 3 to 5 seconds is 0.71 or more, and furthermore, when the measurement time is 6 seconds or more, the multiple correlation coefficient is 0.8 or more. Furthermore, the degree of freedom-adjusted coefficient of determination exceeds 0.5 when the measurement time is 3 seconds or more. Therefore, when the measurement time is 3 seconds or more, it can be said that Equation 9 is at a level that can be used as an estimation model for HF.

Figure 21 shows the correlation between LP.S, LP.m, age, and HF. Figures 22 and 23 show the correlation between TP and LP.S. Figures 21 to 23 show the results of measurements taken on 500 subjects. In Figures 21 to 23, the numbers in the upper rows are correlation coefficients (r values), and the numbers in the lower rows are significance probabilities (p values). Note that logarithmic transformed values were used for LP.S, LP.m, and HF in Figures 21 to 23.

Figure 21 shows that LP.S10 and LP.m10 have a significant correlation with TP90. It also shows that there is a negative correlation between the subject's age and TP90.

22 and 23, it can be seen that LP.S3 to LP.S9 have a correlation coefficient with HF90 of 0.6 or more, and further, LP.S10 to LP.S90 have a correlation coefficient with HF90 of 0.8 or more. In other words, it can be seen that the area of the ellipse calculated by the Lorenz plot method using heart rate data of 3 seconds or more is correlated with HF90 calculated by the maximum entropy method. From these results, it can be seen that by using the Lorenz plot method to estimate HF with Equation 9, it is possible to estimate the subject's HF even if the measurement time of the heart rate data is short.

Based on the second experimental result described above, the acquisition unit 101 may acquire the subject's heart rate data by measuring the subject at HF for 3 seconds or more and less than 60 seconds. The calculation unit 102 may calculate (estimate) the subject's HF by the Lorenz plot method using the heart rate data acquired by measuring the subject at HF for 3 seconds or more and less than 60 seconds. This makes it possible to evaluate the subject's HF standard deviation in a shorter time than conventional methods.

In addition, without being limited to the above, the acquisition unit 101 may acquire the subject's heart rate data by measuring the subject for R seconds or more and less than S seconds. Furthermore, the calculation unit 102 may calculate (estimate) the subject's HF by the Lorenz plot method using the heart rate data acquired by measuring the subject for R seconds or more and less than S seconds. Any time may be set for R and S within the range of 3 to 60 seconds. For example, R and S may be 6 seconds and 60 seconds, 6 seconds and 15 seconds, 3 seconds and 15 seconds, or 3 seconds and 7 seconds, respectively.

(Additional information regarding evaluation of the parasympathetic nervous system)
FIG. 24 is a diagram showing the results of comparing the percentage of HF in TP with the subject's fatigue level. FIG. 24 shows data obtained by performing an autonomic nerve function evaluation on 19 healthy subjects before and after fatigue caused by mental work load, and the vertical axis "% HF" shows the percentage of HF in TP (i.e., HF÷TP×100). The horizontal axis shows a score indicating the subject's fatigue level, and the higher the score, the more fatigued the subject is. According to FIG. 24, it can be seen that the lower the percentage of HF in TP, the higher the subject's fatigue level. Note that TP is calculated as LF+HF, so the lower the percentage of HF in TP, the higher the percentage of LF.

The evaluation device 10 may calculate the ratio of the subject's HF to the subject's TP calculated based on the Lorenz plot method (i.e., HF/TP) and output information indicating the calculated HF ratio. As described above, the evaluation device 10 according to this embodiment is capable of measuring TP in a shorter time than conventional methods by using the Lorenz plot method, and furthermore, is capable of measuring HF in a shorter time than conventional methods. Therefore, the "ratio of HF to TP" of the subject can also be calculated in a shorter time than conventional methods. Therefore, in this embodiment, it is possible to determine the state of fatigue or stress-related disorders in an extremely short time and in a situation where the burden of measurement on the subject is reduced.

(Example output)
FIG. 25 is a diagram showing an example of a screen for outputting the measurement result of the subject. As shown in FIG. 25, the output unit 103 may display the deviation value of the subject's ccvTP and the deviation value of the subject's ccvHF value side by side on the screen. The upper part of FIG. 25 corresponds to the deviation value of the ccvTP, and the lower part of FIG. 25 corresponds to the deviation value of the ccvHF. This allows the subject to check the ccvTP and ccvHF at the same time. In addition, according to this embodiment, it is possible to calculate the deviation value of the ccvTP from heart rate data measured in a time period that was difficult in the past, such as less than 25 seconds, so that the subject can grasp the degree of fatigue and the magnitude of stress of his or her own brain without feeling stress.

In addition, by measuring the subject in less than 7 seconds, for example, it is possible to calculate the deviation value of ccvHF, and evaluate the subject's parasympathetic autonomic nerve function in a short time that was previously impossible. In addition to estimating parasympathetic nerve activity from HF using frequency analysis, estimating parasympathetic nerve activity from the area of the Lorenz plot method can increase the reliability of evaluations in a short time.

