JP2003052655A - Method and apparatus for obtaining electrocardiogram - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain 12 lead electrocardiograms needed to diagnose a disease based on one derived electrocardiogram measured from electrodes brought into contact with or mounted at two positions easy to take on a subject's body. SOLUTION: A method for deriving the electrocardiogram comprises the steps of recording an electrocardiogram waveform measured between the electrodes brought into contact with or mounted at two positions of the subject's body as time series data, deciding an embedding delay time τ by using a self- correlation for the time series data, calculating a correlation dimension, deciding an embedding dimension (d), constituting an d-dimension attractor based on the delay time τ and the dimension (d), rotating the attractor according to a type of each derived electrocardiogram, and generating the derived electrocardiogram from a graphic form projected on a plane to be specified.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a method and an apparatus for deriving a 12-lead electrocardiogram from one measured electrocardiogram waveform.

[0002]

2. Description of the Related Art An electrocardiogram is a waveform chart in which the membrane potential of heart cells is taken out and recorded. Electrodes are attached to the limbs and chest, and changes in the potential difference between the electrodes are measured by connecting them to an electrocardiograph. When the heart has a disease, the electrocardiogram shows a unique waveform different from that in the normal case. There are 12 lead electrocardiograms as standard measurement methods in this electrocardiogram. Standard limb lead (3 leads), monopolar limb lead (3 leads) and chest lead (6 leads)
Induction).

The standard limb lead is bipolar lead, and the I lead is left hand-
Right hand, Lead II represents left foot-right hand, Lead III represents potential difference between left foot-left hand. The electrode on the right leg acts as a ground and does not participate in drawing the ECG waveform. Einthoven
The ven) triangle is an equilateral triangle fitted around the heart as shown in FIG. QRS derived from I, II and III
The wave height can be used to indicate the magnitude and direction of the cardiac electromotive force in the frontal plane by a vector.

The monopolar limb lead is a recording of electrical changes from the right shoulder, the left shoulder, and the diaphragm direction. The chest lead is a recording of electrical changes projected on a horizontal plane. FIG. 8 shows the electrode mounting sites V _{1 to} V ₆ for chest guidance.

As shown in FIG. 9, the electrocardiogram waveforms have P, Q,
The names R, S, T, and U are given. P reflects the excitement of the left and right atrium, but the excitement of the right atrium precedes the left atrium. Q
RS indicates the excitability of the ventricular muscle, and T indicates the process of reducing the excitement. The PQ interval is the time from the onset of right atrial excitation to the time when the stimulus reaches the ventricular myocytes through the atrioventricular node, His bundle, left and right legs, Purkinje fibers, and the QT interval is from the onset of ventricular excitation (depolarization). It is the time until it is depressed (repolarization). Note that J indicates the junction between S and T.

[0006] As described above, in conventional medicine, a total of four electrodes are attached to both hands and the right foot in the standard limb lead and the monopolar limb lead, and six electrodes are attached to the chest in the chest lead to 12 electrodes.
You are measuring different types of electrocardiograms. Heart disease was judged from this electrocardiogram.

[0007]

In the case of the standard limb lead and the monopolar limb lead of the electrocardiogram, the test subject is measured with the electrodes attached to the left and right wrists and legs and lying on his / her back on the bed. , In the case of chest guidance, it is necessary to attach the electrode directly to the chest, so the upper body must be exposed when wearing, and there are many people with resistance. Further, there is a problem that it takes a long time to measure the three types of lead electrocardiograms.

The present inventors presumed that there may be some correlation between the waveforms of the 12-lead electrocardiogram, and repeated measurements and correlations with many subjects. We have found that 12 lead electrocardiographic waveforms are derived from one electrocardiographic waveform.

The problem to be solved by the present invention is based on one electrocardiogram of one lead measured from electrodes which are contacted or attached to two places on the human body which are easy to be taken, and which are necessary for diagnosis of diseases. An object of the present invention is to provide an electrocardiogram deriving method and device capable of obtaining a lead electrocardiogram.

[0010]

In order to solve the above-mentioned problems, the method for deriving a lead electrocardiogram according to the present invention is applied to a human body of a subject.
The electrocardiogram waveform measured between the electrodes that are in contact with or attached to the place is recorded as time-series data, the embedding delay time τ is determined using autocorrelation for this time-series data, and the correlation dimension is calculated to determine the embedding dimension d. Then, a d-dimensional attractor is constructed based on the embedding delay time τ and the embedding dimension d, and the attractor is rotated according to the type of each electrocardiogram and projected from a figure projected on a specified plane,
A 12-lead electrocardiogram including the lead is generated.

