Title:
METHOD AND APPARATUS FOR DETECTING CARDIAC ARRHYTHMIAS
United States Patent 3828768


Abstract:
A continuous electrical wave representing the electrical action of a patient's heart is produced (the ECG wave). Frequency components of the ECG wave that lie predominantly below the frequency range of the normal QRS complex are sensed, these frequency components being typically in the range of about two Hertz to about eight Hertz. These frequency components are integrated upon the occurrence of a QRS complex, and the magnitude of the result is then compared to a reference standard. If the magnitude of the integral exceeds the reference standard an output signal indicative of a cardiac arrhythmia is then produced.



Inventors:
DOUGLAS D
Application Number:
05/271373
Publication Date:
08/13/1974
Filing Date:
07/13/1972
Assignee:
PHYSIOLOGICAL ELECTRONICS CORP,US
Primary Class:
International Classes:
A61B5/0468; (IPC1-7): A61B5/04
Field of Search:
128/2
View Patent Images:
US Patent References:
3698386CARDIAC RHYTHM COMPUTER DEVICE1972-10-17Fried
3616790N/A1971-11-02Harris
3524442ARRHYTHMIA DETECTOR AND METHOD1970-08-18Horth
3352300Cardiac monitor1967-11-14Rose



Primary Examiner:
Kamm, William E.
Attorney, Agent or Firm:
Beehler, Vernon D.
Claims:
I claim

1. The method of analyzing a continuous electrical wave representing the electrical action of a patient's heart in order to detect cardiac arrhythmias, comprising the steps of:

2. The method of claim 1 wherein said high frequency filter is selected to be predominantly responsive to frequencies in the range from 10 to 18 Hertz.

3. The method of claim 1 wherein said low frequency filter is selected to be predominatly responsive to frequencies in the range from 2 to 8 Hertz.

4. The method of claim 1 wherein said step of producing a timing signal includes continuing said timing signal for about 50 milliseconds after discontinuance of an output from said high frequency filter.

5. The method of claim 1 which additionally includes the prior step of first selecting a particular value of said reference standard which is applicable to the particular patient whose cardiac action is to be monitored.

6. The method of claim 1 wherein said step of producing a timing signal includes continuing said timing signal so long as an output in being produced by said high frequency filter and for at least several milliseconds thereafter.

7. The method of analyzing the ECG wave of a patient comprising the steps of:

8. The method of claim 7 which includes the further step of maintaining said timing signal so long as said high frequency components exist and for about 50 milliseconds thereafter.

9. The method of claim 7 wherein said reference standard is previously selected for the particular patient.

10. The method of analyzing the ECG wave of a patient, comprising the steps of:

11. The method claimed in claim 10 wherein the resulting integral is compared to a reference standard.

12. The method claimed in claim 11 wherein said reference standard is previously established to reflect the normal heart action of the patient, and an output signal indicating a cardiac arrhythmia is produced whenever said resulting integral exceeds said standard.

13. The method claimed in claim 12 wherein said reference standard is selected to be significantly greater than the minimum integral of said low frequencies which is produced during normal heart action of the patient.

14. The method of analyzing the ECG wave of a patient, comprising the steps of:

15. The method claimed in claim 14 which includes the further steps of:

16. The method claimed in claim 15 which includes the additional step of maintaining said timing signal for about fifty milliseconds after said high frequency components cease to exist.

17. The method claimed in claim 14 wherein said reference standard is established by:

18. The method claimed in claim 17 which includes the additional step of visually displaying said sample set of QRS complexes before selecting said reference standard.

19. Apparatus for detecting cardiac arrhythmias comprising, in combination:

20. Apparatus as claimed in claim 19 wherein said timing means includes:

21. Apparatus as claimed in claim 20 wherein said generating means is adapted to generate said timing signal throughout the existence of said high frequency components and for at least several milliseconds thereafter.

