JP2018082893A - Bathtub-type electro-cardio monitoring system, detection method of disease seizure during bathing using the same, optimal bathing condition setting method, health status analysis method and control program of these execution - Google Patents

Abstract

PROBLEM TO BE SOLVED: To propose a method for estimating unipolar chest 6 lead/lead (V-V, V, V) from 17 lead enabling actual measurement.SOLUTION: A bathtub-type electro-cardio monitoring system has a processing device, a bathtub being connected to the processing unit, a database and an output device. The system has 15-electrodes in an inner wall of the bathtub in which 13-electrodes are in the upper body side of a bather and 2-electrode in the foot part side. The processing device is characterized by finding the 6 lead signal (unipolar limb lead signal V, V, V, Vand additional lead V, V) of the chest side of the bather having no electrode based on the least squares method, by using the average projection matrix of multiple persons being stored in the database and the signal of 17 lead (bipolar limb lead (I, II, III) and unipolar limb lead (aVR, aVL, aVF, V, V, V, V, V, V, V, V, V, V, V)) enabling actual measurement during bathing from the 15-electrodes provided in the inner wall of the bathtub.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a bathtub type electrocardiogram monitoring system, a method for detecting a disease attack during bathing using the same, an optimal bathing condition setting method, a health condition analysis method, and a control program for carrying out these.

  An electrocardiograph is used to read an electrocardiogram waveform and derive an electrocardiogram for detecting various heart diseases. An electrocardiograph is used to measure an extracellular potential due to a myocardial action potential.

  In general, in order to induce such an electrocardiogram, an electrocardiogram having a standard 12-lead waveform is detected and measured using 10 electrodes, and recording is performed if necessary.

That is, electrodes are mounted at 6 locations for measurement of chest guidance and at 4 locations for limb guidance, respectively. Based on the electrocardiogram detected from these 10 electrodes by an electrocardiograph, the standard 12-lead limb 6-lead signal (I, II, III, aVR, aVL, aVF) and the standard 12-lead chest 6-lead signal ( _{V 1, V 2, V 3} , V 4, V 5, V 6) operation on the outputs.

The relationship between the induced waveform for obtaining such a standard 12-lead electrocardiogram and the electrocardiogram at the measurement site is as follows.
I induction: vL-vR
II Induction: vF-vR
III induction: vF-vL
aVR induction: vR- (vL + vF) / 2
aVL induction: vL- (vR + vF) / 2
aVF induction: vF- (vL + vR) / 2
V ₁ induction: v1- (vR + vL + vF ) / 3
V ₂ induction: v2- (vR + vL + vF ) / 3
V ₃ induction: v3- (vR + vL + vF ) / 3
V ₄ induction: v4- (vR + vL + vF ) / 3
V ₅ induction: v5- (vR + vL + vF ) / 3
V ₆ induction: v6- (vR + vL + vF ) / 3
However, vX (X = L, R, F, 1 to 6) is a potential detected at each electrode mounting position.

  Measurements using such a large number of electrodes are possible when the patient is at rest in a fully equipped hospital.

  However, when performing home medical care or emergency medical care, it is often difficult to use a large number of electrodes and attach each electrode at an appropriate position on the body surface of the living body.

  In general, an ECG signal can be transmitted in about one channel (one lead). Therefore, diagnosis for heart disease is performed by detecting and measuring an electrocardiogram composed of several waveforms within a standard 12-lead waveform using 2 to 4 electrodes.

  On the other hand, it is estimated that 17,000 people die from cardiopulmonary arrest during bathing. In view of this point, various developments have been made as a method and apparatus for obtaining an electrocardiogram during bathing while avoiding the troublesome wearing of the electrode on the body surface (Patent Documents 1-6, 8). And 9).

  Patent Document 1 discloses that three electrodes are attached to the inner wall surface of a bathtub with a magnet and the heart rate during bathing is measured.

  Patent Document 2 shows that a plurality of electrode pairs are attached to the inner wall of a bathtub, the waveform having the largest amplitude is used for processing, and the influence on the electrocardiographic signal amplitude due to the posture position during bathing is suppressed.

  Patent Document 3 discloses that the Seki electrode is not soaked in bathtub water but is installed on both sides on the bathtub wall, and the entire stainless steel bathtub is used as an indifferent electrode. Furthermore, it is disclosed that only one person can perform electrocardiogram measurement even if a plurality of persons are bathed together by holding the Seki electrode with two hands.

  In Patent Document 4, an electrocardiographic amplifier circuit is accommodated in a water-tight case, and three electrodes are attached to the outside. It is disclosed that an electrocardiogram is automatically measured when a case is floated on a hot water surface and a bather approaches.

  Patent Document 5 attaches three electrodes to a bathtub wall, measures an electrocardiogram during bathing, and calculates the power of HFC and LFC of HRV. Comparing the calculation result with a predetermined threshold, the intensity and operating time of the human body stimulation generator such as jet bath device, acoustic device, light control device, fragrance device, temperature control device etc. according to the physiological and psychological state of the human body To control.

  Patent Literature 6 discloses that a sucker is attached to an arbitrary position in a bathtub by fitting a substantially conical conductive band (aluminum) to a sucker that can be attached to and detached from the inner wall of the bathtub according to the physique and posture of the bather. doing.

  Patent Document 8 shows a metal wire type electrode (annular bend type, stranded wire type) capable of measuring a stable electrocardiogram without being affected by fluctuations in bathtub water, and using six electrodes, six limb-leading electrocardiographs. Is derived (I, II, III, aVR, aVL, aVF).

  Patent Document 9 measures I, II, and III induction signals using three detection electrodes and one common electrode. An electrocardiogram signal is extracted using a 1-100 Hz bandpass filter, and a respiratory signal is extracted using a 0.1-1 Hz bandpass filter.