The apparatus for deriving an electrocardiogram according to the present invention includes two electrodes that come into contact with or are worn on two points of the human body of a subject, a means for recording the electrocardiogram waveform between the electrodes as time series data, and the time series data. About the embedding delay time τ determined by using autocorrelation, and means for configuring a d-dimensional attractor based on the correlation dimension d determined by calculation, and rotating this attractor according to each lead electrocardiogram type And a means for generating a full 12-lead electrocardiogram including the lead based on the projected figure.

According to the present invention, twelve lead electrocardiograms required for diagnosing a disease can be generated on the basis of electrocardiograms measured from electrodes that are contacted with or attached to two locations of a human body of a subject.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to FIGS. In the present invention, when two electrodes are attached to two places of the human body of the subject, for example, the right hand and the left hand, the cardiac electromotive force existing between the right hand and the left hand in the human body can be measured between the electrodes. By applying the “Takens embedding theorem” based on the “chaos theory” to the time-series data (lead I) of the electrocardiographic waveform, it is possible to reconstruct the dynamics of the entire electrocardiographic system. The reason is based on the following concept.

The heart interacts with the autonomic nerves and blood vessels, and the heart itself also interacts with the right atrium, left atrium, right ventricle, left ventricle, ventricular septum, etc., and finally one heart organ. Is considered to be a "complex system" because it forms a non-linear movement.

[Complex system theor
y) ”refers to a system in which multiple elements move in relation to each other, in the time series of rhythms created by all elements,
When the time series data of one of the elements is obtained, it is guaranteed that the time series data of other elements can be reconstructed. For the reconstruction, the embedding theorem of Turkens in chaos theory, which forms the core of complex system theory, is used, and one parameter is restored to the other parameter under the differential isomorphism. For this, see the paper “F. Takens, Detectin
g strange attractors in turbulence, Lecture Note i
n Math., 898 (1980), 366-381, Springer-Verlag ”.

The following equation is the Lorentz chaos equation which is a typical example of a complex system.

[Equation 1]

FIG. 1 shows Lorentz chaos attractors drawn directly from this equation on the xy plane, the xz plane, and the yz plane. Further, FIG. 2 shows an attractor when the time series of the variable X is three-dimensionally reconstructed by applying the Turkens embedding theorem. As can be seen by comparing FIG. 2 with the (XY plane) of FIG. 1A, it can be said that the original structure can be almost restored from one variable.

As described above, since the heart is considered to be a complex system, the attractor reconstructed by the Takens embedding theorem can obtain other leads by obtaining one electrocardiogram data (time series). You should be able to restore it. In the present invention, an electrocardiogram waveform between electrodes that are in contact with or attached to two locations on the human body of a subject is recorded as time series data, and the embedding delay time τ is determined using autocorrelation for this time series data to determine the correlation dimension. The embedding dimension d is calculated to determine a d-dimensional attractor based on the embedding delay time τ and the embedding dimension d.

Here, the embedding delay time is a parameter for embedding time series data in a multidimensional phase space in order to draw a chaotic attractor. As shown in FIG. 3A, one observed time series data is s (k) (k =
0, 1, 2, 3 ...). Using this data, the vector T
Create (i) = (s (i), s (i + τ), s (i + 2τ)) (in case of 3D). Τ used here is an embedding delay time, which is a quantity representing a time delay. When this vector T (i) is sequentially plotted in the three-dimensional phase space (coordinate axes are T (i), T (i + τ) and T (i + 2τ) (i = 0, 1, 2, ..., n), Fig. 3 (b)
A trajectory like

Here, the attractor is a method used in a dynamical system, and refers to the point where the curve of the trajectory drawn in the phase space settles down. From a certain experimental fact, it is a geometric model that can be used when making a model of dynamics, and the geometric model used in the theory of dynamical systems is generally a manifold. A manifold is a generalized concept of lines and surfaces.

FIG. 4 (a), which is now measured as lead I
Assuming that the measured data of the electrocardiogram waveform as shown in Fig. 4 is {s (n)}, the embedding delay time τ and the embedding dimension d give
It is possible to reconstruct a d-dimensional attractor T (n) as shown in (b).