<Summary>
According to the embodiment described above, the Lorenz plot method is used instead of the maximum entropy method, so that TP can be calculated from heart rate data measured in a short time. This makes it possible to more quickly evaluate the autonomic nerve function. According to the present embodiment, the deviation value of ccvHF is calculated, so that the index related to the parasympathetic nerve can be used as a common index even between subjects of different ages. According to the present embodiment, the deviation value of ccvLF is calculated, so that the index related to the sympathetic nerve can be used as a common index even between subjects of different ages. According to the present embodiment, the deviation value of ccvLF is calculated, so that the index related to the sympathetic nerve can be used as a common index even between subjects of different ages. According to the present embodiment, the deviation value of ccvTP and the deviation value of ccvHF calculated from heart rate data measured in a short time are simultaneously output, so that the state of the parasympathetic nerve can be more quickly understood than before. According to the present embodiment, the deviation value of ccvTP and the deviation value of ccvHF calculated from heart rate data measured in a short time are simultaneously output, so that the state of the parasympathetic nerve can be more quickly understood than before. According to the present embodiment, the ratio of HF value measured in a short time to TP (HF/TP) is output, so that the state of the parasympathetic nerves (for example, whether the ratio of HF to TP is high or low, etc.) can be grasped more quickly than before. In addition to estimating parasympathetic nerve activity from HF by frequency analysis, the parasympathetic nerve activity can be estimated from the area of the Lorenz plot method, so that the reliability of the evaluation in a short time can be improved. According to the present embodiment, the deviation value of ccvTP and the deviation value of HF calculated from the heart rate data measured in a short time can be output simultaneously, so that the subject can grasp the state of his/her autonomic nerves in a manner that can be compared with other subjects without feeling stress.

The above-described embodiments are intended to facilitate understanding of the present invention, and are not intended to limit the present invention. The flow charts, sequences, elements included in the embodiments, and their arrangements, materials, conditions, shapes, sizes, etc., described in the embodiments are not limited to those exemplified, and may be modified as appropriate. In addition, configurations shown in different embodiments may be partially substituted or combined. The HF value and LF value may be referred to as HF and LF, respectively.

10...Evaluation device, 11...Processor, 12...Storage device, 13...Input device, 14...Output device, 15...Biometric information acquisition device, 100...Storage unit, 100a...Distribution data (ccvTP), 100b...Distribution data (ccvHF), 100c...Distribution data (ccvLF), 101...Acquisition unit, 102...Calculation unit, 103...Output unit

Claims (11)