  Patent document 9 sticks two electrodes in the longitudinal direction of the bathtub, and determines the bathing position based on the electrocardiogram signal amplitude of both. Estimate the bather's “hot” state using the difference in body surface temperature between the back and foot sides. Adjust the bathing environment according to the "no hotness" state. Disclosing hot water from the drainage device, lowering the body surface temperature, and recovering from a “hot” state are disclosed.

Japanese Patent Laid-Open No. 5-95921 JP-A-5-95923 JP-A-5-103766 JP-A-5-111470 Japanese Patent No. 2594217 JP-A-10-155757 Japanese Patent No. 4586008 JP 2008-93100 A Japanese Patent No. 5048568 JP 2008-100080 A

  The conventional method and apparatus for obtaining an electrocardiogram during bathing disclosed in the above patent document have the following characteristics.

  Basically, three electrodes are used. Even when a larger number of electrodes are used, an electrode that outputs a good signal is searched and the output from this electrode is used. The idea of using multiple leads is not shown.

  Moreover, in the patent document which discloses the technique for obtaining an electrocardiogram during bathing, only the measurement of the electrocardiogram is performed, and a scenario for data analysis and use based thereon is not shown.

  Furthermore, there is no scenario regarding the analysis and use of ECG abnormality diagnosis.

  The invention disclosed in Patent Document 10 is an improvement of the invention related to the method and apparatus for constructing a standard 12-lead ECG disclosed in Patent Document 7. That is, according to the invention disclosed in Patent Document 7, a highly accurate standard 12-lead electrocardiogram can be obtained. For example, when a stenosis of a coronary artery that sends blood to the myocardium occurs in the posterior wall of the myocardium in myocardial infarction, Since the electrode mounting site is far from the posterior wall, the sensitivity of the electrocardiogram waveform is low, and it is difficult to reflect the influence on the electrocardiogram waveform due to the stenosis of the coronary artery.

For this reason, V ₇ , V ₈ , V ₉ leads measured at the electrode mounting site on the extension line of the chest lead, and V _3R , V _4R , V depending on the electrode part at the target position with the electrode part of the chest lead. _An additional electrocardiogram is obtained from the _5R and _{V6R leads} .

Here, in order to obtain the additional lead electrocardiograms for leads V ₇ , V ₈ , and V ₉ , it is necessary to attach the electrodes for the additional lead measurement to a position such as the back of the patient. From this point, Patent Document 10 discloses that an additional induction can be easily obtained without attaching an electrode to the electrode used in 12 induction.

  However, Patent Documents 7 and 10 do not disclose any electrocardiogram measurement during input.

Here, the case of arranging the electrode in a bath, unipolar chest 6 induced induction of the front wall _{(V 1 ~V 4, V 3R} , V 4R) is impossible measured. Therefore, the object of the present invention is to propose a method for estimating the monopolar chest 6-lead (V _{1 to} V ₄ , V _3R , V _4R ) from 17 leads that can be actually measured. is there.

  In addition, heart rate and respiratory rate are obtained from electrocardiogram signals measured in real time during bathing, and body condition is estimated from analysis of heart rate variability and related to pressure, sound, light, temperature and incense input. Presents technology to control external environmental elements. Furthermore, a bath-type electrocardiogram monitoring system that makes it possible to accumulate and use data during daily bathing over a long period of time, a method for detecting a disease attack during bathing using the same, a method for setting optimal bathing conditions, and a program for implementing this The purpose is to provide.

The 1st side according to the present invention which achieves the above-mentioned object is a bathtub type electrocardiogram monitoring system, and has a processing device, a bathtub connected to the processing device, a database, and an output device. The inner wall has 15 electrodes having the following positional relationship, 13 electrodes on the upper body side of the bather and 2 electrodes on the foot side,
No. Electrode name Position 1 V _LA left hand 2 V _RA right hand 3 V _LF left foot 4 V _RF right foot 5 V ₅ Intersection of fifth intercostal horizon and left anterior axillary line 6 V ₆ Intersection of fifth intercostal horizon and left central axillary line 7 V ₇ V Intersection of ₆ and left posterior axilla line Intersection of 8 V ₈ V ₆ and left central scapula line
After 9 V ₉ V ₃
10 V ₁₀ center rear 11 V _5R 5th intercostal horizontal line and right anxillary line intersection
12 V _6R Intersection of 5th intercostal horizontal line and right mid-axillary line 13 V _7R Intersection of _6R and right rear axillary line Intersection of 14 V _8R V _6R and right mid-axillary line _{15V 9R} V _3R Post-processing The apparatus includes an average projection matrix of a plurality of persons stored in the database, and 17-lead signals (bipolar limb leads [I, II, III], Tankyokushi induced with _{[aVR, aVL, aVF, V} 5, V 6, V 7, V 8, V 9, V 10, V 5R, V 6R, V 7R, V 8R, V 9R] a) It is characterized in that 6-lead signals (monopolar limb induction signals V ₁ , V ₂ , V ₃ , V ₄ and additional inductions V _3R , V _4R ) on the chest side of the bather who does not have an electrode are obtained by the least square method To do.

In the first aspect according to the present invention to achieve the above object, a first aspect, the average projection _V 5 estimated from the _{matrix, V 6, V 7, V} 5R, V 5 was measured with _{V 6R, V 6} , V ₇ , V _5R , V _6R as a correction term, a projection matrix tuned for each individual, and can be measured during bathing by excluding V ₅ , V ₆ , V _5R , V _6R from the 17-lead signal Using the 13-lead signal, the 6-lead signal on the chest side of the bather who does not have an electrode is obtained by the method of least squares.

  The second aspect according to the present invention for achieving the above object is a method for detecting a disease attack during bathing using the bathtub-type electrocardiographic monitoring system, the method comprising measuring an electrocardiogram signal from an electrode on the inner wall of the bathtub, A step of performing preprocessing on the measured electrocardiogram signal; a step of analyzing an electrocardiogram signal based on a preprocessing result of the measured electrocardiogram signal; and an arrhythmia or myocardium based on the analysis of the electrocardiogram signal It has the process of detecting the abnormal event of an infarction.