The mechanism of deriving another lead from the measured electrocardiogram waveform will be described with reference to the flowchart of FIG. Step 100: Electrocardiogram measurement First, the electrocardiogram is measured from one lead. Step 110: Generation of ECG digital data The measured ECG waveform is stored as digital data (time series data). Measured electrocardiogram data {s (n)}
And Step 120: Reconstruction of attractor Apply the Takens' embedding theorem to reconstruction of attractor, and give parameters as follows.・ Embedding delay time: τ ・ Embedding dimension: d By the embedding theorem of Turkens, the point T (n) representing the trajectory of the attractor can be expressed as follows.

[0023]

[Equation 2] As a result, the observed one-dimensional time series data is reproduced in the d-dimensional phase space.

Steps 130 and 140: Projection of attractor and rotation Attractor reconstructed in the multidimensional phase space is projected to a low dimension by a view direction unit vector. For example, projection from 4 dimensions to 3 dimensions is performed as follows. Given in matrix.

[0025]

[Equation 3]

Further, a rotation operation is applied to each of the axes to the attractor projected in the low dimension. The rotation matrix is as follows.

[Equation 4]

Step 150: Determination of derivation parameters When one derivation is completed, another embedding dimension, embedding delay time, view direction unit vector, rotation angle, projection axis are determined, and steps 120 to 140 are repeated.

Step 160: Time-series data in which each component of the orbital point of the attractor to which the operation of deriving another electrocardiogram and applying rotation is added is projected onto each axis. That is,
One of the 12 leads shown in 200 was measured,
The other 11 is a newly derived waveform.

The above methods are summarized as follows. (1) Measure the electrocardiogram waveform {s (n)}. (2) Embedding in a multidimensional phase space by the Turkens embedding theorem. The points of the attractor's orbit are

[Equation 5]

(3) Projection matrix given the viewing direction unit vector
Use P = {p _ij } to project to lower dimensions. The trajectory points of the attractor projected in the low dimension are

[Equation 6]

(4) Furthermore, the rotation axis and the rotation angle are given to the attractor, and the rotation operation is performed with the rotation matrix R = {r _ij }. The point on the attractor with the rotation operation is

[Equation 7]

Here, each orbital component of the obtained attractor derives a new time series. That is, the new time series {s' _i (n)} projected on the i-th axis is

[Equation 8] It is represented by.

[0033]

EXAMPLES Examples of the present invention will be described below. FIG. 6 shows an embodiment of the present invention. In the figure, 10 and 11 are electrodes, 20 is an electrocardiogram waveform measuring unit, 21 is a 12-lead waveform deriving unit, 22 is an electrocardiogram waveform monitor unit, and 23 is a 12-lead electrocardiogram automatic analysis. Reference numeral 24 denotes an automatic diagnosis unit. Further, when the present invention is used from a remote place outside the hospital via a communication line or the like, as shown by the broken line in the drawing, the communication portion 25
A communication line such as the Internet 26 connecting the hospital and the hospital 27 is required.

In this embodiment, the electrodes 10, 1 are placed on both hands.
1 was attached, the electrocardiogram of the lead I was measured by the electrocardiogram waveform measuring unit 20, and the data was stored as digital data. Specifically, for the electrocardiogram waveform of FIG. 4A, 0.001
30,000 points were sampled every second.

The 12-lead waveform is derived by the 12-lead waveform deriving unit 21 as follows. In the electrocardiogram data,
The embedding delay time is 40 points, the embedding dimension is 4 dimensions, and the XYZW 4
A view direction unit vector (X, Y, Z, W) = (0.5, 0.5, 0.5, 0.
5) Project to 3D phase space with (length 1),
The XY plane rotated about the Z axis at a rotation angle of 0 ° (without rotation) is illustrated.

[0036] By rotating only θ about the Z axis, X-axis, Y-axis, a waveform derived in the Z-axis _{s 'X (n), s} '
_{Calculate with Y} (n), s' _Z (n). According to the Turkens embedding theorem, the attractor is reconstructed with an embedding delay time of 40 points and an embedding dimension of 4. Point T (n) that represents the trajectory of the attractor
Is

[Equation 9]

Next, the attractor embedded in the four-dimensional phase space is set to 3 using the viewing direction unit vector (0.5, 0.5, 0.5, 0.5).
When projected onto a dimension

[Equation 10]

That is, a point on the three-dimensional phase space

[Equation 11] To get

When a point on the three-dimensional phase space is rotated counterclockwise by θ about the Z axis, the Z coordinate does not move. Therefore, calculation using a rotation matrix leads to the X axis and the Y axis. Waveform is obtained.

