JPS5967932A - Detection of original point in wave form recognition of ele-ctrocardiograph - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for detecting an origin in on-line waveform recognition of an electrocardiograph for an electrocardiogram automatic analyzer used in heart disease testing.

In order to diagnose diseases such as bundle branch block, first degree atrioventricular block, WPW, and myocardial infarction, the electrocardiogram automatic analyzer uses the origin (QR3I!), which is the basis of all waveform recognition.
Time (interval) from J starting point)...QR3! i time・
The PQ interval, the interval from the origin to the first wave of the QR3 axis, etc. are used. Therefore, by accurately determining the origin, it is possible to accurately diagnose the above-mentioned bundle branch block, WPW, first degree atrioventricular block, etc.

Regarding the conventional electrocardiogram automatic analysis device, Figures 1 and 3
This will be explained below with reference to the drawings.

The electrocardiogram automatic analysis device includes a ROM 2 for storing a control program, a RAM 3 for storing data that can be changed depending on the application, and sequential control and calculation according to the contents of the ROM 2 as a program memory. an interrupt timer 4 that issues an interrupt signal to the cpU 1 at regular intervals according to a program;
It has an input/output circuit 5 that interfaces the CPU 1 with an external circuit, and includes the CPU, ROM 2, and RAM.
3. The interrupt timer 4 and the input/output circuit 5 are connected by a bus 6 consisting of an address bus, a data bus, a control bus, etc., respectively. In addition, the input/output circuit 5
includes an A-D converter 7 for converting analog signals from the electrocardiograph 10 into digital signals, a CRT 8 for displaying messages and data, and a keyboard 9 for inputting commands and data. It is connected.

The operation of the electrocardiogram automatic analysis device shown in FIG. 1 will be explained below with reference to the system program shown in FIG.

When the power is turned on, the program in the ROMZ automatically starts (101) and proceeds to step (102).

In step 102, questions (for example, examiner number: ■D, age, date, gender, etc.) are displayed on the CRT 8,
When the operator inputs data for the above-mentioned questions from the keyboard 9, the process proceeds to step 103. Step 10
In step 3, the electrocardiograph 10 and interrupt timer 4 are activated by setting the electrocardiograph measurement time and operating the start switch on the keyboard 9, and the process proceeds to step 104. In step 104, a plurality of outputs from the electrocardiograph 10 are sampled and held in synchronization with the interrupt timer 4, and then the sequential A/D converted values are read into the RAM 3 as data DA, and the process proceeds to step 105. In step 105, when the data acquisition for all the leads is completed, the process proceeds to the next step 106.

Furthermore, if the data acquisition for all the leads has not been completed, the process advances to step 103, and steps 103, 104 and 105 are looped until the data for all the leads are acquired.

In step 106, the moving average method, divided average method, weighted average method, etc. are used for each lead data: DA to obtain data: DB from which noise has been removed, and the data is transferred to the above-mentioned RA.
Store it in M3 and proceed to step E07. Step 107
Now, after confirming the approximate position of the QR3 wave shown in Figure 3 by keeping the data: DB constant and slicing it with Hell, we will calculate a predetermined correctable range based on the position of this QR3 wave. After storing data f(x) whose baseline has been corrected by linear approximation, the process advances to step 108. Data with baseline correction: f(xl) can be expressed as a graph as shown in Figure 3 with the address being X. In other words, the data f(xl)
(This is a collection of collected xi. In step 10B, after detecting the general division of each pattern from the position of the three QR waves, detailed patterns are detected, and the process proceeds to step E09.

In step 109, the waveform pattern divisions detected in step 108 are used to measure the peak value, waveform width, waveform shape, time interval, etc. of these waveforms, and the process proceeds to step 110. In step 110, if the measurement of all the leads has been completed, the process proceeds to step 111, and steps o6 to 110 are looped until the measurement of all the leads is completed, and the process proceeds to step 111, in which the measurement of all the leads is completed. In step 111, the measurement result is used in diagnostic logic to obtain a diagnostic result, and the process proceeds to step E12. At step 112, the diagnostic results are output to the CRT 8 and printer 1, and the program proceeds to step 113, ending the program.

By the way, as a conventional method for detecting the origin, there is a method described on pages 484 to 486 of the "New Edition Computer Handbook" published by Corona Publishing Co., Ltd. on May 15, 1976, 7th edition (new edition).
It is explained in detail in the section ``Electrocardiogram automatic diagnosis system''.

This method involves slicing the analog output of the electrocardiograph using a comparator, detecting the approximate divisions from the positions of the QR3 waves, and then determining the divisions of each waveform pattern by taking into consideration the polarity of each waveform. It has been shown that positive discrimination is possible.

