JP6865438B2 - Method and device for creating indicators to determine neuropsychiatric status - Google Patents

Description

The present invention relates to a method for creating an index capable of objectively discriminating a neuropsychiatric state and a device for creating the index.

Depression (major depressive disorder), bipolar disorder, schizophrenia, panic disorder, obsessive-compulsive disorder, autonomic imbalance, sleep disorder, etc. due to various factors such as inheritance, trauma, surrounding environment, stress, etc. May cause mental disorders. Therefore, observing the state of the psychiatric nerve is effective for early detection of illness and appropriate treatment. The presence, type or severity of mental illness may be determined by a physician interview based on known criteria. For example, the depression self-assessment scale (CES-D), the stress self-assessment scale (Hoppins Symptom Checklist: HSCL), the depression assessment scale (HAMD), the Young Mania Rating Scale (YMRS), and the like are used to determine depression. Criteria are used. However, since the conventional judgment criteria include abstract items and there is room for the subjectivity of the patient to enter, a method that can be judged objectively and mechanically is desired.

For example, Patent Document 1 discloses a psychiatric symptom and a method for evaluating the risk of developing a mental illness using the heart rate variability index for using the heart rate variability index for diagnosing a mental illness or detecting a risk of developing a psychiatric disorder. The high frequency index, low frequency index, and their ratios in the frequency analysis step of the method are measured in each of the three states of the subject in the resting state, the task performing state, and the resting state after the task is performed, and the frequency is measured. Obtained by analysis. However, the evaluation method described in Patent Document 1 imparts a task such as random number generation to the subject, and may impose an excessive burden on the subject. There is room for improvement from the perspective of easily creating indicators for discrimination.

Japanese Unexamined Patent Publication No. 2013-233256

Therefore, an object of the present invention is to provide a method for creating an index and a device for creating the index, which can easily and objectively determine the neuropsychiatric state.

The present inventors are investigating a method for discriminating a neuropsychiatric state using easily measurable biological information, and the pulsating interval and the amount of activity of a subject during awakening or sleeping are used to diagnose a mental disorder. And statistical manual (DSM-5), depression self-assessment scale (CES-D), stress self-assessment scale (HSCL), etc. The present invention was completed with the idea that it is effective as an index for this purpose.

That is, the method for discriminating the neuropsychiatric state of the present invention is the beat interval of the subject, the amount of activity expressed by the acceleration or angular velocity accompanying the movement of the subject, and the acceleration TA in the height direction of the subject. The negative acceleration NA is calculated according to the steps of measuring and the following conditions (1) to (3), and the first lying time zone, the second lying time zone, the awakening time zone, and the sleeping time zone are calculated. The gist is that it has a step to perform, a constant term, and a step to create an index represented by an expression including a beat interval of the awakening time zone × a variable term of the amount of activity.
[conditions:
(1) Calculation of negative acceleration NA When TA ≧ 0 when the subject is standing, NA = (-1) × (TA)
If TA <0 when the subject is standing, NA = TA
(2) Calculation of 1st decubitus time zone and 2nd decubitus time zone 1st decubitus time zone L _m1 : NA ≧ C1 (C1 is a constant) is the first predetermined time T ₁ or more. Position time zone L _m2 : There are two or more first _decubitus time zones L _m1 , and the gap time zone L _{sm where NA <C1 between two adjacent first decubitus time zones L m11} and L m12 is the first. 2 If it is within the predetermined time T _{2, the} time zone in which the two adjacent first recumbent time zones L _m11 and L _m12 and the gap time zone L _sm are totaled (3) Calculation of the awakening time zone and the sleeping time zone Sleeping time zone : The longest time zone of the first recumbent time zone L _m1 and the second recumbent time zone L _m2 during the predetermined measurement unit time Awakening time zone: The time zone excluding the sleep time zone from the predetermined measurement unit time]

Since the above method has the conditions (1) to (3), it is possible to objectively and accurately calculate the awakening time zone and the sleeping time zone. In addition, it is possible to easily create an index for discriminating the neuropsychiatric state from the beat interval and the amount of activity, which are easily measurable parameters. In addition, it is possible to objectively determine the neuropsychiatric state by using the created index.

The above method further uses LF, which is a value obtained by definitely integrating the power spectrum obtained by including the step of converting the beat interval into a frequency spectrum, from frequencies Lf1 to Lf2, and a value obtained by definitely integrating the power spectrum from frequencies Hf1 to Hf2. It has a step of calculating a certain HF, and it is preferable that the index includes a variable term of (LF / HF) × activity amount of the sleep time zone. Here, Hf1> Lf1 and Hf2> Lf2.

In the above method, it is preferable that the index includes a variable term of the total time of the first recumbent time zone.

In the above method, it is preferable that the index includes a variable term of the beat interval of the sleep time zone.

In the above method, it is preferable that the index includes the variable term of (LF / HF) / activity amount of the awakening time zone.

In the above method, it is preferable that the index includes the variable term of the beat interval / activity amount during the sleep time zone.

The method further comprising the step of time zone is NA ≧ C1 calculates the third recumbent hours is within a third predetermined time T _3, the third predetermined time T _3, the first predetermined time T _{It is preferably less than 1} and the index includes a variable term for the total time of the third recumbent time zone.

In the above method, the indicators are the variable term of the beat interval in the awakening time zone, the variable term of the LF / HF in the awakening time zone, the variable term of the activity amount in the awakening time zone, the variable term of the LF / HF in the sleeping time zone, Includes at least one of the variable term of the activity amount of the sleep time zone, the variable term of the time of the sleep time zone, the variable term of HF × the activity amount of the awakening time zone, and the variable term of HF / activity amount of the sleep time zone. Is preferable.

In the above method, it is preferable that the index includes at least one of the variable term of HF in the wake time zone and the variable term of HF in the sleep time zone.

In the above method, the indicators are the SDNN variable term of the awakening time zone, the SDNN variable term of the sleeping time zone, the CVRR variable term of the awakening time zone, the CVRR variable term of the sleeping time zone, and the RMSSD variable term of the awakening time zone. At least one of the term, the variable term of RMSSD in the sleep time zone, the variable term of NN50 in the awake time zone, the variable term of NN50 in the sleep time zone, the variable term of pNN50 in the awake time zone, and the variable term of pNN50 in the sleep time zone It is preferable to include one. Here, SDNN is the standard deviation of the beat interval, CVRR is the value obtained by dividing SDNN by the average value of the beat interval and multiplied by 100, and RMSD is the square root of the average value of the square of the difference between adjacent beat intervals. NN50 is the total number of beats with an adjacent beat interval difference of more than 50 ms, and pNN50 is the percentage of beats with an adjacent beat interval difference of more than 50 ms.

In the above method, it is preferable to use the RR interval, which is the interval between the R wave and the R wave in the electrocardiographic signal, as the pulsation interval.

In the above method, it is preferable that the index indicates the presence or absence or degree of depression or autonomic imbalance.

Further, the device for creating an index for discriminating the neuropsychiatric state of the present invention has a time calculation unit for calculating the awakening time zone and the sleep time zone, and an index creating unit for creating an index for discriminating the neuropsychiatric state. The time calculation unit calculates the negative acceleration NA from the acceleration TA in the height direction of the subject under the following conditions (1) to (3), and calculates the negative acceleration NA in the first recumbent time zone and the second recumbent time zone. , The awakening time zone and the sleeping time zone are calculated, and the gist is that the index is expressed by an equation including a constant term and a beat interval of the awakening time zone × a variable term of the amount of activity. Here, the amount of activity is the acceleration or angular velocity associated with the movement of the subject.
[conditions:
(1) Calculation of negative acceleration NA When TA ≧ 0 when the subject is standing, NA = (-1) × (TA)
If TA <0 when the subject is standing, NA = TA
(2) Calculation of 1st decubitus time zone and 2nd decubitus time zone 1st decubitus time zone L _m1 : NA ≧ C1 (C1 is a constant) is the first predetermined time T ₁ or more. Position time zone L _m2 : There are two or more first _decubitus time zones L _m1 , and the gap time zone L _{sm where NA <C1 between two adjacent first decubitus time zones L m11} and L m12 is the first. 2 If it is within the predetermined time T _{2, the} time zone in which the two adjacent first recumbent time zones L _m11 and L _m12 and the gap time zone L _sm are totaled (3) Calculation of the awakening time zone and the sleeping time zone Sleeping time zone : The longest time zone of the first recumbent time zone L _m1 and the second recumbent time zone L _m2 during the predetermined measurement unit time Awakening time zone: The time zone excluding the sleep time zone from the predetermined measurement unit time]

The time calculation unit of the above device can objectively and accurately calculate the awakening time zone and the sleeping time zone according to the conditions (1) to (3). In addition, the index creation unit can easily create an index for discriminating the neuropsychiatric state from the beat interval and the amount of activity, which are easily measurable parameters. In addition, it is possible to objectively determine the neuropsychiatric state by using the created index.