Description:
BACKGROUND OF THE INVENTION

In the field of cardiology it is well known that the muscular action of the heart is controlled by a continuously changing electrical potential. It is a long-established technique to attach electrodes to the body of a patient to thereby obtain a continuous electrical wave representing the electrical action of a patient's heart. This continuous electrical wave (the ECG wave) is then conventionally graphically recorded on a strip-chart to provide the conventional electrocardiogram, or it may be momentarily displayed on an oscilloscope (the cardioscope) for purpose of visual examination of the wave form by a doctor or a specially trained nurse.

Each recurrence of the normal heart beat corresponds with characteristic points on the ECG wave which are conventionally designated by the letters P, Q, R, S, and T. The Q, R and S portions of the wave when taken together are referred to as the QRS complex, and the occurrence of the QRS complex indicates that ventricular depolarization has taken place in the patient'heart.

A number of analytical methods have heretofore been developed for determining, in accordance with a preconceived method or formula, that the QRS complex has occurred. However, to differentiate between a QRS complex which corresponds to a normal heart action, and a QRS complex which corresponds to abnormal heart action, has proven a far more difficult problem.

Many different types of abnormal heart actions have been identified and studied, and it is known that some are relatively harmless while others indicate a fairly immediate threat to the life of the patient. The cardiac arrhythmias include certain ventricular ectopic beats or contractions, and more particularly those known as ventricular tachycardia, coupled beats, ventricular premature complex (VPC), and fusion beats. It is known that all of these abnormalities indicate a probability that, if the patient is not properly treated, his heart may soon cease to pump blood because of ventricular fibrillation.

Thus it is of extreme importance to be able to analyze the ECG wave in accordance with a predetermined method or formula in order to determine whether the heart action falls within one of the highly dangerous categories mentioned above. This approach to the problem is particularly applicable to the coronary care units of hospitals, which have been established in rapidly increasing numbers since the year 1962. These coronary care units are devoted exclusively to the care and treatment of patients who have had one or more heart attacks. The patient himself is unable to determine whether his heart action is characterized by a highly dangerous type of abnormality. Furthermore, it is often the case that there will be many normal heart beats, then a single abnormal one, and then many more normal beats thereafter. In other instances the abnormal beats occur with some regularity and a significant frequency. The patient may have a heart beat rate of 60 per minute, and if the highly dangerous beats are occurring at the rate of one per minute or less then treatment on an emergency basis may not be required while if the dangerously abnormal beats are occurring at the rate of five or more per minute then this may indicate an emergency situation.

Thus the object and purpose of the present invention is to provide a method or formula for analyzing the ECG wave in a predetermined manner, in order to identify the dangerously abnormal type of heart beats.

REFERENCE TO PRIOR ART

Among the applicable prior art is U.S. Pat. No. 3,138,151 entitled "Detector and Alarm Ventricular Impulses" and issued June 23, 1964.

Also included among the prior art is German Pat. No. 2,109,179 issued Oct. 14, 1971. The German patent discloses the method of analyzing an ECG wave into a relatively high frequency band and a relatively low frequency band, and then comparing the energy levels in the two frequency bands in order to detect abnormality of the heart function.

SUMMARY OF THE INVENTION

The present invention is based upon my study and analysis of the ECG waves of many patients. From this extensive study I have drawn certain conclusions, and the conclusions in turn have been used to construct a method or formula for analysis of the ECG wave. The conclusions will be stated first, and the method will be summarized thereafter.

My conclusions which form the basis for the present invention are as follows:

1. All of the ventricular ectopic beats (ventricular tachycardia, coupled beats, VPC, and fusion beats) are characterized by a significant energy content at the low frequencies which lie on the low end of, or entirely below, the frequency range of the normal QRS complex.

2. All of the ventricular ectopic beats are characterized by a QRS complex which is of longer time duration than a normal QRS complex for the same patient.

3. The frequency components in the ECG wave which reliably indicate the occurrence of a QRS complex are significantly different for heart patients taken as a group, than for people of excellent health. For example, the heart beats of astronauts (assumed to be persons of excellent health) may be most reliably detected by monitoring frequencies centered at 30 Hertz. By contrast, I have concluded that the occurrence of a QRS complex (whether normal or abnormal) in a heart patient is most reliably indicated by frequency components centered at 14 Hertz, and preferably in the range of about 10 Hertz to about 18 Hertz.