FIG. 5 is a flowchart showing a method of origin detection previously carried out by the inventor of the present invention. This origin detection method is a method of detecting the search point X as the origin when there are five consecutive comparison points within a set error range with a difference of comparison point C from the search point X. The program will be explained in detail.

Sampling frequency: 20011z (5m5e) is 2n times the commercial frequency (n = 1.2.3...)
c), the output of the electrocardiograph 10 is subjected to A-D conversion, and the input data becomes converted data f fXl after removing noise and correcting the baseline. The data f (xl
An example is shown in FIG.

FIG. 4A is an example of an electrocardiographic waveform of normal QRS fi. Further, FIG. 4B is an example of an electrocardiographic waveform of the Q RS axis having a delta wave showing signs of WPW.

After a temporary origin is detected in advance by a pattern recognition program by comparison at a slice point determined according to the maximum value of the R wave, the origin detection program shown in FIG. 5 is executed.

The origin detection program starts at step 201 and proceeds to step 202. In step 202, ■ The address number (data number) of the temporary origin obtained in advance is sent, and ■ The search starting point i returned from the temporary origin m is
(However, i = m-2), ■ To determine the search range, 130 m5 from the search starting point i
ec Set the search end point n (however, n = 1-26) returned, ■ Cent the search start point i to initialize the address counter X consisting of a hard counter or soft counter corresponding to the search point, ■ Search point A counter C, which is a hard counter or a soft counter corresponding to the number C between X and the comparison point, is set to 1 to initialize it, and the process proceeds to step 203. In step 203, the value of address counter X (
In this case, the magnitude of x=i) and the value of the final search point n is determined, and if X is greater than or equal to n, the process proceeds to step 204, and if X is smaller than n, the process proceeds to step 300.

In step 204, the value of the counter C corresponding to the score of the search point If it is larger, proceed to step 209.

In step 205, C is a number from 1 to 5 (FfX)
= f (X) - f (x - C) ... Perform the calculation of formula 11), and calculate the voltage determined by the absolute value of the calculation result F (X) of formula (1) and the removal of noise such as gentle noise and ripple. Number of digits of A-D converter corresponding to range: A
(However, in this case, an 8-bit A-D converter is used to
= 1), and if the absolute value I F (Xi I is within A digits corresponding to the voltage range, step 20
6, and if the absolute value I F (Xl 1 is greater than the A digit corresponding to the voltage range, step 2
Proceed to 07.

In step 206, in order to update the comparison point, 1 is added to the counter C corresponding to the score between the search point X and the comparison point (at this time, C becomes 2), and the process returns to step 204.

At this time, since the counter C is 2, step 204
The process then proceeds to step 205. Therefore, if the absolute value IF of the calculation result (X11 is within the number of digits corresponding to the voltage range 5 times in a row = A), step 204
205 and 206 are looped five times, and the value of the counter C becomes six. Therefore, the value of the counter C becomes greater than 5, and the process proceeds from step 204 to step 209.
The value of the address counter X corresponding to the search point is stored in the memory at address B or register B, and step 2
The program proceeds to step 10 and ends.

Further, in step 207, the counter is set to 1 again to initialize it, and the process proceeds to step 208. In step 208, in order to update the address counter X and proceed to the next search point, 1 is subtracted from the counter X, and step 2
Proceed to 03. In this way, the absolute value IF of the calculation result
(Number of digits where Xi l corresponds to the voltage range: A
If it does not fall within the range 5 times in a row, step 203
・204・205・207・208・203 loop or stem 203・204・(205・206・2
04) Through the loop of 205, 207, 208, and 203, the value n of the search end point is thicker than the value of the counter
Proceed to. In step 300, an error message is displayed to indicate that the origin was not found, and the program proceeds to step 210 and ends.

However, in this method, the absolute value I F (Xl l of the calculation result of the above equation (1) is successively found within the number of digits corresponding to the voltage range of the error setting: A, and the search point at that time is set as the origin. Therefore, in the case of an abnormal waveform (Figure 4B), the origin may be mistaken and the origin may be mistakenly detected as the origin at the top of the P wave, or the origin may not be detected. Furthermore, even when using the comparator to find the origin using hardware, it is difficult to set the slice point when there is a lot of noise or an abnormal waveform, making it impossible to accurately detect the origin. .

An object of the present invention is to find the first inflection point from the direction of the slope of the next point of the search end point when the above method cannot be used to find the origin, and to find the first inflection point. By adding a method using the origin as the origin, the present invention attempts to provide a method that can accurately detect the origin even when an abnormal waveform such as wpw or myocardial infarction is shown.

In the method of the present invention, instead of the error display 300 in FIG.
The feature is that a routine is added to find the first point of inflection from the direction of the slope of the point next to the search end point, and the point of inflection is detected as the origin, which will be explained below using examples.