The device further comprises a value obtained by definitely integrating the power spectrum obtained by including the step of converting the beat interval into a frequency spectrum from frequencies Lf1 to Lf2, and a value obtained by definitely integrating the power spectrum from frequencies Hf1 to Hf2. It is preferable to have a processing unit for calculating a certain HF, and the index includes a variable term of (LF / HF) × activity amount of the sleep time zone. Here, Hf1> Lf1 and Hf2> Lf2.

In the above apparatus, it is preferable that the index includes the variable term of the total time of the first recumbent time zone.

The method and apparatus of the present invention can objectively and accurately calculate the awakening time zone and the sleeping time zone according to the conditions (1) to (3). In addition, it is possible to easily create an index for discriminating the neuropsychiatric state from the beat interval and the amount of activity, which are easily measurable parameters. In addition, it is possible to objectively determine the neuropsychiatric state by using the created index.

The explanatory diagram of the power spectrum integral which concerns on this invention is shown. The block diagram which shows the structure of the index making apparatus which discriminates the neuropsychiatric state which concerns on Embodiment 1 of this invention is shown. The block diagram which shows the structure of the index making apparatus which discriminates the neuropsychiatric state which concerns on Embodiment 2 of this invention is shown. The block diagram which shows the structure of the index making apparatus which discriminates the neuropsychiatric state which concerns on Embodiment 3 of this invention is shown.

1. 1. Method for Creating an Index for Discriminating Psychoneurological State The method for creating an index for discriminating the neuropsychiatric state of the present invention is the amount of activity expressed by the beat interval of the subject and the acceleration or angular velocity accompanying the movement of the subject. The negative acceleration NA is calculated according to the step of measuring the acceleration TA in the height direction of the subject and the conditions (1) to (3) described later, and the first decubitus time zone and the second decubitus position are calculated. A step of calculating the time zone, the awakening time zone and the sleeping time zone, and a step of creating an index represented by an equation including a constant term and a variable term of the beat interval × activity amount of the awakening time zone. Have. The details of each step will be described.

(A) Measurement step In the above method, the beat interval of the subject, the amount of activity expressed by the acceleration or angular velocity accompanying the movement of the subject, and the acceleration TA in the height direction of the subject are measured. Have steps.

The beat interval refers to the interval between heartbeats or pulses (unit: ms). The heartbeat interval is acquired by reading the interval between R waves from the electrocardiogram or measuring the interval between adjacent heartbeats. The pulse interval is acquired by measuring the interval between adjacent pulses. The beat interval or its pulsation is said to be an index of physical and psychological stress, and reflects the balance of the neuropsychiatric states of the sympathetic and parasympathetic nerves, which are the autonomic nervous system.

As the pulsation interval, it is preferable to use the RR interval (hereinafter, referred to as "RRI"), which is the interval between the R wave and the R wave in the electrocardiographic signal. In RRI, since the peak of the signal appears clearly, erroneous recognition of the peak position is unlikely to occur, so that the accuracy of the beat interval can be improved.

The amount of activity is the acceleration or angular velocity associated with the movement of the subject. Acceleration A is a composite value of the acceleration accompanying the movement of the subject and the gravitational acceleration acting on the subject, and is ^{expressed as a ratio to the gravitational acceleration g (= 9.8 m / s 2} ) (unit: dimensionless). amount). Specifically, as expressed by the following equation (1), the acceleration A is the square of the accelerations x, y, and z in the X-axis, Y-axis, and Z-axis directions, which are the accelerations associated with the movement of the subject. It is the value obtained by subtracting (g / g) = 1 as ^{the gravitational acceleration g (= 9.8 m / s 2} ) acting on the subject from the square root of the sum (where the unit g is the magnitude of the gravitational acceleration). Represent). When the subject is standing and there is no movement, the values of x, y, and z are almost 0, so that the magnitude of the acceleration in the height direction of the subject is almost 1.

_{The angular velocity Ω is the square root of the sum of squares of the angular velocities ω x} , ω _y , and ω _z around the X-axis, Y-axis, and Z-axis of the subject, and the unit is rad / s or 1 / s. The angular velocity Ω is expressed by the following equation (2).

Since the angular velocity detects rotation, it is suitable for detecting, for example, the frequency of turning over of a subject during sleep time. In order to make it easier to detect the posture of the subject, the amount of activity is preferably acceleration.

In the case of psychiatric disorders, the amount of activity during sleep is excessively high, and the amount of activity during wakefulness is excessively low. Therefore, it can be said that the amount of activity also indicates the state of neuropsychiatry.

The acceleration TA in the height direction of the subject is used for calculating the first lying position time zone, the second lying position time zone, the awakening time zone, and the sleeping time zone, which will be described later. The height direction is the direction from the subject's feet to the head. In general, when the accelerometer is adjusted so that the acceleration in the height direction becomes a negative value when standing, the acceleration in the height direction in the lying position tends to be larger than the acceleration in the height direction in the standing or sitting position. is there. When the subject is standing and there is no movement, the magnitude of the acceleration in the height direction of the subject is 1. In the case of mental disorders, the activity in the sleep time zone is excessively high, and the activity in the awake time zone is excessively low. Therefore, it can be said that the acceleration TA in the height direction of the subject indicates the state of the neuropsychiatric nerve.

In the measurement step, it is preferable to measure the pulsation interval, the amount of activity, and the acceleration TA in the height direction for a predetermined measurement time. The predetermined measurement time refers to the total time for measurement. In the index creation step (C) described later, the data of the awakening time zone is used, but since it is necessary to specify the sleep time zone for calculating the awakening time zone, the predetermined measurement time is one or more sleep time zones. It is preferable that the time length is such that one or more awakening time zones can be obtained. Alternatively, the predetermined measurement time may be a length of time during which two or more awakening time zones and two or more sleeping time zones can be obtained. Therefore, the predetermined measurement time is preferably 2 days or more, more preferably 3 days or more, and further preferably 4 days or more. Further, the longer the predetermined measurement time is, the more reliable the data can be acquired. However, in consideration of the burden on the subject, the predetermined measurement time can be set, for example, within 14 days, more preferably within 10 days. .. The sleep time zone is a time zone in which the subject is sleeping, and the awakening time zone is a time zone in which the subject is awake. These time zones will be described in detail later.

(B) Time calculation step In the above method, the negative acceleration NA is calculated under the following conditions (1) to (3), and the first lying position time zone, the second lying position time zone, the awakening time zone, and the sleeping time zone are calculated. Has a step of calculating.

Condition (1) Calculation of negative acceleration NA When TA ≧ 0 when the subject is standing, NA = (-1) × (TA)
If TA <0 when the subject is standing, NA = TA

Negative acceleration NA has a sign of minus and is ^{expressed as a ratio to gravity acceleration g (= 9.8 m / s 2} ) (unit: dimensionless quantity). For the calculation of the negative acceleration, the acceleration TA in the height direction measured in the measurement step (A) is used.

Condition (2) Calculation of 1st decubitus time zone and 2nd decubitus time zone 1st decubitus time zone L _m1 : NA ≧ C1 (C1 is a constant) is the first predetermined time T ₁ or more. Lying time zone L _m2 : There are two or more first lying time zones L _m1 , and the gap time zone L _sm where NA <C1 between two adjacent first lying time zones L _m11 and L _m12 If it is within the second predetermined time T _{2, the} time zone in which the two adjacent first recumbent time zones L _m11 and L _m12 and the gap time zone L _sm are totaled.

The recumbent position indicates a supine position, for example, a supine position, which is a supine position, a lateral decubitus position, which is a lying position, and a prone position, which is a prone position, such as sleep, napping, and nap. Includes time. In the present invention, the recumbent time zone is divided into a first recumbent time zone and a second recumbent time zone for calculation.

The first recumbent hours L _m1 NA ≧ C1 (C1 is a constant) is the time period is a first predetermined time above T _1. A negative acceleration NA of C1 or higher indicates that the subject is in the recumbent position. In the calculation of the first lying time zone L _m1 , a threshold value based _{on the first predetermined time T 1} is set in order to improve the calculation accuracy of the awakening time zone and the sleeping time zone.