4. I have concluded that the low frequency components of the QRS complex which indicate that the heart beat is of the highly dangerous variety, i.e., a ventricular ectopic beat, are centered at about 4 Hertz and typically in the range of about 2 Hertz to about 8 Hertz.

5. In the abnormal QRS complex the measurable duration of the low frequency components is greater than that of the higher frequency components.

According to my invention the occurrence of a cardiac arrhythmia is detected in the following manner. The ECG wave is produced in conventional fashion. The ECG wave is sensed in any appropriate manner to determine the occurrence of a QRS complex, whether normal or abnormal. At the same time, the ECG wave is independently sensed for those frequency components which lie predominantly below the frequency range of the normal QRS complex. These independently-sensed low-frequency components are then integrated throughout a time period which commences upon detection of a QRS complex and extends at least throughout its time duration. The magnitude of the integral is then compared to a reference standard, and if it exceeds the standard the conclusion will be drawn that a cardiac arrhythmia has occurred.

According to another phase of my invention I have determined that the reference standard referred to above is not a constant which can be used indiscriminately for every patient in a group of patients, but rather, that its value will differ significantly from one patient to the next. My invention also includes a method of determining the reference standard applicable to a particular patient.

DRAWING SUMMARY

FIG. 1 is a typical electrocardiogram of a heart patient having a mixture of normal and abnormal beats;

FIG. 2 is a schematic block diagram of an apparatus suitable for carrying out the method of my invention;

FIG. 3 is a waveform diagram of bigeminy (alternating normal and abnormal) showing how the various wave forms are related in accordance with my invention;

FIG. 4 is a more detailed schematic diagram of the apparatus shown in FIG. 2;

FIG. 4(a) and 4(b) show the pass bands of filters used in the circuit of FIG. 4;

FIG. 4(c) shows a timing pulse; and

FIG. 5 is a still more detailed schematic diagram of certain portions of the circuit of FIG. 4.

PREFERRED EMBODIMENT

In FIG. 1 there is shown the electrocardiogram of a heart patient characterized by a mixture of normal and abnormal beats. Specifically, two normal beats spaced at a proper time interval are followed by one abnormal beat which is premature, and hence follows much too closely after the second one of the normal beats. Different characteristic portions of the wave forms are identified by the conventional symbols P, Q, R, S, and T. FIG. 1 is indicative in a general way of the problem to be solved. No effort is made here to illustrate the wave forms corresponding to the numerous different types of ECG abnormalities.

FIG. 2 illustrates an apparatus suitable for carrying out the method of the present invention. An input means is used to produce a continuous electrical wave representing the electrical action of a patient's heart (the ECG wave), and in FIG. 2 this is indicated simply by the letters ECG. The ECG wave is applied both to a QRS Detector, a high bandpass filter, and to a V.E. Filter, a low bandpass filter. The letters "V. E." in the diagram refer to ventricular ectopic. The output of the QRS Detector is fed to a Timer. The main function of the Timer is to measure out a time period which at least equals, and perhaps exceeds, the time duration of the QRS complex, and for this purpose it produces an output identified as "Timing Pulse."

The output of the V. E. Filter is fed to an Integrator. The output of the Timer is also fed to the Integrator, enabling the Timer to control the Integrator so that integration of frequency components passed through the V. E. Filter will continue at least throughout the duration of the QRS complex. The output of the Integrator is fed to a Comparator, to which a Reference Standard is also supplied, and the Comparator produces an output indicating a potentially life-threatening cardiac arrhythmia and which is accordingly marked as "LTCA pulse."

At the usual heart beat rate of about 60 per minute the heart beats are occurring regularly once each second. For a normal heart beat the time duration of the QRS complex is about 50 milliseconds to about 80 milliseconds. For a ventricular ectopic beat the time duration of the QRS complex is always prolonged, and will be about 120 milliseconds to about 200 milliseconds. The relatively low frequency portion of the abnormal QRS complex tends to have a longer time duration than the relatively high frequency portion.