FIG. 6 is a flowchart showing an embodiment of the present invention in which an inflection point detection routine is added in place of the error display step 300 of FIG.

That is, in this flowchart, steps 301 to 305 are applied instead of step 300 in FIG.
In addition, the initial setting of the search end point n in step 201 is set to a point 4Q msec (in this case, n is 1-8) back from the search start point i.

Thus, when the program shown in the flowchart is executed and the origin is not detected in steps 203 to 208, the value n of the search end point becomes larger than the value of the counter X for the search point, and step 203 The process then proceeds to step 301.

In step 301, in order to determine whether the direction of the slope of the search end point is upward to the right or upward to the left, data f (Xl) of address n-1, which has passed by one from the search end point n, and two data returned from the address n-1. Data f at address n-3
(x-2), data f (Xl is data f (
If the data f(X) is larger than the data f(x-2), the process proceeds to step 302, and if the data f(X) is equal to or smaller than the data f(x-2), the process proceeds to step 304. In order to find the direction of the slope with high accuracy, the influence of noise etc. can be reduced by increasing the number of comparison points in step 301 and using the sum of the differences.

In step 302, the address counter X is updated and decremented by 1 in order to proceed to the next search point, and the process proceeds to step 303. In step 303, in order to detect the downward left inflection point,
Compare data f (Xl) at address X with data f (x-1) at address Proceeding to step 209, the value X of the address counter at that time is stored in register B or address B, and this program ends.
x-1), it is determined that there is no inflection point, and the process returns to step 302, updates the address, and loops steps 302 and 303 until an inflection point is found.

In step 304, 1 is added to update the address counter, and the process proceeds to step 305. In step 305, data f(x) at address X and data f(x+1) at address
The data f(x) is compared with the data f(x).
If (x+l) is large, it is assumed that there is an inflection point, and the process proceeds to step 209, where the value X of the address counter at that time is stored in register B, and the program ends. Further, if the data f (Xl) is equal to or smaller than the data f (x+1), it is determined that there is no inflection point and the process returns to step 304;
Update address and step 30 until you find the inflection point
4. Loop 305.

As explained above, in the present invention, for a normal waveform like the one shown in FIG. In addition, for abnormal waveforms caused by WPW, myocardial infarction, etc., by applying a method of detecting the inflection point as the origin from the direction of the slope of the point next to the search end point, it is possible to detect the inflection point as the origin. It is possible to accurately detect the origin that was incorrectly detected or undetectable using the detection method.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an electrocardiogram automatic analysis device. Second
The figure shows the system program for an automatic electrocardiogram analyzer. FIG. 3 is a diagram in which data f (x) subjected to baseline correction is expressed by an address X. FIG. 4A shows a normal electrocardiogram waveform. FIG. 4B shows a dramatic waveform caused by WPW, myocardial infarction, etc. FIG. 5 is a flowchart showing the origin detection program previously carried out by the inventor of the present invention. FIG. 6 is a flowchart showing an origin detection program in an embodiment of the present invention. Explanation of symbols 1...CPU, 2...ROM, 3...RAM,
4... Interrupt timer, 5... Input/output circuit, 7...
A-D converter, 8...CRT, 9...keyboard,
DESCRIPTION OF SYMBOLS 10... Electrocardiograph, 11... Printer, m... Temporary origin, i... Search start point, n... Search end point, X...
Address counter, C... counter.

Claims (2)

[Claims] (1) Remove noise and correct the baseline from the data imported from the electrocardiograph to obtain data f (X), and obtain the data f (
X) Detect the maximum value of the R wave included in the data, compare the slice point determined according to the maximum value of the R wave with the data f( A search start point and a search end point are determined, and the positions of the reference search points are sequentially changed at predetermined intervals between the search start point and the search end point, and each of these reference search points and a predetermined number of the data f (
In the method of detecting the reference search point as the origin when the difference from Xl continuously falls within a predetermined voltage range, if the origin cannot be detected, the data f at the search end point A method for detecting an origin in waveform recognition of an electrocardiograph, characterized in that the inflection point of the search direction is determined, and the inflection point is detected as the origin. (2) Determine whether the difference between each of the reference search points and the predetermined number of data f (Xl) that follows it is within the voltage range, and Determine the slope of f ( Claim 1: When the slope is upward to the left, the inflection point is determined by the following equation (4).
The method for detecting the origin in waveform recognition of an electrocardiograph as described in . If(xl −f (x−C)l≦A... (1
)f(x) > f(x-2)...
・(2) f(x) < f (x-1)
−(3) f(×)〈 f(xl1)
... (4) However, A is a predetermined voltage range, C is an integer of 1, 2, 3, etc., and X is a search point.