The first predetermined time T ₁ can be set, for example, preferably 30 minutes or more, more preferably 45 minutes or more, still more preferably 1 hour or more, but set to 2 hours or less, or 1.5 hours or less. You can also. It can be said that the time zone in which NA ≧ C1 is less than the first predetermined time T ₁ is actually in the lying position, but it can be evaluated as the original sleep from the viewpoint of activity rhythm such as a nap or a nap for a relatively short time. In order to prevent the time that cannot be calculated as the sleep time zone, the time zone in which NA ≧ C1 is _{less than the first predetermined time T 1} is not regarded as the lying time zone.

The second recumbent time zone L _m2 is a gap time zone in which there are two or more first recumbent time zones L _m1 and NA <C1 between two adjacent first recumbent time zones L _m11 and L _m12. When L _sm is within the second predetermined time T ₂ , it is the total time zone of the two adjacent first recumbent time zones L _m11 and L _m12 and the gap time zone L _sm. NA <C1 indicates that the patient is in a posture other than the recumbent position. The reason why the second recumbent time zone is calculated in this way is that when the first recumbent time zone L _m1 is fragmented, the gap between the two first recumbent time zones L _m11 and L _m12. This is because it is calculated as one recumbent time zone (second recumbent time zone L _m2 ) including the time zone L _sm. This makes it easier to calculate the sleep time zone even if the subject frequently takes a posture other than the recumbent position during the sleep time zone.

The second predetermined time T ₂ can be set, for example, preferably 30 minutes or more, more preferably 45 minutes or more, still more preferably 1 hour or more, and may be set to 2 hours or less, or 1.5 hours or less. it can. Further, the first predetermined time T ₁ and the second predetermined time T ₂ may be the same or different.

The value of the constant C1 (unit: dimensionless quantity) is not particularly limited, but is preferably -0.75 or more, more preferably -0.62 or more, and preferably -0.5 or more, for example. More preferred. The constant C1 is allowed to be -0.4 or less, or -0.45 or less.

Condition (3) Calculation of awakening time zone and sleeping time zone Sleeping time zone: The longest time zone of _{the first decubitus time zone L m1} and the second decubitus time zone L _{m2 during a predetermined measurement unit time Awakening time zone} : Time zone excluding sleep time zone from predetermined measurement unit time

The sleep time zone is a time zone in which the subject is sleeping, and refers to the longest time zone of _{the first lying position time zone L m1} and the second lying position time zone L _{m2 during a predetermined measurement unit time.} The awakening time zone refers to the time zone in which the subject is awake, that is, the time zone in which the subject is awake, and is a time zone other than the sleeping time zone. When the subject self-reports, there is room for beliefs and misunderstandings, but in the present invention, since the measured acceleration TA in the height direction of the subject is used, the awakening time zone and the sleeping time zone are used. The calculation can be performed objectively and accurately.

The predetermined measurement unit time is a time length set for estimating the daily sleep time zone. The predetermined measurement unit time is preferably 12 hours or more, more preferably 18 hours or more, and preferably 24 hours or less. That is, it is preferable that the number of awakening time zones and sleeping time zones is one for each day. Since it is common for both day shift workers and night shift workers to be awake around 18:00, the starting point of the predetermined measurement unit time is preferably set from 17:00 to 19:00.

When calculating the awakening time zone and the sleeping time zone under the above condition (3), for example, the acceleration represented by the acceleration TA or the negative acceleration NA in the height direction is subjected to a morphology calculation on the acceleration-time waveform. The later value is preferable. Morphology operations are used in image processing for noise removal. Therefore, if the value after performing the morphology calculation on the acceleration-time waveform is applied to each conditional expression, the obtained acceleration is shorter than the predetermined measurement time (for example, the predetermined measurement time). Values that change within 1/150 hours are removed. Therefore, the entire outline of the acceleration-time waveform is extracted, and it becomes easy to calculate the awakening time zone and the sleep time zone. The morphology calculation may be performed on the acceleration TA in the height direction or on the negative acceleration NA.

Prior to the morphology calculation, the acceleration TA in the height direction may be binarized. In the binarization process, for example, if the negative acceleration NA is equal to or higher than the threshold value C2 (unit: dimensionless quantity), the negative acceleration NA is regarded as 0, and if the negative acceleration NA is less than C2, it is regarded as 1. By the binarization process, the processing time required for the morphology operation can be shortened. The value of the threshold value C2 is not particularly limited, but is preferably -0.75 or more, more preferably -0.62 or more, and further preferably -0.5 or more, for example. The threshold value C2 is allowed to be -0.4 or less, or -0.45 or less. Although the example of performing the morphology calculation on the data of the negative acceleration NA has been described, the negative acceleration NA may be calculated after performing the morphology calculation on the acceleration TA in the height direction.

The morphology operation includes, for example, an expansion operation for thickening a line, a contraction operation for thinning a line, an opening process for performing an expansion operation after a contraction operation, and a closing process for performing a contraction operation after an expansion operation. When calculating the awakening time zone and the sleeping time zone using the acceleration after the morphology calculation, it is preferable that the morphology calculation is at least one of the opening process and the closing process performed in a predetermined time width. Further, it is more preferable to perform both the opening process and the closing process as the morphology operation. By combining the expansion calculation and the contraction calculation, it becomes easy to extract the entire outline of the acceleration-time waveform, so that it becomes easier to calculate the awakening time zone and the sleep time zone.

The number of times the opening process and the closing process are performed is not particularly limited, but it is preferable that the opening process and the closing process are performed once or more, and it is more preferable that the opening process and the closing process are performed twice or more each.

The time width for performing the expansion calculation and the contraction calculation may be appropriately set, but it is preferable to increase the time width during processing each time the number of processing is repeated. By gradually increasing the time width of the opening process and the closing process in this way, it is possible to prevent the data of the acceleration changed in a short time as compared with the predetermined measurement time from being removed.

(C) Index Creation Step The above method has a step of creating an index represented by an equation including a constant term and a variable term of beat interval in awakening time zone × activity amount. Since it is necessary to process a large amount of data, the index is preferably created by a machine such as a computer.

The index is an expression containing a constant term and a variable term. The constant term a ₀ is an arbitrary real number, and its value is not particularly limited and may be 0. The variable term is represented by the product of the nth-order variable x _{i and the coefficient a i} , which is an arbitrary real number. _{The variable term of the pulsation interval of the awakening time zone × the amount of activity means that the variable term includes the variable x i} of the nth order (the pulsation interval of the awakening time zone) × (the amount of activity of the awakening time zone). Means. The degree n of the variable x _i is preferably 1 or more or 2 or more in order to simplify the expression, and 5 or less or 4 or less in order to prevent the expression from becoming complicated.

The method of the present invention can easily create an index for discriminating the neuropsychiatric state from the beat interval and the amount of activity, which are easily measurable parameters. In addition, it is possible to objectively determine the neuropsychiatric state by using the created index.

(C) In the index creation step, the beat interval and the amount of activity in the awakening time zone are preferably the average value of the beat interval in the awakening time zone × the amount of activity, and the measured beats in all the awakening time zones. It is more preferable that the average value of the interval × the amount of activity. By setting the variable term in this way, it becomes easy to create an index. Similarly, in other variable terms described later, it is preferable to use the average value of the variables in the awakening time zone or the sleeping time zone.

(C) Prior to the index creation step, it is preferable to calculate at least one of the value of the constant term and the value of the coefficient of the variable term included in the formula representing the index by multivariate analysis. This makes it possible to quantitatively estimate the relationship between the neuropsychiatric state as the objective variable and the beat interval x activity amount in the awakening time zone as the explanatory variable. In calculating the coefficients of the constant term and the variable term, it is preferable to use data on the presence / absence, type or severity of mental disorders of a plurality of subjects. The number of subjects required for multivariate analysis is preferably the number of explanatory variables (variable terms) × 10 or more, more preferably the number of explanatory variables × 50 or more, and further preferably the number of explanatory variables × 100 or more. Is.

The neuropsychiatric state, which is the objective variable, is preferably quantified according to the presence, type, or severity of mental disorders. Examples of the quantification of the neuropsychiatric state include a method of quantifying the presence or absence of a psychiatric disorder and a method of classifying the psychiatric disorder according to the type or severity of the psychiatric disorder based on known criteria. Known criteria include DSM (Diagnostic and Statistical Manual of Mental Disorders), which is a criterion for mental disorders, HSCL, which is a criterion for stress, and CES-, which is a criterion for depression. D, HAMD, YMRS, PANSS (Positive and Negative Syndrome Scale), which is a criterion for schizophrenia, Toho Medical Index, which is a criterion for autonomic imbalance, and the like can be used.