  The first aspect of the second aspect according to the present invention that achieves the above object is the step of obtaining an electrocardiogram analysis result from an electrocardiogram signal measured by the step of measuring the electrocardiogram signal as the preprocessing, A preprocessing step for removing baseline fluctuations and hum in the electrocardiogram signal; and a step of emphasizing the QRS after the preprocessing and detecting a QRS feature point.

  The second form of the second aspect according to the present invention that achieves the above object is the step of dividing QRS feature point data detected by the step of detecting the QRS feature points into data for each heart rate, and the division. The method includes a step of detecting a heart rate abnormality based on RRI time-series data detected in the step, and a step of determining a medical condition caused by arrhythmia when the heart rate abnormality is detected by the step.

  The third form of the second aspect according to the present invention that achieves the above object is the step of dividing QRS feature point data detected by the step of detecting the QRS feature points into data for each heart rate, and the division. And a step of determining the state of myocardial infarction in response to the ST level variation detected in the step.

  The third aspect according to the present invention that achieves the above object is an optimum bathing condition setting method using the bathtub type electrocardiographic monitoring system, the step of measuring an electrocardiographic signal from an electrode on the inner wall of the bathtub, A step of pre-processing the measured electrocardiogram signal; a step of analyzing heart rate variability based on a pre-processing result of the measured electrocardiogram signal; and a frequency parameter of heart rate variability based on the analysis of the heart rate variability And a step of analyzing the comfort level based on the obtained frequency parameter of the heartbeat variability.

  The first form of the third aspect according to the present invention that achieves the above object is characterized in that the step of analyzing the comfort level based on the frequency parameter of the heartbeat variability is determined based on the bath water temperature as the comfort level. To do.

  The second form of the third aspect according to the present invention that achieves the above object is characterized in that the step of analyzing the comfort level based on the frequency parameter of the heartbeat variability is determined based on the bathing time as the comfort level. To do.

  The 4th side according to the present invention which achieves the above-mentioned object is a health condition analysis method using the bathtub type electrocardiogram monitoring system, the step of measuring an electrocardiogram signal from the electrode which the inner wall of the bathtub has, and the measurement Performing pre-processing on the measured ECG signal, analyzing the heart rate fluctuation based on the pre-processing result on the measured ECG signal, and analyzing the analysis data of the heart rate fluctuation at predetermined intervals. And a step of detecting a parameter of a predetermined health condition based on the analysis for each predetermined period.

  The 5th side according to the present invention which achieves the above-mentioned object is a control program which controls execution of the disease seizure detection method during bathing using a bathtub type electrocardiogram monitoring system, and has in the processing unit, the inner wall of the bathtub A step of measuring an electrocardiogram signal from an electrode; a step of performing pre-processing on the measured electrocardiogram signal; a step of analyzing an electrocardiogram signal based on a pre-processing result for the measured electrocardiogram signal; And a step of detecting an abnormal event of arrhythmia or myocardial infarction based on an analysis of an electrocardiogram signal.

  The first aspect of the fifth aspect according to the present invention that achieves the above object is the step of measuring an electrocardiogram signal from an electrode on the inner wall of the bathtub in the processing device, and a step before the measured electrocardiogram signal. A step of performing processing, a step of analyzing heart rate variability based on a pre-processing result for the measured electrocardiogram signal, a step of analyzing the analysis data of the heart rate variability for each predetermined period, and for each predetermined period And a step of detecting a parameter of a predetermined health condition based on the analysis.

It is one structural example of the bathtub type electrocardiogram monitoring system according to the present invention. It is a figure explaining the electrode position installed in the bathtub 1 used for the electrocardiogram monitoring according to this invention. It is a figure which expands and shows electrode arrangement of a bathtub for easy understanding It is a figure which shows the data analysis flow according to this invention. It is a figure which shows the waveform of a general electrocardiogram signal. It is a flowchart of an electrocardiogram signal preprocessing. It is a figure which shows the processing flow when performing the characteristic analysis of an electrocardiogram signal. It is a figure which shows the processing flow which detects the detection (arrhythmia) of an abnormal event. It is a figure which shows the processing flow which detects the detection (myocardial infarction) of an abnormal event. It is a figure which shows the processing flow of heart rate fluctuation | variation analysis. It is a figure which shows the processing flow when performing optimal bathing condition setting (tub water temperature). It is a figure which shows the processing flow when performing optimal bathing condition setting (bathing time). It is a figure which shows the processing flow when performing a health condition fluctuation | variation analysis.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. These examples are for facilitating the understanding of the present invention, and the application of the present invention is not limited to these examples. Further, the scope of protection of the present invention covers the same or similar scope as the claims.

  FIG. 1 is a configuration example of a bathtub type electrocardiogram monitoring system according to the present invention. The bathtub 1, the database 3, and the output device 4 are connected around the processing device 2 such as a personal computer.

  The processing device 2 detects the electric potential of the electrode stretched on the wall of the bath 1 during bathing by executing a stored program, and performs electrocardiographic monitoring according to the present invention.

The projection matrix used when estimating the unipolar chest 6 leads (V _{1 to} V ₄ , V _3R , V _4R ) of the front wall, which will be described later, from 17 leads that can be actually measured is the electrode arrangement, the electrode, the bather, This is related to the relative position of the bather, the size of the bather, etc. In fact, it is difficult to obtain the projection matrix of each person at the time of bathing. Therefore, as the projection matrix, 23-lead measurement signals of a plurality of persons are obtained in advance, stored in the database 3, and used.