In carrying out my invention I prefer to construct the Timer so that it will continue to operate at least several milliseconds beyond the duration of the QRS complex as detected by the QRS Detector. In this manner I am able to have the Integrator pick up essentially the entire energy content of the low-frequency components that are sensed through the V. E. Filter, that is, the frequency components that lie predominantly below the normal frequency range of a normal QRS complex and are preferably in the frequency range of about 2 Hertz to about 8 Hertz.

Reference is now made to FIG. 3 illustrating the relationship of actual wave forms which occur during the operation of my invention. At the top of FIG. 3 there is shown a prerecorded ECG wave which was used for purpose of the test. Next, below the ECG wave there is shown the Integrator Output which resulted. These wave forms are actual tracings taken from a test which I made, using the method of my invention, on Thursday, Apr. 20, 1972. At the bottom of FIG. 3 there is shown the Comparator Output.

It will be seen from FIG. 3 that for each occurrence of a normal QRS complex in the ECG wave the Integrator Output peaks at a relatively small value. However, for each abnormal QRS complex the Integrator peaks at a much higher value. The Comparator Output produces a square pulse each time that the Integrator peaks at the relatively high value. Therefore, each output pulse from the Comparator indicates the occurrence of an abnormal QRS complex.

Reference is now made to FIG. 4 which schematically illustrates in somewhat greater detail the presently preferred circuit for carrying out the method of the present invention.

In FIG. 4 it is indicated that an input signal is taken from the patient and fed to a pre-amplifier in order to provide a standard 1 volt signal, and that an input signal can also be taken from a monitor. Only one of these inputs is used. Whichever input is used, an input attenuation control (not shown) may be adjusted to produce a nominal one volt signal which refers to the vertical height (either positive or negative) of the R wave from its base line.

As shown in FIG. 4 the ECG wave feeds both the Ventricular Ectopic Filter and the QRS Detector, the same as shown in FIG. 2. An added feature is that each of these circuits feeds both into and around an accompanying Inverter followed by a Summing Network having a summing point S1 or S2 also. Each Summing Network includes a full-wave rectifier. The reason for use of these Inverters is that sometimes the electrodes are connected to the patient in the wrong polarity, and also in some instances the electrical polarization of the heart occurs in the proper manner but with inverted polarity. The method of my invention is not concerned with detecting the polarity; therefore, the use of an inverter in conjunction with each circuit serves to produce the proper form and magnitude of the wave regardless of polarity. That is what is desired for the purposes of my invention.

The QRS Detector is, in general, responsive to frequencies in the range from 10 to 30 Hertz. FIG. 4a illustrates a preferred pass band for the QRS detector. As there shown, a filter is peaked at 14 Hertz and has a smaller receptivity to the low frequency of 10 Hertz and the high frequency of 18 Hertz. I have found that sensing the ECG wave for frequency components in this frequency range provides a reliable indication of the occurrence of a QRS complex, regardless of whether the QRS complex is normal or abnormal. In other words, this range of frequencies does not so much characterize the normal QRS complex but rather is common to both normal and abnormal. Furthermore, the QRS filter will not pass any significant amount of the P and T waves, because of their lower frequency content.

FIG. 4B illustrates the preferred frequency range for the Ventricular Ectopic Filter. As shown in the drawing this filter is peaked at 4 Hertz and has a smaller energy reception at 2 Hertz at the low frequency end and at 8 Hertz on the high frequency end. I have found this frequency range to be optimum for indicating the Ventricular Ectopic form of the QRS complex.

In response to each QRS complex the Timer circuit produces a Timing Pulse which is approximately fifty milliseconds longer than the duration of those frequencies passed by the QRS filter.

More specifically, the QRS filter is used to sense the frequency components of the ECG wave in the range of 10 Hertz to 18 Hertz, and the continuance of these frequencies at a significant energy level is considered as representing the continuation of the QRS complex itself. When these frequency components drop below their significant level the timing pulse generated by the Timer nevertheless continues for approximately fifty milliseconds. It is this timing pulse which keeps the Integrator turned on. A typical shape of the Timing Pulse is shown in FIG. 4(c).

