CA1123919A - Method and apparatus for monitoring electrocardiographic waveforms - Google Patents
Method and apparatus for monitoring electrocardiographic waveformsInfo
- Publication number
- CA1123919A CA1123919A CA322,460A CA322460A CA1123919A CA 1123919 A CA1123919 A CA 1123919A CA 322460 A CA322460 A CA 322460A CA 1123919 A CA1123919 A CA 1123919A
- Authority
- CA
- Canada
- Prior art keywords
- measure
- ecg
- signal
- width
- qrs complex
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
- A61B5/366—Detecting abnormal QRS complex, e.g. widening
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
- A61B5/364—Detecting abnormal ECG interval, e.g. extrasystoles, ectopic heartbeats
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
Abstract
ABSTRACT OF THE DISCLOSURE
Apparatus for the recognition of ventricular premature beats by an identifying width of a QRS complex by dividing the area by the height and further noting when the width measurement is at least 40 percent wider than the average width of the immediately preceding several QRS com-plexes. Additional apparatus for indicating ventricular premature beats by determination of an early width determina-tion which exceeds the average width by at least 20 percent and is additionally followed by the compensatory pause as determined by interbeat interval measurement.
Apparatus for the recognition of ventricular premature beats by an identifying width of a QRS complex by dividing the area by the height and further noting when the width measurement is at least 40 percent wider than the average width of the immediately preceding several QRS com-plexes. Additional apparatus for indicating ventricular premature beats by determination of an early width determina-tion which exceeds the average width by at least 20 percent and is additionally followed by the compensatory pause as determined by interbeat interval measurement.
Description
~ 23~
M~THOD AMD APPARATUS FOR MONITORING
E~ECTROCARDIOGRAPHIC WA~EFORMS
The invention relates to electrocardiographic monitoring systems which detect abnormal ECG waveforms.
In the field of cardiology, it is common to continuously monitor the ECG signal of a patient for analysis. Normally, it is impractical to continuously monitor the ECG waveform and systems have evolved for automatically monitoring and analyzing the waveform.
~he ECG waveform is normally comprised of a series of characteristic points conventionally designated by the letters P, Q, R, S and T. ~he Q, R a~d S portions of the wave when taken together are referred to as the QRS complex. Means for determining ~hat the QRS complex has occurred are commonly referred to as R-wave detectors, a representative example of which being disclosed in U.SO Patent No. 3,490,811 issued ~uly 6, lg71, to ~arris for Electrocardiographic R-Wave Detector. However, it i5 often desirable to further dif~erentiate between a QRS complex which corresponds to a normal heart action, and a QRS complex which corxesponds to abnormal heart action~
One such abnormal heart action of interest is that of ectopic beats. Ectopic bea~ are characterized by departure from a "~ormal" interbeat intexval and~or departure from a "~ormal"
wave~orm morphology. "~ormal" interbeat interval and/or waveæorm morphology for one patient may differ ~rom that which is "Normal" ~or another. Ectopic beats deserving o~ special , 3~
attention include ventxicular premature beats ~VPB), atrial premature beats (APB), and successive groupings of such premature beats.
Ven~ricular premature beats (VPB's) have been identified by monitoring the interval between successive QRS complexes and the width and/or area of such complexes, and signaling the occurrence of such VPB if those parameters differ by more than a predetermined amount from that which is "Normal" for the particular patient.
An example of these techniques is disclosed in some detail in U.S. Patent 3,616,791 issued November 2, 1971, to G. J. Harris for Electrocardiographic Morphology Recognition System. In that patent, certain QRS complexes are identified as VPB's if their width is greater than normal. That determination of width i~
made by measuring th`e area under the rectified ECG signal and comparing it with the average area That area determination is made by integrating the rectified waveform. Be~ore actually indicating that a "wide" QRS complex is actually a VPB, the interval between QRS complexes is normally required to exhibit a so-called compensatory pause. The QRS complex of a VPB will usually occur earlier than expected and the next succeeding QRS
complex will occur after a longer than no~mal interval or compensatory pause. This combination of a "wide" QRS complex occurring earlier than usual and followed by a compensatory pause is generally a reliable indicator of a VPB.
It would be more preferable to identify a VPB based on width alone without analysis of the interbeat interval. Howe~er, such analysis has not heretofore been relied on, in part because the area, and thus the width measure, of the QRS complex may increase as the height or amplitude of the waveform increases due to changes in patient respiration, drift in the signal baseline and/or other causes.
Accordingly, it is a pxincipal object of the present inven~ion to provide a method and apparatus for analyziny the ECG wave in a manner which simply and accurately identifies VPB's, and to provide improvad mea~s for accurately identifying the occurrence of the QRS complex in an ECG waveform whereby accurate identification of VPB's a~d other heart characteristics may be'made.
The present invention is bas,ed on the recognition of VPB's upon,the width of a particular QRS complex exceeding an average width for such complexes by a predetexmined amount. The de~ermination of wid~h for each QRS complex is done indirectly with a measurement of QRS area normalized to remove variations in waveform amplitude by dividing the area by the height thereof.
Although this width measure does not result in a signal which is identical to ~he width of the complex, it does result in a signal which is proportional to that width.
According to a preferred embodiment o~ the invention, a particula.r hear~beat is designated a premature ventricular contraction if the QRS complex is determined to be at least 40% wider than the average width of the immediately-preceding several QRS complexes. Supplementally, a VPB may also be indicated if the width of a QRS complex is early, exceeds the average'width by at least 20/~ and is additionally ~ollowed by a compensatory pause as determined by interbea~ interval measure-ments~
~3~
The l.CG signal is fed in~o two parallel filters, the outputs of which are rectified and peak de-tected. The bandwidth of one fil-ter is -that of a normal Q~S complex such that its output is proportional to height. The bandwidth of the other filter is below that of the first such tha-t -the filter's impulse response is in turn proportional to the area of the complex.