  FIG. 2 is a diagram for explaining electrode positions installed in the bathtub 1 used for electrocardiographic monitoring according to the present invention. The electrode arrangement | positioning in the top view of the bathtub 1 is shown to the upper direction of FIG. 2, and the outline of the bather's side surface and the height position of an electrode are shown to the downward direction. The wall of the bathtub 1 is provided with 13 electrodes in the region corresponding to the head side during bathing and 2 electrodes in total on the foot side. The position of point C is the position of the heart.

FIG. 3 is an enlarged view showing the electrode arrangement of the bathtub 1 for easy understanding. As shown in an enlarged view in FIG. 3, in the bathtub 1, electrodes cannot be provided at the positions of V _{1 to} V _{4 on} the front wall.

  The example of FIG. 2 is shown using the thoracic guidance axis published by Simonson, but the same principle can be applied to the Chou arrangement and other arrangements. The installation positions of the 15 electrodes are the corresponding positions as follows.

No. Electrode name Position 1 V _LA left hand 2 V _RA right hand 3 V _LF left foot 4 V _RF right foot 5 V ₅ Intersection of fifth intercostal horizon and left anterior axillary line 6 V ₆ Intersection of fifth intercostal horizon and left central axillary line 7 V ₇ V Intersection of ₆ and left posterior axilla line Intersection of 8 V ₈ V ₆ and left central scapula line
After 9 V ₉ V ₃
10 V ₁₀ center rear 11 V _5R 5th intercostal horizontal line and right anxillary line intersection
12 V _6R Intersection of 5th intercostal horizontal line and right middle axillary line 13 V _7R Intersection of _6R and right rear axillary line 14 V _8R Intersection of _6R and right central axillary line After 15 V _9R V _3R

In FIG. 2, the induction axis for monopolar chest guidance is a line connecting the electrical center C of the heart and 15 electrode mounting sites around the chest. The polarity of the induction axis is (+) at the electrode mounting site and (−) at the electrical center of the heart. V ₂ is induction front, V ₆ is views of the electromotive force vector from the left side.

  The electrocardiogram of 23 leads is monitored during bathing from the electrode arrangement shown in FIG.

  The 23-lead ECG can be acquired with standard 12 leads (8 actual measurement signals + 4 derived signals) and additional 11 leads (9 actual measurement signals + 2 derived signals).

  The standard 12-lead 8 actual measurement signal is obtained as follows.

8 measured signals = three bipolar lead (I, II, III) +3 single-pole limb leads (aVR, aVL, aVF) is +2 single-pole chest lead _(V 5, V _6).

  The 9 actual measurement signals for the additional 11 leads are obtained as follows.

9 Actual measurement signal = additional monopolar chest 9 lead, that is, V ₇ , V ₈ , V ₉ , V ₁₀ , V _5R , V _6R , V _7R , V _8R , V _9R

  The six derived signals (6 derived signals) are as follows.

6 derived signals = 4 unipolar chest leads (V ₁ , V ₂ , V ₃ , V ₄ ) in standard 12 leads + 2 additional leads (V _3R , V _4R )

  That is, the present invention is characterized in that the chest side potential (6 derivation signal) where no electrode is arranged is derived from the electrodes arranged in the bathtub 1 by calculation.

  Here, bipolar induction is a method of obtaining an electrocardiogram by measuring a potential difference between two electrodes. For example, standard limb guidance (I, II, III guidance). On the other hand, monopolar induction is a method of obtaining an electrocardiogram by measuring a potential difference between a measurement electrode and a potential base point called an “indifferent electrode”.

  Further, the above derivation method uses the least square method based on the lead vector principle.

Average projection matrix and the measured signal 17 derived, i.e. _{(I, II, III, aVR} , aVL, aVF, V 5, V 6, V 7, V 8, V 9, V 10, V 5R, V 6R, V 7R, _V8R , _V9R ) are used.

  This derivation calculation procedure will be described later, but the outline will be described first for easy understanding.

  An optimal average projection matrix is obtained from the database 3 that estimates the unipolar chest 6 lead of the anterior wall by the least square method using a signal of 17 leads (the number of leads that can be actually measured during bathing). All bathers who are subjects to be measured use this average projection matrix and the 17-lead signal actually measured at the time of bathing to estimate the monopolar chest 6 lead of the front wall.

As another method, using the signal of 13 leads (the number of leads that can be measured during bathing excluding V ₅ , V ₆ , V _5R , and V _6R ), the monopolar chest 6 on the front wall is obtained by the least square method. An optimum average projection matrix is obtained from the database 3 for estimating the guidance. Correction terms of individualized projection matrix, the average projection _V 5 estimated from the _{matrix, V 6, V 5R, V} 6R guided actually measured _{V 5, V 6, V 5R} , determined from the difference between _{V 6R} induction. A projection matrix tuned for each individual is constructed using the optimum average projection matrix and the correction term for the individual projection matrix. Bathing person to be measured, by using a signal personalized tuning the projection matrix to the 13 induction was measured during bathing _{(V 5, V 6, V} 5R, number of time measurable induction bathing excluding the V _6R), before Estimate monopolar chest 6 lead in the wall.

  Next, the calculation process of the derivation will be described.

  The electrocardiogram v of the body surface is a projection of a lead vector l and a heart vector h.

  However, l depends on the electrode position.

  If three electrocardiograms V are measured with three independent leads L, h can be calculated.

Substituting equation (2) into equation (1),

  That is, according to the induction theory, the electrocardiogram v on the arbitrary lead vector l can be calculated by the equation (3), and the electrocardiogram of the arbitrary induction (position) can be derived from the linear combination of the three measured electrocardiograms V. is there.

  The chest 6 lead is derived using the measured signal 17 lead as an objective.

In order to obtain the projection matrix A = {a _ij }, the projection matrix A is obtained by the least square method using data in the database 3 (FIG. 1).

  The projection matrix A thus obtained reflects an average volume conductor model.