An acquisition unit that acquires heart rate data of a subject to be evaluated , the heart rate data being measured for 3 seconds or more and less than 60 seconds , and the age of the subject ;
a calculation unit that calculates an ellipse area and an average distance in a Lorenz plot method based on the RR interval of the heart rate data, and calculates an autonomic nerve activity amount of the subject based on the ellipse area, the average distance and the age of the subject;
An autonomic nervous system evaluation device having the same. A storage unit that stores distribution data showing a distribution of autonomic nerve activity by age of subjects,
the calculation unit calculates a standard deviation of the autonomic nerve activity of the subject by using the autonomic nerve activity of the subject, the heart rate data of the subject, and the distribution data;
The autonomic nerve evaluation device according to claim 1 . The calculation unit is
Plotting a plurality of points, each of which represents two consecutive RR intervals among the RR intervals of the heartbeat data as x and y, on a graph having an x-axis and a y-axis as one axis;
a first ellipse is calculated, the major axis of which is a first standard deviation obtained by multiplying a standard deviation of a distance between the origin and a projection point of the points plotted on the graph onto the x=y axis by a predetermined coefficient, and the minor axis of which is a second standard deviation obtained by multiplying a standard deviation of a distance between the origin and a projection point of the points plotted on the graph onto the x=-y axis by the predetermined coefficient;
a second ellipse is calculated for points that are inside the first ellipse among the points plotted on the graph, with a third standard deviation as its major axis being obtained by multiplying the standard deviation of the distance between the origin and the projection points of the points plotted on the graph onto the x=y axis by the predetermined coefficient, and a fourth standard deviation as its minor axis being obtained by multiplying the standard deviation of the distance between the origin and the projection points of the points plotted on the graph onto the x=-y axis by the predetermined coefficient;
Calculating the amount of autonomic nerve activity of the subject by substituting the area of the second ellipse, which is the ellipse area, the average distance, and the age of the subject into a predetermined calculation formula.
The autonomic nerve evaluation device according to claim 1 . The calculation unit is
Plotting a plurality of points, each of which is an x-axis and a y-axis representing two non-consecutive RR intervals among the RR intervals of the heartbeat data, on a graph having an x-axis and a y-axis;
a first ellipse is calculated, the major axis of which is a first standard deviation obtained by multiplying a standard deviation of a distance between the origin and a projection point of the points plotted on the graph onto the x=y axis by a predetermined coefficient, and the minor axis of which is a second standard deviation obtained by multiplying a standard deviation of a distance between the origin and a projection point of the points plotted on the graph onto the x=-y axis by the predetermined coefficient;
a second ellipse is calculated for points that are inside the first ellipse among the points plotted on the graph, with a third standard deviation as its major axis being obtained by multiplying the standard deviation of the distance between the origin and the projection points of the points plotted on the graph onto the x=y axis by the predetermined coefficient, and a fourth standard deviation as its minor axis being obtained by multiplying the standard deviation of the distance between the origin and the projection points of the points plotted on the graph onto the x=-y axis by the predetermined coefficient;
Calculating the amount of autonomic nerve activity of the subject by substituting the area of the second ellipse, which is the ellipse area, the average distance, and the age of the subject into a predetermined calculation formula.
The autonomic nerve evaluation device according to claim 1 . the calculation unit plots a plurality of points, each of which represents the two non-consecutive RR intervals as x and y, and each of which has a different number of other RR intervals present between the two non-consecutive RR intervals, on a graph having an x-axis and a y-axis as one axis;
The autonomic nerve evaluation device according to claim 4 . The acquisition unit acquires the heart rate data of the subject by measuring the heart rate of the subject for 3 seconds or more and less than 25 seconds.
The autonomic nerve evaluation device according to any one of claims 1 to 5. acquiring heart rate data of a subject to be evaluated, the heart rate data being measured for a period of at least 3 seconds and less than 60 seconds, and the age of the subject ;
calculating an ellipse area and an average distance in a Lorenz plot method based on the RR interval of the heart rate data, and calculating an autonomic nerve activity of the subject based on the ellipse area, the average distance and the age of the subject;
An autonomic nerve evaluation method performed by an autonomic nerve evaluation device, comprising: acquiring heart rate data of a subject to be evaluated, the heart rate data being measured for a period of at least 3 seconds and less than 60 seconds, and the age of the subject ;
calculating an ellipse area and an average distance in a Lorenz plot method based on the RR interval of the heart rate data, and calculating an autonomic nerve activity of the subject based on the ellipse area, the average distance and the age of the subject;
A program for causing a computer to execute the following. a storage unit that stores distribution data indicating a distribution of a first index related to the Lorenz plot method by age of subjects;
An acquisition unit that acquires the age of a subject to be evaluated and heart rate data of the subject;
Plotting a plurality of points, each of which represents two RR intervals of the heartbeat data as x and y, on a graph having one axis as the x-axis and the other axis as the y-axis, thereby calculating an ellipse area (LP.S) and an average distance (LP.m) in the Lorentz plot method;
Calculating a first index for the subject related to the Lorenz plot method based on the ellipse area (LP.S) and the average distance (LP.m);
calculating a standard deviation of the first index of the subject based on the distribution data, the age of the subject, and the first index of the subject;
a calculation unit that calculates a standard deviation value of the autonomic nerve activity of the subject by substituting a standard deviation value of the first index of the subject into a predetermined calculation formula that outputs a standard deviation value of the autonomic nerve activity of the subject when a coefficient corresponding to a measurement time of the heartbeat data and the standard deviation value of the first index of the subject are input;
An autonomic nervous system evaluation device having the above configuration. storing distribution data indicating a distribution of the first index related to the Lorenz plot method by age of subjects in a storage unit;
obtaining the age of a subject to be evaluated and heart rate data of said subject;
Plotting a plurality of points, each of which represents two RR intervals of the heartbeat data as x and y, on a graph having one axis as the x-axis and the other axis as the y-axis, thereby calculating an ellipse area (LP.S) and an average distance (LP.m) in the Lorentz plot method;
Calculating a first index for the subject related to the Lorenz plot method based on the ellipse area (LP.S) and the average distance (LP.m);
calculating a standard deviation of the first index of the subject based on the distribution data, the age of the subject, and the first index of the subject;
calculating a standard deviation value of the amount of autonomic nerve activity of the subject by substituting the standard deviation value of the first index of the subject into a predetermined calculation formula that outputs a standard deviation value of the amount of autonomic nerve activity of the subject when a coefficient corresponding to a measurement time of the heart rate data and the standard deviation value of the first index of the subject are input;
An autonomic nervous system evaluation method comprising: storing distribution data indicating a distribution of the first index related to the Lorenz plot method by age of subjects in a storage unit;
obtaining the age of a subject to be evaluated and heart rate data of said subject;
Plotting a plurality of points, each of which represents two RR intervals of the heartbeat data as x and y, on a graph having one axis as the x-axis and the other axis as the y-axis, thereby calculating an ellipse area (LP.S) and an average distance (LP.m) in the Lorentz plot method;
Calculating a first index for the subject related to the Lorenz plot method based on the ellipse area (LP.S) and the average distance (LP.m);
calculating a standard deviation of the first index of the subject based on the distribution data, the age of the subject, and the first index of the subject;
calculating a standard deviation value of the amount of autonomic nerve activity of the subject by substituting the standard deviation value of the first index of the subject into a predetermined calculation formula that outputs a standard deviation value of the amount of autonomic nerve activity of the subject when a coefficient corresponding to a measurement time of the heart rate data and the standard deviation value of the first index of the subject are input;
A program for causing a computer to execute the following.

Images