Therefore, after the frequencies in the range of about 10 Hertz to about 18 Hertz have fallen below a significant value, the frequencies in the lower range of about 2 Hertz to about 8 Hertz which are passed through the V. E. Filter continue to be integrated by the Integrator. The result of this integrating process is a voltage magnitude which is fed to the Comparator throughout the duration of Timing Pulse.

As also shown in FIG. 4 a standard reference voltage is supplied to the Comparator. If the magnitude of the resulting integral produced by the Integrator output exceeds the reference standard, then the Comparator produces an output pulse in the manner illustrated in FIG. 3. However, if the magnitude of the integral is less than the reference standard then no output pulse is produced.

As shown in FIG. 4 the Timer output is also applied to a QRS Driver which applies the QRS Pulse to a White Light. There are at least two reasons why a separate QRS output is useful. One reason is that this output signal, indicated as for example by the flashing of the light bulb, demonstrates that the machine is working and that the patient's heart is working also. Another reason for this output is that an irregular, slow, or rapid flashing of the QRS indicator light would indicate heart rate, premature beats and other information available from a timed relation of successive QRS complexes.

In the circuit of FIG. 4 the LTCA Pulse from the Comparator output is fed to a monostable multivibrator having an "on" cycle of two hundred fifty milliseconds and which is identified on the drawings as "250 MS MONO." This device is used simply to impose a 1/4 second time standardization in the V. E. output. The monostable multivibrator drives a V. E. Driver which is coupled to both a horn and a red light bulb which represent parallel loads for the LTCA output signal. Whenever the comparator produces an output pulse the red bulb will light up and the horn will sound for 1/4 second.

According to the present invention it is preferred to establish the value of the Reference Standard voltage for each individual patient. The reason is that the characteristics of the heart action differ significantly from one individual to the next, and therefore a reference standard established for the population as a whole or for a particular group of people would not be entirely accurate as applied to a particular individual. I have found that it is most advantageous to establish the reference standard for an individual patient by empirical means. For example, the apparatus as shown in FIG. 4 may be connected to the patient, with the ECG wave being also coupled to an oscilloscope to be visually displayed at the same time. I prefer to initially set the reference standard voltage which is applied to the comparator so that it will be too low. The result of this procedure is that all of the heart beats will produce a V. E. output indication. A trained person is observing the visual display of the ECG wave and knows that the output indications are false. Then the Reference Standard voltage is adjusted to a higher value until some of the output indications disappear. The Reference Standard is then higher than the minimum value of the integral produced by the Integrator. The Reference Standard is increased further until all V. E. output signals disappear during normal heart action. This is one of several methods of achieving standardization. In general the approach is that the Reference Standard will be set to a value significantly greater than the largest integral from a sample set.

My novel method of detecting cardiac arrhythmias is believed to be 100 percent reliable (not considering some artifacts), where the reference standard has been empirically selected for the individual patient in the manner outlined above. The ratio of integrator output of a clearly abnormal wave form as compared to a clearly normal wave form is about two to one, and this difference is quite adequate for reliable operation of the instrumentation.

DETAILED CIRCUIT

Reference is now made to FIG. 5 which illustrates in detail the main portions of the circuit of FIG. 4. Included in the circuit of FIG. 5 are eight operational amplifiers, each one of which is constructed as an integrated circuit, and these devices are identified as IC1 . . . IC8, respectively.

Referring briefly back to FIG. 4, it will be noted that the drawing indicates the location in the circuit of each one of these operational amplifiers. For example, IC1 and IC2 are contained in the QRS Filter, while IC8 is contained in the Comparator.

Referring again to FIG. 5, suitable filter sections are combined with the two operational amplifiers, IC1 and IC2, so as to provide the overall characteristics of amplification and band pass which are desired. Thus, C1 and the equivalent resistance looking into R1 form one pole of a high-pass filter section. R1 and C2 form one pole of a low-pass filter section. R1 plus R2 and C3, modified somewhat by R3, form a second pole of the low-pass filter. C4 and R5 form one pole of another high-pass filter section. C5 together with the equivalent resistance provided by the combination of R4, R6, and R31 provides another low-pass filter section. The gain characteristic is essentially as shown in FIG. 4(a).