  Rewrite equation (5) as a determinant.

v _j changes with time, so when finding the optimal A, it contains at least data for one heartbeat.

Therefore,

 Rewrite equation (7) into a determinant.

here,

Reprinting equation (8)

Estimated value
And the difference between the measured values u _i is minimized.

e _i is minimized when orthogonal to 17 measurement leads v _j . That is,

Expanding all of Equation (14),

To organize this,

Therefore,

Here, a _i is a projection matrix (17 × 1) for deriving the guidance i. v is a measured 17-lead electrocardiogram (ECG) signal (formula (10)), u _i is an actual measurement signal (N × 1) that is a reference for derived leads (i = 1, 2, 3, 4, 3R, 4R). 1).

The above equation (12) is used to derive v _i .
Is the derived signal vector of the induction i (time-series data N × 1).

In order to derive signals of 6 leads (v _{1 to} v ₄ , v _3R , v _4R ) from 17 measured signals using all 23 lead ECG signals of the database 3, six average projection matrices a _i (a ₁ to a _4, a _3R, seek a _4R).

6-lead ECG of v ₁ to v ₄ , v _3R and v _4R using the 17-lead signal that can be directly measured during bathing and the average projection matrix a _i (a _{1 to} a ₄ , a _3R , a _4R ) Deriving the signal.

Further, in order to increase the accuracy, a method of tuning the average projection matrix a _i (a _{1 to} a ₄ , a _3R , a _4R ) for each individual is used. Perform tuning according to the following procedure.

Using the total lead ECG data of database 3, the average projection matrix (a ₅ , a ₆ , a _5R from the measured signal of 13 leads (excluding v _5, v _6, v _5R, v _6R during the 17 lead) , a _6R ) (using the least squares method with lead vectors).

When bathing, using 13 induction signals from the measurement signals of the induced _{17 (v 5, v 6,} v 5R, except v _6R), personalized projection matrix _{a 5t, a 6t, a 5Rt} , seeking a _6Rt (read (Least square method based on vector principle is used).

Differences between four average projection matrices (a ₅ , a ₆ , a _5R , a _6R ) and four individual projection matrices (a _5t , a _6t , a _5Rt , a _6Rt ) Δa _5, Δa _6, Δa _5R, Δa _{Find 6R} .

Then the average of the differences Δa _5, Δa _6, Δa _5R, Δa _6R
Ask for.

6 individual projection matrices (a _1t , a _2t , a _3t , a _4t , a _3Rt , a _4Rt ) from 6 average projection matrices (a ₁ , a ₂ , a ₃ , a ₄ , a _3R , a _4R ) Is obtained by the following equation.

As a result, signals of 6 leads (v _{1 to} v ₄ , v _3R , v _4R ) that are individually tuned are derived using the above 6 individual projection matrices. Α _i may be determined empirically as a decimal number between ± 1.0.

The following effects can be expected from the electrocardiogram signal corresponding to the electrode placement position obtained by the above method and the six derived signals (v _{1 to} v ₄ , v _3R , v _4R ) derived.

  First, real-time analysis of electrocardiogram signals enables immediate detection of disease attacks during bathing and ensuring safety.

  It is estimated that about 17,000 people die from cardiopulmonary arrest during bathing. Avoiding dangers due to seizures during bathing by immediate detection of abnormal cardiac function events such as arrhythmia, conduction disorders, acute myocardial infarction (especially the right ventricle and posterior wall, which are usually low in sensitivity), drainage treatment, and alerting Is possible.

  Second, it is possible to optimize the individual bathing environment. For this reason, it is possible to achieve the highest comfort level by evaluating the comfort level based on the evidence or constantly tracking the comfort level.

  Third, daily health management and early screening for diseases are possible. That is, monitoring of long-term health changes and accumulation of electrocardiographic signals during daily bathing over a long period of time are useful for constructing an optimal bathing environment, early prediction of disease onset, and daily health management by estimating changes in health status.

  Hereinafter, a specific method and system for obtaining the above-described effect obtained by real-time analysis of the electrocardiogram signal will be described.

  FIG. 4 is a diagram showing a data analysis flow according to the present invention. Processing following the subsequent flow is executed in the processing device 2 of FIG.

  Here, a waveform of a general electrocardiogram signal is shown in FIG. 5 for the following explanation. The electrocardiogram signal appears periodically in the order of P wave, Q wave, R wave, S wave, and T wave. A QRS composite wave is composed of the Q wave, the R wave, and the S wave.

[ECG signal preprocessing]
In FIG. 4, electrocardiographic signal preprocessing is performed as initial processing (step S0). FIG. 6 is a flowchart of this ECG signal preprocessing.

  In FIG. 6, an electrocardiographic signal obtained from the bathtub 1 during bathing is measured (step S10). If the signal to be measured is not good (step S11, N), the signal is measured again after a time of about 10 seconds (step S12).

  Based on a good measurement signal, electrocardiogram data is set (steps S11, Y, step 13).

  Baseline fluctuations and hum noise are removed as preprocessing for the electrocardiographic data (step S14). Next, the QRS complex wave is emphasized by the matched filter (step S15), and the QRS complex wave is detected by the peak search (step S16).

  Returning to FIG. 4, the characteristic analysis of the electrocardiogram signal is performed based on the detected QRS composite wave (step S1). The feature analysis of the electrocardiogram signal is performed according to the flow shown in FIG.

[Characteristic analysis of ECG signal]
7, the data for each heartbeat is divided based on the electrocardiogram signal subjected to noise processing in steps S13 and S14 in FIG. 6 and the QRS feature point data detected in step S16 in FIG. 6 (step S20). .

  That is, the divided data enables identification of each of the P wave, QRS composite wave, and T wave (steps S21, S23, and S27).