For multivariate analysis, linear regression analysis such as simple regression analysis and multiple regression analysis, binary logistic regression analysis, polynomial logistic regression analysis, cumulative logistic regression analysis and other logistic regression analysis can be used. Let Y be an index indicating the neuropsychiatric state. In the case of linear regression analysis, the index Y is represented by Y = a ₀ + a ₁ x ₁ + a ₂ x ₂ + ... + a _i x _i . It is preferable to use binary logistic regression analysis to create an index for determining the presence or absence of mental disorders. In that case, the index Y is preferably represented by Y = 1 / {1 + exp (−b)}, b = a ₀ + a ₁ x ₁ + a ₂ x ₂ + ... + a _i x _i . Note that a ₁ , a ₂ , ... a _i are coefficients of explanatory variables and are arbitrary real numbers, but a ₁ ≠ 0. x ₁ , x ₂ , ... x _i is a variable, and i is an integer of 1 or more. The index created by binary logistic regression analysis represents the neuropsychiatric state as a binary of 0 and 1, or a number between 0 and 1.

In order to improve the accuracy of discriminating the neuropsychiatric state using the index, a variable term other than the variable term of the pulsation interval and the amount of activity in the awakening time zone may be included. In the formula of the index, if the coefficient of each variable term is an arbitrary real number, its magnitude relation is not particularly specified.

The above method further uses LF, which is a value obtained by definitely integrating the power spectrum obtained by including the step of converting the beat interval into a frequency spectrum from frequencies Lf1 to Lf2, and a value obtained by definitely integrating the power spectrum from frequencies Hf1 to Hf2. It may have a step of calculating a certain HF.

LF is a value obtained by definitely integrating the power spectrum obtained by including the step of frequency spectrum conversion of the beat interval, which is the time signal f, from frequencies Lf1 to Lf2, and HF is the value obtained by definitely integrating the power spectrum from frequencies Hf1 to Hf2. It is an integrated value, and Hf1> Lf1 and Hf2> Lf2. ^{For example, LF is a constant integration of the power spectrum F 2} (first power spectrum) obtained by squaring the beat interval, which is the time signal f, converted into a frequency spectrum (frequency spectrum F) from frequencies Lf1 to Lf2. The HF can be a value obtained by definitely integrating the power spectrum F ² (first power spectrum) from the frequencies Hf1 (> Lf1) to Hf2 (> Lf2). The unit of LF and HF calculated using the first power spectrum F ^{2 is ms 2} . As a method of frequency spectrum transformation, for example, a fast Fourier transform (FFT), a wavelet analysis, a maximum entropy method, or the like can be used. In this specification, the case where FFT is used will be described as an example, but of course, other methods can also be used.

_{In the present specification, the discrete Fourier transform G of the beat interval RRI k obtained} by spline-interlacing the beat interval and re-sampling at the sampling interval Δt is expressed by the following equation (3), and the power spectrum F ² (first power). The spectrum) (unit: ms ² / Hz) is expressed by the following equation (4). Here, k represents a time series, N represents the number of data, S is an arbitrary scale, and generally S = 1 in the power spectrum.

On the other hand, as the values of LF and HF, the method of the present invention is also obtained by definitely integrating the power spectrum F (second power spectrum) (unit: ms) obtained from the value obtained by converting the beat interval into a frequency spectrum in a predetermined interval. include. As described above, if the value obtained by converting the beat interval into the frequency spectrum is used as the power spectrum, the values of LF and HF can be calculated more easily. The units of LF and HF calculated using the second power spectrum F are dimensionless quantities. The power spectrum F (second power spectrum) is represented by the following equation (5).

A detailed calculation method of LF and HF will be described with reference to FIG. FIG. 1 is an explanatory diagram of a power spectrum integral according to the present invention. The vertical axis of FIG. 1 is the power spectral density (unit: ms ² / Hz), and the horizontal axis is the frequency (unit: Hz). LF is a value obtained by definitely integrating the power spectrum (for example, the first power spectrum F ² ) from, for example, 0.04 Hz (Lf1) to 0.15 Hz (Lf2), and is the portion hatched by diagonal lines in FIG. The area. Here, Lf1 <Lf2. On the other hand, HF is a value obtained by definitely integrating the power spectrum (for example, the first power spectrum F ² ) from, for example, 0.15 Hz (Hf1) to 0.4 Hz (Hf2), and is hatched by vertical lines in FIG. It is the area of the part. Here, Hf1 <Hf2. In FIG. 1, the integration range is set so that both Lf2 and Hf1 are equal to 0.15 Hz, but if the relationship of Lf1 <Hf1 and Lf2 <Hf2 is satisfied, Lf2 and Hf1 have the same value. May be different. Here, the method of power spectrum integration ^{has been described using the first power spectrum F 2} , but the definite integral by the second power spectrum F can also be performed in the same manner.

The power spectrum obtained by frequency spectrum conversion is divided into LF, which is a component derived from fluctuations in blood pressure and is also called a Mayer-Wave-related component, and HF, which is a component derived from respiration. The blood pressure fluctuation component LF is a power spectrum around 0.1 Hz and is related to both sympathetic nerve activity and parasympathetic nerve activity. On the other hand, the respiratory component HF has a power spectrum around 0.3 Hz and is considered to be related to parasympathetic nerve activity. From the above, the integration range of LF indicating sympathetic nerve activity and parasympathetic nerve activity preferably includes at least 0.1 Hz and Lf1 <0.1 <Lf2. Further, Lf1 is more preferably 0.03 Hz or higher, and further preferably 0.04 Hz or higher. Further, Lf1 is preferably 0.05 Hz or less, and more preferably 0.045 Hz or less. Lf2 is preferably 0.13 Hz or higher, more preferably 0.14 Hz or higher, preferably 0.16 Hz or lower, and more preferably 0.15 Hz or lower. The integration range of HF indicating parasympathetic nerve activity preferably includes at least 0.3 Hz and Hf1 <0.3 <Hf2. Hf1 is more preferably 0.14 Hz or higher, further preferably 0.15 Hz or higher, and may be 0.17 Hz or lower, or 0.16 Hz or lower. Hf2 is preferably 0.38 Hz or higher, more preferably 0.39 Hz or higher, more preferably 0.41 Hz or lower, and even more preferably 0.4 Hz or lower.

(C) Before the index creation step, it is determined whether at least one of the calculated beat interval, activity amount, LF, and HF contains an abnormal value, and the value is determined to be an abnormal value. It is preferable to exclude the calculated value from the target of index creation. As a result, it is possible to prevent the abnormal value from affecting the index and the discrimination result using the index.

When the above method has a calculation step of LF and HF, it is preferable that the index includes a variable term of (LF / HF) × activity amount of the sleep time zone. Since LF indicates both sympathetic and parasympathetic activity, LF / HF represents the predominance of sympathetic activity over parasympathetic activity. The psychiatric state is a state of balance of the autonomic nervous system, and LF and LF / HF can be said to indicate the state of the psychiatric nerve. By setting the variable term as described above, an index with high discrimination accuracy can be obtained.

In the above method, it is preferable that the index includes a variable term of the total time of the first recumbent time zone. The first recumbent time zone indicates the time zone in which the patient is in the recumbent position including the sleeping time zone, but if there is an irregular life or symptoms such as chronic fatigue, depression, or insomnia, The total time in the first recumbent time zone tends to be long. Therefore, it can be said that the total time of the first recumbent time zone also indicates the state of the neuropsychiatric nerve. Therefore, by setting the variable term as described above, an index with high discrimination accuracy can be obtained. The total time of the first recumbent time zone is preferably the total time of the first recumbent time zone per predetermined measurement unit time.

It is preferable that the index includes a variable term of the beat interval of the sleep time zone. As described above, since the beat interval in the sleep time zone indicates the state of the psychiatric nerve, by setting the variable term in this way, an index with high discrimination accuracy can be obtained. Among them, the indicators are the constant term, the beat interval of the awakening time zone x the variable term of the activity amount, the (LF / HF) of the sleeping time zone x the variable term of the activity amount, and the variable of the total time of the first lying time zone. It is preferable that it is expressed by an equation consisting of a term and a variable term of the beat interval of the sleep time zone. By setting the index in this way, the accuracy of discriminating the neuropsychiatric state can be further improved.