[0040]

[Equation 12] Waveform derived in the Z-axis is _{s 'Z (n) = t} ' is _n3.

The 12-lead electrocardiogram thus derived can be monitored or printed out by the electrocardiogram waveform monitor 22 to diagnose a disease. In addition, the derived waveform is a 12-lead ECG automatic analysis unit 23.
By using the above, the automatic diagnosis unit 24 can immediately judge a case and display or print out the result by comparing the closeness with a pattern (Minnesota code or the like) associated with a large number of cases in advance. Further, the output of the 12-lead electrocardiogram automatic analysis unit 23 is obtained by using the communication unit 25 and a communication system such as the Internet 26.
By transmitting the data to the hospital 27 or the like, it is possible to receive an appropriate instruction from the doctor.

[0042]

As described above, the present invention has the following effects. (1) A 12-lead electrocardiogram required for diagnosis of a disease can be obtained from an electrocardiogram measured by contacting or mounting electrodes at two locations that are easily taken by the human body of a subject. (2) The subject only has to attach the electrodes to the two easy-to-take places and does not need to attach the electrodes to the chest, so that there is no resistance and the measurement can be performed in a short time. (3) Since components other than the pair of electrodes can be configured by electronic means, the size and weight can be reduced, and the device can be portable and worn on the body. (4) It can be easily used for personal health checks, and can be used to check the physical condition of drivers and operators in facilities where no mistakes are allowed, such as before or during operation, or at nuclear power plants, or to check for acute heart disease. . (5) It can be used for remote care and diagnosis of patients who are located at a remote place (home, remote place) from a medical facility. (6) It is possible to grasp the lesion of a heart disease patient, such as an elderly person or a diabetic patient, who is less likely to have pain peculiar to the heart and is less likely to notice the lesion.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a Lorentz chaos attractor drawn on an xy plane, an xz plane, and a yz plane.

FIG. 2 is a diagram when only the time series of the variable X is three-dimensionally reconstructed by applying the Takens' embedding theorem.

FIG. 3 is a diagram for explaining a method according to the present invention.

FIG. 4A is a waveform diagram showing an example of time series data of an electrocardiogram waveform, and FIG. 4B is a diagram showing an example of an attractor of the electrocardiogram.

FIG. 5 is a flowchart showing a 12-lead electrocardiogram deriving procedure according to the embodiment of the present invention.

FIG. 6 is a block diagram showing a configuration of an exemplary embodiment of the present invention.

FIG. 7 is an explanatory diagram of a standard limb lead and a monopolar limb lead among 12-lead electrocardiograms.

FIG. 8 is an explanatory diagram of the chest lead in the 12-lead electrocardiogram.

FIG. 9 is an explanatory diagram of an electrocardiogram waveform pattern.

[Explanation of symbols]

10, 11 electrodes 20 ECG waveform measurement unit 21 12 Lead Waveform Derivation Unit 22 ECG waveform monitor 23 12-Lead ECG Automatic Analysis Unit 24 Automatic diagnosis section 25 Communication part 26 Internet 27 hospitals

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Toshiyuki Shimizu             3-6-1, Hakata Station, Hakata-ku, Fukuoka City, Fukuoka Prefecture             No. Computer Convenience Co., Ltd.             Within F-term (reference) 4C027 AA02 CC00 FF00 FF01 GG00                       GG05 GG10 GG16 HH02 HH11                       JJ01 KK01

Claims (2)

[Claims] 1. An electrocardiogram waveform measured between electrodes contacted or attached to two locations on a human body of a subject is recorded as time-series data, and the embedding delay time τ is determined using autocorrelation for this time-series data to determine the correlation. The dimension is calculated to determine the embedding dimension d, and a d-dimensional attractor is constructed based on the embedding delay time τ and the embedding dimension d, and the attractor is rotated according to the type of each electrocardiogram to obtain a specified plane. From the projected figure, all 12 including the guidance
A method for deriving an electrocardiogram, which comprises generating a lead electrocardiogram. 2. An electrode for contacting or wearing at two points on a human body of a subject, a means for recording an electrocardiographic waveform between the electrodes as time series data, and the time series data determined by autocorrelation. Means for constructing a d-dimensional attractor based on the embedding delay time τ and the embedding dimension d determined by the calculation of the correlation dimension, and rotating this attractor according to the type of each electrocardiogram,
An electrocardiogram deriving device comprising: means for generating a figure projected on a specified plane; and means for generating a total 12-lead electrocardiogram including the lead based on the projected figure.