  The PQ interval is obtained by P wave identification (step S21) (step S22). R-wave identification (step S24), QRS width detection (step S25), and Q-wave identification (step S26) are performed by QRS composite wave identification (step S23).

  Further, identification of the QT interval (step S28) and detection of ST level fluctuation (step S29) are performed by T wave identification (step S27).

  Returning to FIG. 4, based on the result of the electrocardiogram signal feature analysis (FIG. 7), detection of an abnormal event (arrhythmia) (step S2) and detection of an abnormal event (myocardial infarction) (step S3) are performed.

  Detection of an abnormal event (arrhythmia) (step S2) is performed according to the processing flow shown in FIG. On the other hand, detection of an abnormal event (myocardial infarction) (step S3) is performed according to the processing flow shown in FIG. Next, detection processing of these abnormal events will be described sequentially.

[Detection of abnormal event (arrhythmia)]
In FIG. 8, RRI time series data is calculated based on the QRS feature point data obtained by the electric signal preprocessing of FIG. 6 (step S30).

  The RRI time series is a sequence of heartbeat intervals (RRI) arranged on a time axis. The RR interval is plotted and generated from the feature point data of QRS. Since RRI data is not uniform in time interval, it must be resampled at equal intervals before analysis in the frequency domain.

  Based on the obtained RRI time series data, when it is determined that the heart rate is abnormal (step S31, Y), is it a bradycardia (step S32), is a ventricular extrasystole (step S33), or is not atrial? It is determined whether it is contraction (step S34), and if it is any, an alarm is transmitted (step S40).

  This alarm is transmitted by displaying sound, light, etc. by an appropriate output device 4 in the system of FIG. Alternatively, it may be transmitted and displayed at an appropriate place by communication.

  Returning to FIG. 8, if the heart rate abnormality is not any of bradycardia, premature ventricular contraction, and premature atrial contraction, it is determined whether it is tachycardia (step S35).

  If it is tachycardia (step S35, Y), if it is a narrow QRS (step S36, Y), the presence or absence of regularity is determined (step S37A).

  If the tachycardia is regular (steps S37A, Y), it is determined that “sinus tachycardia, atrioventricular nodal reentry tachycardia, atrial flutter, atrial tachycardia, WPW syndrome” (step 38A).

  On the other hand, if the tachycardia is not regular (steps S37A, N), it is determined that “atrial fibrillation, atrial flutter, atrial tachycardia, multi-source atrial tachycardia” (step 38B). In either case, an alarm is transmitted (step S40).

  Further, if the QRS is not narrow (step S36, N), the regularity of tachycardia is determined (step S37B). If there is regularity (step S37B, Y), it is determined that “ventricular tachycardia, superior tachycardia, WPW syndrome” (step S39A), and an alarm is transmitted (step S40). If there is no regularity (step S37B, N), an alarm is issued (step S40) as "atrial fibrillation, vibration coarse movement" (step S39B).

[Detection of abnormal events (myocardial infarction)]
In FIG. 9 showing the processing flow of abnormal event detection (myocardial infarction), when the ST level fluctuation obtained by the electrocardiographic signal feature analysis of FIG. 7 is monitored and the ST level fluctuation rises (step S41, Y), 2 When the induction (II, III, aVF) or more exceeds 0.1 mV (step S42, Y), it is determined as acute inferior myocardial infarction (step 42A), and an alarm is transmitted (step S46).

  Further, when the continuous two or more leads of the anterior chest exceed 0.2 mV (step S43, Y), it is determined as acute anterior myocardial infarction (step 43A), and an alarm is transmitted (step S46).

  Further, when the continuous posterior wall 2 lead or more exceeds 0.1 mV (step S44, Y), it is determined as acute posterior myocardial infarction (step 44A), and an alarm is transmitted (step S46).

  On the other hand, when the ST level fluctuation does not increase (step S41, N), when two or more inductions exceed 0.1 mV (step S47, Y), when the serum CK-MB value or troponin (TnT) is positive (S48, Y) When non-ST elevation change myocardial infarction (step S49A), when serum CK-MB value or troponin (TnT) is not positive (S48, N), myocardial ischemia (step S49B) is issued (step S46). ).

[Heart rate fluctuation analysis]
Returning to FIG. 4, heart rate variability analysis is performed (step S4) after the ECG signal pre-processing (step S0, FIG. 6).

  FIG. 10 is a processing flow of this heart rate fluctuation analysis. FIG. 10 shows only some examples of HRV analysis parameters. There are many other items, all of which are not listed.

  In FIG. 10, RRI time series data is calculated (step S50) based on the feature point data (step S16, FIG. 6) of the QRS obtained by the ECG signal preprocessing.

  An abnormal value and noise are removed from the calculation result (step S51), and nonlinear analysis (step S52A) and linear analysis (step S52B) are performed.

  In the nonlinear analysis (step S52A), a phase space analysis (step S511), a chaos analysis (step S512), and a complexity analysis (step S513) are performed. Then, corresponding to each analysis, parameters 511A, 512A, and 513A are obtained.

  On the other hand, in linear analysis (step S52B), time domain analysis (step S521) and frequency domain analysis (step S522) are performed. Then, corresponding to each analysis, parameters 521A and 522A are obtained.

  These parameters are stored in the database 3 as a result of heart rate variability (HRV) analysis.

  Next, returning to FIG. 4, based on the result of heart rate fluctuation analysis (step S4), bath water temperature setting (step S5), bathing time setting (step S6), and health condition fluctuation analysis (step 7) are set as optimum bathing conditions. I do.

  The processing of each step will be described below.

[Optimal bathing condition setting (bath water temperature)]
FIG. 11 is a processing flow for setting the bath water temperature. 11, on the first day (step S60, Y), and inputs the initial bath temperature by bathers as T = _{T 0} (step S61).