It is preferable that the index includes the variable term of (LF / HF) / activity amount of the awakening time zone. Since both (LF / HF) and the amount of activity indicate the state of the psychiatric nerve, by setting the variable term in this way, an index with high discrimination accuracy can be obtained.

It is preferable that the index includes a variable term of beat interval / activity during sleep time. Since both the beat interval and the amount of activity indicate the state of the psychiatric nerve, by setting the variable term in this way, an index with high discrimination accuracy can be obtained. Among them, the indicators are the constant term, the beat interval of the awakening time zone x the variable term of the activity amount, the (LF / HF) of the sleeping time zone x the variable term of the activity amount, and the total time of the first decubitus time zone. It is expressed by an expression consisting of a variable term, a variable term of the beat interval of the sleep time zone, a variable term of (LF / HF) / activity amount of the awakening time zone, and a variable term of the beat interval / activity amount of the sleep time zone. Is preferable. By setting the index in this way, the discrimination accuracy can be further improved.

The method may further comprise the step of time zone is NA ≧ C1 calculates the third recumbent hours is within a third predetermined time T _3. In that case, it is preferable that the third predetermined time T ₃ is less than the first predetermined time T ₁ and the index includes the variable term of the total time of the third recumbent time zone. The third recumbent time zone indicates a time zone in which the patient is in the recumbent position other than the first recumbent time zone, but the time (third predetermined time T ₃ ) is less than the first predetermined time T _1. It can be said that it is a relatively short time. However, if the patient is in a lying position during the awakening period even for a short period of time, for example, the sympathetic nerve and the parasympathetic nerve may be out of balance. It can be said that it shows. Therefore, by setting the variable term as described above, an index with high discrimination accuracy can be obtained. The total time of the third recumbent time zone is preferably the total time of the third recumbent time zone per predetermined measurement unit time.

The third predetermined time T ₃ can be set, for example, preferably 30 minutes or more, more preferably 45 minutes or more, still more preferably 1 hour or more, and may be set to 2 hours or less, or 1.5 hours or less. it can.

The indicators are the variable term of the beat interval of the awakening time zone, the variable term of the LF / HF of the awakening time zone, the variable term of the activity amount of the awakening time zone, the variable term of the LF / HF of the sleeping time zone, and the variable term of the sleeping time zone. It is preferable to include at least one of an activity variable term, a sleep time zone time variable term, an awakening time zone HF × activity variable term, and a sleep time zone HF / activity variable term. As described above, the beat interval, LF / HF, and amount of activity all indicate the state of the neuropsychiatric nerve. In addition, when the sleep time is excessively short or long, the balance between the sympathetic nerve and the parasympathetic nerve is lost, and it can be said that the sleep time also indicates the state of the psychiatric nerve. Therefore, by setting the variable term as described above, an index with high discrimination accuracy can be obtained.

The indicators are the constant term, the beat interval in the awakening time zone x the variable term for the amount of activity, the variable term for the sleeping time zone (LF / HF) x the amount of activity, and the variable for the total time in the first recumbent time zone. The variable term of the beat interval of the sleep time zone, the variable term of (LF / HF) / activity amount of the awakening time zone, the variable term of the beat interval / activity amount of the sleep time zone, and the third lie. The variable term of the total time of the rank time zone, the variable term of the beat interval of the awakening time zone, the variable term of LF / HF of the awakening time zone, the variable term of the activity amount of the awakening time zone, and the variable term of the sleep time zone. It is preferably expressed by an equation consisting of a variable term of LF / HF, a variable term of the amount of activity in the sleep time zone, and a variable term of the time in the sleep time zone. By setting the index in this way, the discrimination accuracy can be further improved.

In addition, the indicators are the constant term, the beat interval in the awakening time zone x the variable term for the amount of activity, the variable term for the sleeping time zone (LF / HF) x the amount of activity, and the total time in the first recumbent time zone. Variable term of (LF / HF) / activity amount in the awakening time zone, beat interval / activity variable term in the sleeping time zone, and total time variable term in the third recumbent time zone. , The variable term of the time of the sleeping time zone, the variable term of HF × the amount of activity in the awakening time zone, and the variable term of HF / the amount of activity in the sleeping time zone may be expressed by an equation. By setting the index in this way, the accuracy of discriminating the neuropsychiatric state can be further improved.

The index may include at least one of the variable term of HF in the wake time zone and the variable term of HF in the sleep time zone. Since the HF in the wake time zone or the sleep time zone indicates the state of the psychiatric nerve, it is possible to create an index with higher discrimination accuracy. Among them, the indicators are the constant term, the beat interval of the awakening time zone x the variable term of the activity amount, the (LF / HF) of the sleeping time zone x the variable term of the activity amount, and the total time of the first decubitus time zone. , The variable term of the beat interval of the sleep time zone, the variable term of (LF / HF) / activity amount of the awakening time zone, the variable term of the beat interval / activity amount of the sleep time zone, and the first 3 Variable term of total time in lying time zone, variable term of beat interval in awakening time zone, variable term of LF / HF in awakening time zone, variable term of activity amount in awakening time zone, sleep time LF / HF variable term of the band, activity variable term of the sleep time zone, time variable term of the sleep time zone, HF x activity variable term of the awakening time zone, HF / of the sleep time zone It is preferably expressed by an equation consisting of a variable term of activity amount, a variable term of HF in the awakening time zone, and a variable term of HF in the sleeping time zone. By setting the index in this way, the discrimination accuracy can be further improved.

The indicators are SDNN variable term of awakening time zone, SDNN variable term of sleeping time zone, CVRR variable term of awakening time zone, CVRR variable term of sleeping time zone, RMSSD variable term of awakening time zone, sleeping time. Includes at least one of the RMSDV variable term of the band, the NN50 variable term of the awakening time zone, the NN50 variable term of the sleeping time zone, the pNN50 variable term of the awakening time zone, and the pNN50 variable term of the sleeping time zone. Is preferable. In addition, the indicators are the SDNN variable term of the awakening time zone, the SDNN variable term of the sleeping time zone, the CVRR variable term of the awakening time zone, the CVRR variable term of the sleeping time zone, the RMSSD variable term of the awakening time zone, It may include all of the RMSDV variable term of the sleep time zone, the NN50 variable term of the awakening time zone, the NN50 variable term of the sleep time zone, the pNN50 variable term of the awakening time zone, and the pNN50 variable term of the sleep time zone. More preferred. SDNN (Standard Deviation of the Normal to Normal Interval) is the standard deviation of the beat interval. CVRR (Coefficient of Variation of R-R intervals) is a value obtained by dividing the SDNN by the average value of the beat interval (preferably the heartbeat interval) and multiplying by 100. RMSDD (Root Mean Square of the Successive Differences) is the square root of the mean square of the difference between adjacent beat intervals. NN50 is the total number of beats with an adjacent beat interval difference of more than 50 ms, and pNN50 is the percentage of beats with an adjacent beat interval difference of more than 50 ms. Since SDNN, CVRR, RMSSD, NN50, and pNN50 all indicate the state of the psychiatric nerve, it is possible to create an index with even higher discrimination accuracy.

The above method may further include a step of calculating VLF, which is a value obtained by definitely integrating the power spectrum obtained including the step of converting the beat interval into a frequency spectrum from frequencies Lf3 to Lf4. In that case, the index may include at least one of the VLF variable term of the awakening time zone and the VLF variable term of the sleeping time zone. Here, Lf3 <Lf4. Since VLF is considered to be related to sympathetic nerve activity, the neuropsychiatric state can also be determined by setting the index in this way. Lf3 is preferably 0.0025 Hz or higher, more preferably 0.003 Hz or higher, and even more preferably 0.0033 Hz or higher. Lf3 is preferably 0.005 Hz or less, and more preferably 0.004 Hz or less. Lf4 is preferably 0.025 Hz or higher, more preferably 0.03 Hz or higher, preferably 0.06 Hz or lower, more preferably 0.05 Hz or lower, and 0.04 Hz. The following is more preferable. Further, Lf4 may be less than Lf1 or more than Lf1, but it is preferable that Lf4 = Lf1.