The output of the QRS Filter passes through resistor R7 to an Inverter consisting of IC3 and resistor R8. The Inverter output passes through a capacitor C7 and diode D2 to a Summing Point S2. The direct output of the QRS Filter also passes through a capacitor C6 and a diode D1 to the same summing point. The voltage at the summing point S2 appears across a load consisting of the parallel combination of capacitor C8 and resistor R11. It will be noted that the summing circuit is also a full-wave rectifier. It will also be noted that capacitor C8 provides a filtering function, to filter out the ripple voltage in the fully rectified wave, and this separate and distinct function has been indicated in FIG. 4 by the separate box entitled "Ripple Filter."

The Timer includes operational amplifier IC4. A fixed reference voltage is developed from a source of minus 10 volts through a voltage divider including resistors R12 and R14, and is coupled through resistor R13 to the inverting input of the operational amplifier. The output of the QRS Filter, after passing through the full-wave rectifier, summing circuit, and ripple filter as previously described, is then applied to the non-inverting input of amplifier IC4. The circuit functions in this manner. When the level of the applied signal exceeds the fixed reference voltage, an output is produced at the output terminal of the amplifier. When the level of the applied signal does not exceed the reference voltage, no output is produced from the amplifier. The fixed reference voltage is so selected as to avoid a response to noise or spurious signals, but to provide a response to the signals in the frequency band of about 10 to about 18 Hertz so long as they have significant value.

In the Timer a capacitor C9 is coupled between the noninverting input terminal of IC4 and its output terminal. This provides a regenerative feedback loop. Once an output signal has been developed by the amplifier, capacitor C9 tends to maintain that output signal, and delay its discontinuance or shut-off. Therefore, when the input signal falls below the level of the fixed reference voltage, the output signal from the operational amplifier does not shut off until some time afterwards. This time delay is approximately 50 milliseconds. The typical configuration of this timing pulse, or T Pulse, is shown in FIG. 4(c).

The T Pulse is applied to resistor R25 to the gate of FET1, which controls the action of the Integrator.

The V. E. Filter includes operational amplifier IC5. A high-pass filter section is formed by capacitor C10 in series with the equivalent resistance provided by resistors R16, R17, R32, and R18. A low-pass filter section is provided by R16 and C11. Another low-pass filter section is formed by capacitor C12 together with the equivalent resistance of the circuit. The over-all circuit provides the desired amount of gain, and a frequency response essentially as shown in FIG. 4(b).

The output of the V. E. Filter is supplied through a capacitor C14 and a diode D4 to a summing point S1 above resistor R23. It is also supplied through a resistor R19 to an Invertor which includes IC6 and R20. The invertor output is supplied through C13 and D5 to the same summing point. As before, the summing circuit is also a full-wave rectifier. No filtering is provided in this portion of the circuit, however.

The Integrator includes operational amplifier IC7, whose non-inverting input terminal is grounded. The output of the summing circuit that follows the V. E. Filter is supplied through a resistor R24 to the inverting input terminal of IC7. A capacitor C15 is connected between the inverting input of IC7 and its output terminal. The drain and source of field effect transistor FET 1 are connected across C15. When a QRS pulse is not being received, the impedance between drain and source of FET 1 is low, C15 is essentially shorted out, and IC7 does not integrate.

When timing pulse T is applied, however, FET 1 is turned off, effectively putting C15 back in the circuit. The series combination of R24 and C15 then provide an integrator whose output appears at the inverting input of IC7. The output of IC7 therefore produces the integral of the applied signal (from V. E. Filter) throughout the continuation of the timing pulse T.

The Integrator output goes to the non-inverting input of Comparator IC8. The Reference Standard (selected as earlier described) is applied to the inverting input. A positive-going LTCA pulse is produced on the output terminal of IC8 whenever the Integrator output exceeds the Reference Standard.

As will be understood by those skilled in the art, what has been described are preferred embodiments in which modifications and changes may be made without departing from the spirit and scope of the accompanying claims.