  On the other hand, if bathing is not the first day (step S60, N), the bath water temperature is set to T = T0 + ΔT (step S61A). Here, the range of ΔT is ± 5 ° C., and the change step is 0.5 ° C. The bath water temperature is set at such a temperature (step S62).

  Next, an electrocardiogram signal is measured (step S63). If the quality of the measured electrocardiogram signal is not good, the measurement is performed again after a time of about 10 seconds (steps S64N, 63A).

  Based on the obtained electrocardiogram signal, heart rate fluctuation is analyzed (step S65). This heart rate variability analysis is the same as the processing described above with reference to FIG. 10, and the heart rate variability frequency parameters LF (low frequency), HF (HF), which are stress indicators for measuring the balance of the autonomic nervous function finally obtained. High frequency), LF / HF (step 66, FIG. 10, 522A))), comfort level analysis is performed (step S67).

  The HF component reflecting the respiratory fluctuation appears in the heartbeat fluctuation only when the parasympathetic nerve is in tension (activated). On the other hand, the LF component appears in heart rate variability both when the sympathetic nerve is tense and when the parasympathetic nerve is tense. Therefore, in the comfort level analysis (step S67), the high frequency fluctuation component (HF component) and the low frequency component (LF component) are extracted, and the magnitudes of both are compared to analyze the comfort level.

  The result is accumulated as data in the database 3 (step S68). Based on the accumulated data, a comfort curve is constructed (step S68A), and the comfort curve is sequentially updated. The change of the comfort level thus obtained with respect to the water temperature is shown in a graph (step S69), the bath water temperature at which the comfort level is optimal is obtained, and the optimal bath water temperature is specified (step S69A).

  Here, when the optimum bathtub water temperature setting in FIG. 11 is performed for each individual, the bather is identified using an operation panel, a weight scale, a face recognition camera, an electrocardiogram after bathing, and the like. In data storage (step S68, FIG. 11), data is stored separately for each individual.

  Therefore, the bath water temperature can be set (step S62, FIG. 11) based on the individual data stored for each individual.

  Returning to FIG. 4, FIG. 12 shows a processing flow in the case of performing optimal bathing condition setting (bathing time) based on the heart rate fluctuation analysis (step S <b> 4).

  The processing flow in FIG. 12 is basically the same as the processing described in FIG. The difference is that, based on the comfort level analysis (step S76), instead of the water temperature change, the change in the comfort level is left as data with the passage of time as the axis (steps S77, S78).

  In FIG. 12, in order to set the bathing time for each individual, the comfort level data for each individual stored in the database 3 is used.

A bather is identified using an operation panel, a weight scale, a face recognition camera, an electrocardiogram after bathing, and the like (steps S70 and 70A). The optimum bath water temperature _Topt corresponding to the identified bather is specified and set from the database 3 (step S71).

  An electrocardiogram signal during bathing is measured (step 72). If the quality of the measured electrocardiogram signal is not good, the measurement is performed again after about 10 seconds (steps S73N and 72A).

  Based on the obtained electrocardiogram signal, heart rate fluctuation is analyzed (step S74). This heart rate variability analysis is the same as the processing described above with reference to FIG. 10, and the heart rate variability frequency parameters LF (low frequency), HF (HF), which are stress indicators for measuring the balance of the autonomic nervous function finally obtained. High frequency), LF / HF (step S75, FIG. 10, 522A)), a comfort level analysis is performed (step S76).

  The result of the comfort level analysis is accumulated in the database 3 (step S77). Based on this accumulated data, the comfort curve is updated with time on the horizontal axis. The change of the comfort level obtained in this way with respect to the bathing time is shown in a graph (step S77A), and the processing from the electrocardiogram signal measurement (step S72) is repeated until the inflection point at which the comfort level changes is detected (step S72). S78, N).

  Then, when an inflection point at which the bath water temperature rises and the comfort level changes is detected over time (step S78, Y), a signal indicating that the optimum bathing time has been reached is transmitted (step S79). This can prompt the bather to leave the bathtub 1.

  Furthermore, returning to FIG. 4, it is possible to grasp the health condition variation from the data obtained from the electrocardiographic signal obtained in the present invention (step S7).

  FIG. 13 is a flowchart of such health condition fluctuation analysis. HRV analysis result data accumulated and stored in the database 3 is acquired by accumulation of heart rate variability (HRV) analysis result data (step 53) in FIG. 10 (step S80), and abnormal values are removed as preprocessing (step S81). . Further, the missing data is supplemented as continuous change data by interpolation processing as preprocessing (step S82).

  With regard to the HRV data stored in this manner, continuous change data is obtained every day (step S83A), every week (step S83B), every month (step S83C), every season (step S83DC), and every year (step S83E). To analyze.

  Based on these analysis results, bather biorhythm analysis (step S84A), physiological period estimation (step S84B), physical condition change detection (step S84C), and lifestyle change detection (step S84D) can be performed.

[Real-time processing]
Furthermore, as an extension of the present invention, it is possible to perform automatic start and end of data measurement by real-time signal processing and automatic detection of bath water level and electrocardiogram signal.

  That is, the basic water level before bathing is detected.

  When a bather enters the bathtub, a change in the water level and the appearance of an electrocardiogram signal are detected, and data measurement is automatically started.

  When the bather leaves the bathtub, the water level change and the disappearance of the electrocardiogram signal are detected, and the data measurement is automatically terminated.

  Moreover, it is possible to ensure safety by automatic drainage and alarm transmission. It can be configured to perform automatic drainage processing and alarm dispatch processing by immediate detection of abnormal events of cardiac function such as arrhythmia, conduction disorder, acute myocardial infarction (especially the right ventricle and posterior wall which are usually low detection sensitivity) is there.