The above method may further include a step of calculating ULF, which is a value obtained by definitely integrating the power spectrum obtained including the step of converting the beat interval into a frequency spectrum from frequencies Lf5 to Lf6. In that case, the index may include at least one of the variable term of ULF in the wake time zone and the variable term of ULF in the sleep time zone. Here, Lf5 <Lf6. Since ULF is considered to be related to sympathetic nerve activity, the neuropsychiatric state can also be determined by setting the index in this way. Lf5 is preferably 0.0015 Hz or less, more preferably 0.001 Hz or less, and may be 0 Hz or more, but more preferably 0 Hz. Lf6 is preferably 0.0025 Hz or higher, more preferably 0.003 Hz or higher, further preferably 0.0033 Hz or higher, and preferably 0.0045 Hz or lower, preferably 0.004 Hz. More preferably: Further, Lf6 may be less than Lf3 or more than Lf3, but it is preferable that Lf3 = Lf6.

For the calculation of VLF and ULF, for example, the first power spectrum F ² may be used or the second power spectrum F may be used as the power spectrum.

Indicators can indicate the presence or degree of psychiatric disorders such as depression, depression (major depressive disorder), bipolar disorder, schizophrenia, panic disorder, obsessive-compulsive disorder, autonomic imbalance, and sleep disorders. preferable. By using the created index, it is possible to determine the presence or absence and degree of specific mental disorders.

The present invention further includes a method for discriminating a neuropsychiatric state, which comprises a step of discriminating the neuropsychiatric state using the created index. In that case, the index standard (for example, if the index is a numerical value X or more, it is determined to be depression, etc., and the index ranking standard) is a conventionally known judgment standard (for example, in the case of depression, CES-D). Etc.) are preferably the same or correlated. As the index standard, the above-mentioned judgment standard can be used as a preferable objective variable. As a result, when the neuropsychiatric state is quantified using an index, the meaning of the number becomes clear. A specific example of a method for determining a neuropsychiatric state will be described. Prepare the index created by the above method. Here, an example is shown in which the index is expressed by an expression including a constant term and a beat interval in the awakening time zone × a variable term of the amount of activity, but the variable term “beat interval in the awakening time zone × activity” is shown. As the value of "amount", for example, the average value of the beat interval x activity amount of the subject in all the measured awakening time zones is substituted. By comparing the obtained index with the standard, the neuropsychiatric state can be determined. It is preferable that the determination of the psychiatric state using the index is performed by a machine such as a computer.

In the step of determining the neuropsychiatric state, a threshold value may be set for the index to determine the neuropsychiatric state. If the calculated index is above the threshold, you have (or may have) a mental illness, and if it is below the threshold, you are not (or may have) a mental illness. It is unlikely that this is the case). The value of the threshold value is not particularly limited, but for example, in the case of an index obtained by binary logistic regression analysis, the threshold value may be set to 0.4 or more, 0.5 or more, or 0.6 or less. For example, the threshold value can be set to 0.5, and less than 0.5 can be determined as a healthy state, and 0.5 or more can be determined as a depressed state.

According to the index creation method of the present invention, it is possible to create and discriminate an index that can be objectively discriminated while reducing the burden on the subject. Such indicators are useful for checking and screening for the presence, type or severity of mental illness.

2. Device for creating an index for discriminating the neuropsychiatric state The device for creating an index for discriminating the neuropsychiatric state of the present invention creates a time calculation unit for calculating the awakening time zone and the sleeping time zone, and an index for discriminating the neuropsychiatric state. It has an index creation unit. Hereinafter, the "device for creating an index for discriminating the neuropsychiatric state" may be simply referred to as a "device". Examples of the device include a personal computer capable of transmitting and receiving various data and performing various arithmetic processes, a computer such as a microcomputer, a tablet terminal, and a smartphone.

(Embodiment 1)
FIG. 2 is a block diagram showing a configuration of the device 10 (10A) according to the first embodiment of the present invention. As shown in FIG. 2, the device 10A is provided with a receiving unit 11 for receiving data transmitted from the sensor 50 capable of measuring the pulsation interval, the amount of activity, and the acceleration TA in the height direction of the subject. May be good. Here, the amount of activity is the acceleration or angular velocity associated with the movement of the subject.

The sensor 50 includes a measuring unit 51 that detects the pulsation interval, the amount of activity, and the acceleration TA in the height direction. The beat interval refers to a heartbeat or pulse interval, and it is preferable to use the RR interval (RRI) as the beat interval. In this embodiment, an example in which RRI is measured as the beat interval and acceleration is measured as the amount of activity is shown.

When the heartbeat interval is used as the beat interval, it is preferable that the sensor 50 is small and lightweight, and the electrode on the back surface of the main body is attached to the skin of the chest of the subject together with the main body. Further, the sensor 50 does not need to have the electrode and the main body integrated, and clothes, underwear, a belt or the like having an electrode made of conductive fibers or a conductive sheet / film may be used as the electrode. The measuring unit 51 of the sensor 50 measures the electrocardiographic signal with the electrodes in close contact with the chest of the subject, calculates the RRI based on the electrocardiographic signal, and transmits it to the device 10A. Electrodes can be placed on the abdomen, back, and lumbar region as well as on the chest. The RRI may be calculated by the receiving unit 11 or the time calculating unit 12.

The measuring unit 51 of the sensor 50 may measure the pulse wave. The pulse wave can be obtained by irradiating a person's fingertip, earlobe, or the like with near-infrared rays having a wavelength of 700 nm to 1200 nm, and measuring the amount of near-infrared rays reflected in contact or non-contact. The sensor that measures the pulse wave has an advantage that it is easy to attach to the body, and in particular, the non-contact type that measures the pulse wave has a possibility of becoming widespread because it eliminates the trouble of attaching the sensor to the body.

The measurement unit 51 of the sensor 50 measures the amount of activity represented by the acceleration or angular velocity accompanying the movement of the subject and transmits it to the time calculation unit 12. Further, the measurement unit 51 of the sensor 50 measures the acceleration TA in the height direction and transmits it to the time calculation unit 12. The amount of activity is preferably acceleration. As a result, it is not necessary to measure the amount of activity and the acceleration TA in the height direction separately. The type of the sensor 50 for measuring the acceleration is not particularly limited, and for example, a piezoresistive type acceleration sensor, a piezoelectric type acceleration sensor, a capacitance type acceleration sensor, or the like can be used. Since the piezoresistive acceleration sensor uses a semiconductor, it is small and easy to mass-produce. Piezoelectric accelerometers can easily detect relatively high accelerations. Capacitive accelerometers are more sensitive than piezoresistive accelerometers, have a wider range of detectable accelerations, and are less temperature dependent. The type of the sensor for measuring the angular velocity is not particularly limited, and for example, a rotary type, a vibration type, a gas type, an optical fiber type, and a ring laser type angular velocity sensor can be used.

The receiving unit 11 may be provided in the time calculation unit 12 or the index creating unit 13. As a method of transmitting / receiving data from the sensor 50 to the device 10A, wireless communication may be used or wired communication may be used. In particular, when data is transmitted / received by wireless communication, it is preferable to reduce the frequency of transmission / reception by transmitting a plurality of, for example, three RRIs at once in order to improve the life of the built-in battery.

The time calculation unit 12 receives the data of the pulsation interval, the amount of activity, and the acceleration TA in the height direction of the subject transmitted from the sensor 50 via the reception unit 11. The time calculation unit 12 calculates the negative acceleration NA from the acceleration TA in the height direction of the subject under the following conditions (1) to (3), and calculates the first decubitus time zone, the second decubitus time zone, and awakening. Calculate the time zone and sleep time zone.
[conditions:
(1) Calculation of negative acceleration NA When TA ≧ 0 when the subject is standing, NA = (-1) × (TA)
If TA <0 when the subject is standing, NA = TA
(2) Calculation of 1st decubitus time zone and 2nd decubitus time zone 1st decubitus time zone L _m1 : NA ≧ C1 (C1 is a constant) is the first predetermined time T ₁ or more. Position time zone L _{m2 : The gap time zone L sm in} which there are two or more of the first recumbent time zones L _m1 and NA <C1 between two adjacent first recumbent time zones L _m11 and L _m12. If it is within the second predetermined time T _{2, the time zone obtained by summing the} two adjacent first recumbent time zones L _m11 and L m12 and the gap time zone L _sm (3) Calculation of the awakening time zone and the sleeping time zone Sleep time zone: The longest time zone of the first recumbent time zone L _m1 and the second recumbent time zone L _m2 during the predetermined measurement unit time Awakening time zone: The sleep time zone from the predetermined measurement unit time Excluded time zone]
The details of the calculation method of the first lying time zone, the second lying time zone, the awakening time zone, and the sleeping time zone are as described in "1. Method of creating an index for discriminating the neuropsychiatric state".