1 Bathtub 2 Processing device 3 Database 4 Output device

Claims (12)

A bathtub type electrocardiogram monitoring system,
A processing device;
A bathtub connected to the processing device, a database, and an output device;
On the inner wall of the bathtub, there are 15 electrodes having the following positional relationship, 13 electrodes on the upper body side of the bather and 2 electrodes on the foot side,
No. Electrode name Position 1 V _LA left hand 2 V _RA right hand 3 V _LF left foot 4 V _RF right foot 5 V ₅ Intersection of fifth intercostal horizon and left anterior axillary line 6 V ₆ Intersection of fifth intercostal horizon and left central axillary line 7 V ₇ V Intersection of ₆ and left posterior axilla line Intersection of 8 V ₈ V ₆ and left central scapula line
After 9 V ₉ V ₃
10 V ₁₀ center rear 11 V _5R 5th intercostal horizontal line and right anxillary line intersection
12 V _6R Intersection of 5th intercostal horizontal line and right mid-axillary line 13 V _7R Intersection of _6R and right rear axillary line Intersection of 14 V _8R V _6R and right mid-axillary line _{15V 9R} V _3R Post-processing The device
An average projection matrix of a plurality of persons stored in the database and 17-lead signals (bipolar limb leads [I, II, III], monopolar limbs) that can be actually measured at the time of bathing from the 15 electrodes on the inner wall of the bathtub induced with _{[aVR, aVL, aVF, V} 5, V 6, V 7, V 8, V 9, V 10, V 5R, V 6R, V 7R, V 8R, V 9R] a), the least square method The 6-lead signals (monopolar limb induction signals V ₁ , V ₂ , V ₃ , V ₄ and additional inductions V _3R , V _4R ) on the chest side of the bather who does not have an electrode are obtained.
Bathtub type electrocardiographic monitoring system characterized by that. In claim 1,
Tuning the _V 5 estimated from the average projection _{matrix, V 6, V 7, V} 5R, V 5 was measured with _{V 6R, V 6, V 7} , V 5R, as correction term the difference between _{V 6R,} the personalized The bather who does not have an electrode by the least squares method using the projected matrix and the 13-lead signal that can be actually measured during bathing by removing V ₅ , V ₆ , V _5R , and V _6R from the 17-lead signal Obtaining the 6-lead signal on the chest side of
Bathtub type electrocardiographic monitoring system characterized by that. A method for detecting a disease attack during bathing using the bathtub-type electrocardiographic monitoring system according to claim 1 or 2,
Measuring an electrocardiogram signal from an electrode on the inner wall of the bathtub;
Pre-processing the measured electrocardiogram signal;
Analyzing an electrocardiogram signal based on a pre-processing result for the measured electrocardiogram signal;
Detecting an abnormal event of arrhythmia or myocardial infarction based on the analysis of the electrocardiogram signal;
A method for detecting a disease attack during bathing, comprising: In claim 3,
As the pretreatment,
Obtaining an electrocardiogram analysis result from the electrocardiogram signal measured by the step of measuring the electrocardiogram signal;
A pre-processing step for removing baseline fluctuations and hum with respect to the electrocardiogram signal;
Emphasizing the QRS after the preprocessing and detecting QRS feature points;
A method for detecting a disease attack during bathing, comprising: In claim 4,
Dividing QRS feature point data detected by the step of detecting the QRS feature points into data for each heart rate;
Detecting heart rate abnormality based on RRI time-series data detected in the dividing step;
A method for detecting a disease attack during bathing, comprising: a step of determining a medical condition caused by arrhythmia when an abnormal heart rate is detected by the above step. In claim 4,
Dividing QRS feature point data detected by the step of detecting the QRS feature points into data for each heart rate;
And detecting a state of myocardial infarction corresponding to the amount of ST level fluctuation detected in the dividing step. An optimal bathing condition setting method using the bathtub-type electrocardiographic monitoring system according to claim 1 or 2,
Measuring an electrocardiogram signal from an electrode on the inner wall of the bathtub;
Pre-processing the measured electrocardiogram signal;
Analyzing heart rate variability based on a pre-processing result for the measured electrocardiogram signal;
Obtaining a heart rate variability frequency parameter based on the analysis of the heart rate variability;
Analyzing the comfort level based on the obtained frequency parameter of the heart rate variability. In claim 7,
The step of analyzing the comfort level based on the frequency parameter of the heart rate variability is determined based on the bath water temperature as the comfort level. In claim 7,
The step of analyzing the comfort level based on the frequency parameter of the heart rate variability is determined based on the bathing time as the comfort level. A health condition analysis method using the bathtub-type electrocardiogram monitoring system according to claim 1 or 2,
Measuring an electrocardiogram signal from an electrode on the inner wall of the bathtub;
Pre-processing the measured electrocardiogram signal;
Analyzing heart rate variability based on a pre-processing result for the measured electrocardiogram signal;
A health condition analysis method comprising: analyzing the heart rate variability analysis data for each predetermined period; and detecting a parameter for a predetermined health condition based on the analysis for each predetermined period. A control program for controlling execution of a disease seizure detection method during bathing using the bathtub-type electrocardiographic monitoring system according to claim 1 or 2,
In the processing device,
Measuring an electrocardiogram signal from an electrode on the inner wall of the bathtub;
Pre-processing the measured electrocardiogram signal;
Analyzing an electrocardiogram signal based on a pre-processing result for the measured electrocardiogram signal;
And a step of detecting an abnormal event of arrhythmia or myocardial infarction based on the analysis of the electrocardiogram signal. A control program for controlling execution of a health condition analysis method using the bathtub-type electrocardiographic monitoring system according to claim 1,
In the processing device,
Measuring an electrocardiogram signal from an electrode on the inner wall of the bathtub;
Pre-processing the measured electrocardiogram signal;
Analyzing heart rate variability based on a pre-processing result for the measured electrocardiogram signal;
A control program, comprising: a step of analyzing the analysis data of the heart rate variability every predetermined period; and a step of detecting a parameter of a predetermined health state based on the analysis every predetermined period.