It is preferable that the time calculation unit 12 classifies the data transmitted from the sensor 50 into the data of the awakening time zone and the data of the sleeping time zone. The time calculation unit 12 of the device 10A calculates the beat interval in the awakening time zone and the value of the activity amount in the awakening time zone.

The time calculation unit 12 may calculate the total time of the first recumbent time zone. As a result, the index creation unit 13 can create an index including the variable term of the total time in the first lying time zone.

Time calculating unit 12 further, NA ≧ C1 at a time zone may calculate the third recumbent hours is within a third predetermined time T _3. As a result, the index creation unit 13 can create an index including the variable term of the total time in the third recumbent time zone.

Although not shown, the time calculation unit 12 may have a morphology calculation unit. The morphology calculation unit performs the morphology calculation in order to remove the noise of the negative acceleration-time waveform. Further, the time calculation unit 12 may have a binarization processing unit that binarizes the magnitude of the negative acceleration value with a predetermined value as a threshold value before the processing by the morphology calculation unit. The morphology operation and the binarization process can be performed by the method described above.

The index creation unit 13 creates an index for discriminating the neuropsychiatric state by using the data of the awakening time zone and the sleep time zone obtained by the time calculation unit 12. The index is expressed by an expression including a constant term, a pulsation interval and activity amount, and a variable term of pulsation interval × activity amount in the awakening time zone. In the index creation unit 13, it is preferable that the coefficients of the constant term and the variable term included in the index are created by multivariate analysis (more preferably regression analysis). The index creation unit 13 may have an average calculation unit that calculates an average value of data in each time zone. Although not shown, the average calculation unit can calculate, for example, the average value of the beat interval in the awakening time zone and the average value of the activity amount in the awakening time zone.

The index created by the index creating unit 13 may include a variable term of the total time of the first lying time zone, or may include a variable term of the total time of the third lying time zone. Not limited to these indexes, the indexes created by the index creation unit 13 are the variable term of the beat interval of the sleep time zone, the variable term of the beat interval / activity amount of the sleep time zone, and the beat of the awakening time zone. It may include at least one of a variable term of the dynamic interval, a variable term of the activity amount of the awakening time zone, a variable term of the activity amount of the sleeping time zone, and a variable term of the time of the sleeping time zone.

Although not shown, the device 10A may have a function of removing outliers in data. That is, the device 10A may include an outlier detection unit and an outlier removal unit. The outlier detection unit and the outlier removal unit can be preferably provided in the front stage of the time calculation unit 12 or the front stage of the index creation unit 13. As a result, the data from which the abnormal values have been removed can be transmitted to the index creation unit 13, so that an index with high accuracy can be created.

Further, the computer 1 may be provided with a discriminating unit 20 for discriminating the neuropsychiatric state based on the index created by the index creating unit 13. By comparing the created index with the standard, the discriminating unit 20 can objectively and mechanically discriminate the presence or absence and degree of mental disorder.

Although not shown, the computer 1 may be provided with a notification unit that notifies the subject or the like of the determination result of the neuropsychiatric state from the determination unit 20. The notification method is not particularly limited to voice, still image, video and the like. It is also possible to change the content of the notification according to the level of expertise of the person to be notified, such as a specialist such as a doctor or counselor, the subject or his / her family. The discrimination result may be transmitted to a notification device other than the computer 1 and the result may be notified to the subject or the like. Examples of the notification device include an external monitor, a mobile phone, a smartphone, a tablet terminal, a speaker, an earphone, and the like.

(Embodiment 2)
The second embodiment is a configuration example of an apparatus having a processing unit for calculating LF and HF. FIG. 3 shows a block diagram showing the configuration of the device 10 (10B) according to the second embodiment of the present invention. The same components as those of the device 10A of the first embodiment are designated by the same numbers, and the description thereof will be omitted. The device 10B includes a time calculation unit 12, an index creation unit 13, and a processing unit 14 (hereinafter, referred to as “first processing unit 14A”).

The first processing unit 14A calculates the data related to the frequency domain among the data required for creating the variable term. Specifically, the first processing unit 14A definitely integrates the power spectrum obtained including the step of converting the beat interval into the frequency spectrum from frequencies Lf1 to Lf2, and the power spectrum is obtained from frequencies Hf1 to Hf2. It is preferable to calculate HF, which is a definite integral value up to. In addition, it is preferable that the first processing unit 14A also calculates LF / HF.

Although not shown, the first processing unit 14A may include a frequency spectrum conversion unit and a power spectrum integration calculation unit. The frequency spectrum conversion unit converts the RRI, which is a time signal transmitted from the reception unit 11, into a frequency spectrum by using a frequency spectrum conversion method such as FFT. Next, the power spectrum integration calculation unit calculates the power spectrum from the spectrum obtained by the frequency spectrum conversion unit and performs integration in a predetermined frequency range to obtain LF and HF. The specific calculation method of LF and HF is as described above in "1. Method of creating an index for discriminating a neuropsychiatric state". As the power spectrum, for example, the first power spectrum F ² may be used, or the second power spectrum F may be used.

In the time calculation unit 12 of the device 10B, the sleep time zone of the LF and HF calculated by the first processing unit 14A and the amount of activity transmitted from the receiving unit 11 based on the above conditions (1) to (3). Extract the data that meets the calculation conditions of. As a result, the value of the sleep time zone (LF / HF) and the amount of activity in the sleep time zone are calculated.

The first processing unit 14A includes a VLF which is a value obtained by constantly integrating the power spectrum obtained by including the step of converting the beat interval into a frequency spectrum from frequencies Lf3 to Lf4, and a step of converting the beat interval into a frequency spectrum. At least one of ULF, which is a value obtained by constantly integrating the obtained power spectrum from frequencies Lf5 to Lf6, may be calculated. By calculating the VLF in the first processing unit 14A, the index creating unit 13 can create an index including the variable term of the VLF. Further, by calculating the ULF in the first processing unit 14A, the index creating unit 13 can create an index including the variable term of the ULF.

In the index creation unit 13, using the data obtained by the time calculation unit 12 and the first processing unit 14A, the pulsation interval of the awakening time zone × the variable term of the activity amount and the (LF / HF) × of the sleep time zone Create an index represented by an equation that includes the variable term of activity.

In the second embodiment, the index further includes the variable term of sleep time zone (LF / HF) × activity amount. However, when the device 10B is provided with the first processing unit 14A, the index is created. Is the variable term of (LF / HF) / activity amount in the awakening time zone, the variable term of LF / HF in the awakening time zone, the variable term of LF / HF in the sleeping time zone, and the variable of HF × activity amount in the awakening time zone. HF / activity variable term in sleep time zone, HF variable term in wake time zone, HF variable term in sleep time zone, VLF variable term in wake time zone, VLF variable term in sleep time zone , At least one of the variable term of ULF in the wake time zone and the variable term of ULF in the sleep time zone may be included.

(Embodiment 3)
Embodiment 3 is a configuration example of an apparatus having a processing unit that further processes beat interval data into data such as SDNN. FIG. 4 shows a block diagram showing the configuration of the device 10 (10C) according to the third embodiment of the present invention. The same components as those of the devices 10A and 10B of the first and second embodiments are designated by the same numbers, and the description thereof will be omitted. The device 10C includes a time calculation unit 12, an index creation unit 13, and a processing unit 14 (hereinafter, referred to as “second processing unit 14B”).

It is preferable that the second processing unit 14B calculates at least one of SDNN, CVRR, RMSSD, NN50, and pNN50 using the beat interval data. This makes it possible to create an index containing at least one of SDNN, CVRR, RMSSD, NN50, and pNN50 as a variable term.

Although not shown, the apparatus 10C may be provided with both the first processing unit 14A and the second processing unit 14B. This makes it possible to create an index with high discrimination accuracy.

(Verification)
According to the method of the present invention, the following objective variables and explanatory variables were set for the data of 31 subjects including depressed patients and healthy subjects, and a regression equation was created by binary logistics regression analysis. SPSS Statistics (manufactured by IBM) was used as the statistical analysis software.
Objective variable: Binary data with depression as 1 and healthy as 0 Explanatory variables: Awakening time zone pulsation interval x activity amount, sleep time zone (LF / HF) x activity amount, 1st decubitus time zone Total time, beat interval of sleep time zone The regression equations obtained are Y = 1 / {1 + exp (−b)}, b = a ₀ + a ₁ × (beat interval of awakening time zone × activity amount). ) + A ₂ x {(LF / HF) x activity in sleep time zone} + a ₃ x (total time in first recumbent time zone) + a ₄ x (beat interval in sleep time zone), a ₀ = 6. 94, a ₁ = −0.034, a ₂ = −0.004, a ₃ = 0.002, a ₄ = 11.412.

The index created for 31 subjects was applied. When the score calculated by the index was 0.5 or more, it was determined to be depressed, and when it was less than 0.5, it was determined to be healthy. The detection rate of subjects judged to be depressed based on the CES-D interview results was 85%, and the detection rate of subjects determined to be healthy based on the CES-D interview results was 89%. It was found that the neuropsychiatric state can be discriminated with high accuracy by using the index created in the present invention.

1: Computers 10, 10A, 10B, 10C: Index creation device for discriminating the neuropsychiatric state 11: Reception unit 12: Time calculation unit 13: Index creation unit 14: Processing unit 14A: First processing unit 14B: Second processing Unit 20: Discrimination unit 50: Sensor 51: Measurement unit

Claims (12)

Steps to measure the pulsatile interval of the subject, the amount of activity expressed by the acceleration or angular velocity associated with the movement of the subject, and the acceleration TA in the height direction of the subject.
Negative acceleration NA is calculated according to the following conditions (1) to (3), and the first lying time zone, the second lying time zone, the third lying time zone, the awakening time zone, and the sleeping time zone are calculated. Steps and
LF, which is a definite integral value of the power spectrum obtained by including the step of converting the beat interval into a frequency spectrum from frequencies Lf1 to Lf2, and HF, which is a definite integral value of the power spectrum from frequencies Hf1 to Hf2, are calculated. Steps to do and
A constant term, a beat interval of the awakening time zone × a variable term of the activity amount, a variable term of (LF / HF) × the activity amount of the sleeping time zone, and a variable of the total time of the third lying time zone. Has a term and a step to create an index represented by an expression containing
In the step of creating the index, the constant term, the coefficient of the beat interval of the awakening time zone × the variable term of the activity amount, and the sleeping time zone are obtained by multivariate analysis using the data of a plurality of subjects. (LF / HF) × A method for creating an index for discriminating a neuropsychiatric state, which comprises calculating a coefficient of a variable term of activity amount. Here, Hf1> Lf1 and Hf2> Lf2.
[conditions:
(1) Calculation of negative acceleration NA When TA ≧ 0 when the subject is standing, NA = (-1) × (TA)
If TA <0 when the subject is standing, NA = TA
(2) Calculation of the 1st decubitus time zone, the 2nd decubitus time zone, and the 3rd decubitus time zone _{1st decubitus time zone L m1} : NA ≧ C1 (C1 is a constant) is the first predetermined time T ₁ or more Second decubitus time zone L _m2 : There are two or more of the first decubitus time zones L _m1 , and NA <C1 between _{two adjacent first decubitus time zones L m11} and L _m12. When a certain gap time zone L _sm is within the second predetermined time T _2, the time zone obtained by summing the two adjacent first recumbent time zones L _m11 and L _m12 and the gap time zone L _sm.
Third recumbent time zone: A time zone in which NA ≥ C1 is _{within the third predetermined time T 3} , provided that the third predetermined time T ₃ is less than the first predetermined time T ₁ (3) Awakening. Calculation of time zone and sleep time zone Sleep time zone: The longest time zone of the first decubitus time zone L _m1 and the second decubitus time zone L _m2 during a predetermined measurement unit time Awakening time zone: Predetermined measurement Time zone excluding the sleep time zone from the unit time] The method of claim 1, wherein the index includes a variable term for the total time of the first recumbent time zone. The method according to claim 1 or 2, wherein the index includes a variable term of the beat interval of the sleep time zone. The method according to any one of claims 1 to 3, wherein the index includes a variable term of (LF / HF) / activity in the awakening time zone. The method according to any one of claims 1 to 4, wherein the index includes a variable term of beat interval / activity amount in the sleep time zone. The index is the variable term of the beat interval of the awakening time zone, the variable term of the LF / HF of the awakening time zone, the variable term of the activity amount of the awakening time zone, and the variable term of the LF / HF of the sleeping time zone. , At least the variable term of the activity amount of the sleep time zone, the variable term of the time of the sleep time zone, the variable term of HF × the activity amount of the awakening time zone, and the variable term of HF / activity amount of the sleep time zone. The method according to any one of claims 1 to 5 , which comprises any one. The method according to any one of claims 1 to 6 , wherein the index includes at least one of the variable term of HF in the awakening time zone and the variable term of HF in the sleeping time zone. The indicators are the SDNN variable term of the awakening time zone, the SDNN variable term of the sleeping time zone, the CVRR variable term of the awakening time zone, the CVRR variable term of the sleeping time zone, and the RMSSD of the awakening time zone. Variable term, RMSD variable term of the sleep time zone, NN50 variable term of the awakening time zone, NN50 variable term of the sleep time zone, pNN50 variable term of the awakening time zone, pNN50 of the sleep time zone The method according to any one of claims 1 to 7 , which comprises at least one of the variable terms of.
Here, SDNN is the standard deviation of the beat interval, CVRR is the value obtained by dividing the SDNN by the average value of the beat interval and multiplied by 100, and RMSSD is the average value of the square of the difference between adjacent beat intervals. The square root and NN50 are the total number of beats with an adjacent beat interval difference of more than 50 ms, and pNN50 is the ratio of beats with an adjacent beat interval difference of more than 50 ms. The method according to any one of claims 1 to 8 , wherein the RR interval, which is the interval between the R wave and the R wave in the electrocardiographic signal, is used as the pulsation interval. The method according to any one of claims 1 to 9 , wherein the index indicates the presence or absence or degree of depression or autonomic imbalance. A time calculation unit that calculates the awakening time zone and sleep time zone of the subject,
LF is a value obtained by definitely integrating the power spectrum obtained by including the step of converting the beat interval of the subject into a frequency spectrum from frequencies Lf1 to Lf2, and a value obtained by definitely integrating the power spectrum from frequencies Hf1 to Hf2. The processing unit that calculates HF and
It has an index creation unit that creates an index to determine the neuropsychiatric state, and
The time calculation unit calculates the negative acceleration NA from the acceleration TA in the height direction of the subject under the following conditions (1) to (3), and calculates the first decubitus time zone, the second decubitus time zone, and the second . 3 It calculates the lying time zone, the awakening time zone, and the sleeping time zone.
The index is a constant term, a beat interval of the awakening time zone × a variable term of the activity amount, a variable term of (LF / HF) × the activity amount of the sleep time zone, and the third recumbent time zone. It is expressed by an expression that includes the variable term of the total time and
In the index creation unit, using the data of a plurality of subjects, the constant term, the coefficient of the beat interval of the awakening time zone × the variable term of the activity amount, and the coefficient of the variable term of the sleep time zone ( LF / HF) × A device for creating an index for discriminating a neuropsychiatric state, which comprises calculating a coefficient of the variable term of the amount of activity.
Here, the amount of activity is an acceleration or an angular velocity accompanying the movement of the subject, and Hf1> Lf1 and Hf2> Lf2.
[conditions:
(1) Calculation of negative acceleration NA When TA ≧ 0 when the subject is standing, NA = (-1) × (TA)
If TA <0 when the subject is standing, NA = TA
(2) Calculation of the 1st decubitus time zone, the 2nd decubitus time zone, and the 3rd decubitus time zone _{1st decubitus time zone L m1} : NA ≧ C1 (C1 is a constant) is the first predetermined time T ₁ or more Second decubitus time zone L _m2 : There are two or more of the first decubitus time zones L _m1 , and NA <C1 between _{two adjacent first decubitus time zones L m11} and L _m12. When a certain gap time zone L _sm is within the second predetermined time T _2, the time zone obtained by summing the two adjacent first recumbent time zones L _m11 and L _m12 and the gap time zone L _sm.
Third recumbent time zone: A time zone in which NA ≥ C1 is _{within the third predetermined time T 3} , provided that the third predetermined time T ₃ is less than the first predetermined time T ₁ (3) Awakening. Calculation of time zone and sleep time zone Sleep time zone: The longest time zone of the first decubitus time zone L _m1 and the second decubitus time zone L _m2 during a predetermined measurement unit time Awakening time zone: Predetermined measurement Time zone excluding the sleep time zone from the unit time] The device according to claim 11 , wherein the index includes a variable term of the total time of the first recumbent time zone.

Images