The peak-detected area is then divided by the peak-detected height to obtain a wid-th rneasure which is compared with an average of preceding width measures.
According to another aspect of the inventionr an improved R-wave detector is provided, which detector exhibits increased accuracy for controlling the peak detection of the height and area measures of a QRS complex. More particularly, a variable threshold is associated with generally conventional R-wave detection means whereby to increase sensitivity to ~ rela-tively small amplitude R-waves and decrease sensitivity to T-waves of relatively large amplitude. Both the VPB and the R-wave detection circuitry adapt to individual patients and to variations over time in the QRS complex of a single patient through use of averaging techniquesO
In summary of the above, therefore, the present invention may be seen as providing in an ECG monitoring system, the method of indirectly providing a measure of the width of a QRS complex appearing in an ECG waveform comprlsing the steps of: generating from the ECG waveform a siynal proportional to the area of the QRS complex; generating from the ECG waveform a slgnal propor-tional to the peak height o~ the QRS complex; and dividing the area signal by the peak amplitude signal to provide a normalized signal representative of the width of the QRS complex of the ECG signal.
B
sd~
: . .
The above method may be carried out by way oE an ECG
monitoring system for generating a signal indicatiny a ventricular premature beat comprising means responsive to an ECG waveform for providing a measure proportional to the area of the QRS
complex of the ECG waveform, means responsive to the ECG wave-form for providing a measure proportional to the peak height of the QRS complex of the ECG waveform, and means for dividing the area measure by the peak height measure thereby to provide a normalized measure proportional to the width of the QRS
complex of the ECG waveform.
Figure 1 depicts a typical ECG of an individual having a mixture of normal and abnormal heartbeats, the uppermost trace being the ECG waveform fox a single normal heartbeat;
Figure 2 is a detailed schematic block diagram of apparatus for practicing the invention;
Flgure 3, appearing on the same sheet as Figure 1, is a more detailed schematic diagram of certain portions of the .
R-wave de.e_tor forming part of the apparatus of Figure 2;
. .......................... ' .
. ' ' ,~
sd~ 4A-~3~3~L9 F'igure 4 is a more detailed schematic diagram of a band-pass filter utilized for its impulse response in ~he apparatus o~
Figure 2, Figure 5 is a flow diagram of a combined hard-wired a~d S computer-based hybrid of the circuitry of Figure 2; and Figuxe 6A and 6B in combination comprise the general flow diagram for the programmed computer-based portion of Figure 5.
In Figure 1 thPre is illustrated in the bottom trace an electrocardiogram ~ECG) ~f a heart patient characterized by a mixture of normal and abnormal beats. The ECG in the uppex trace of Figure 1 represents an enlarged single normal heartbeat of the patient. The different characteristic portions o~ the wave~orm are identified by the conventional signals P, Q, R, S
~ 1 and T. In the lower trace, two normal beats spaced at a proper time interval are followed by cne abnormal beat which is premature and which is followed by another normal beat~ The premature beat ollows the second normal beat more closely and is spaced from the final normal beat by more than the normal interval~ That increased interval is commonly referred to as the compensatory pause.
The ECG monitoring system illustrated in Figure 2 receives the ECG signal of Figure 1 as its principa} input and analyzes that waveform to detect VPB's as well as other types o~ ectopic beats. ~he circuitry of Figure 2 is particularly suited ~or analyzing ECG signals previously recorded, as during ambulatory monitoring; but is similarly applicable to on-line or real-time situations. The playback o~ prev~ously recorded ECG signals may 3~
occur at the same rate at which the si~nals were recorded (lX~, or, at multiples of the speed at which they were recorded,~e.g.
60X, 120X). Speed~up of playback expedites the analysis of the recorded signal. The following discussion of the circuitry of Figure 2 assumes playback at the same speed (lX~ as recorded.
The various time constants associated with the circuitry of Figure 2 will thus have a value for a lX playbacX speed, and alternate volume for 60X and 120Xo The capabi}ity of selectively processing the ECG signal at multiple playback speeds may be accomplished eit~er by having selective channels, each having different appropriate time constants or by a single channel having components with different time constants selectively insertable therein. Thi5 selectivity may be accomplished in a program-controlled, microprocessor-based system~
In Figure 2, the ECG signal, derived from the playback heads of a recorder is connected as an input to low-pass filter 10.
Low-pass filter 10, in addition to comprising the input to the overall analyzer system of Figure 2, is also the first staqe of an R-wave detector, and includes slew-rate limiter 12, band-pass filter 14, full-wave rectifier 16, peak detector 18, variable threshold 20, comparator 21, and non-xetriggerable one-shot 2?
(to be described in greater detail). Low-pass filter 10 is designed to remove 50/60 Hz, line noise (at lX) from the input ECG signal.
The output of low-pass filter 10 is connected to the input of slew-rate limiter 12 (of conventional design) and scaled to suppress pacemaker spikes, noise associated with tape dropout, musc~e artifact and the like. Normally, the signals passed ~z~
by slew-xate limiter 12 will include normal ECG waveforms, VPB's and other ectopic beats and possibly some muscle artifact.
The output of slew-rate lLmiter 12 is input to multistage band-pass filter 14. A first portion of filter 149 shown ~n greater detail as part of Figure 3, is shared by tha R-wave detection circuitry as well as the "area" detector (hereinafter described). The latter portion of filter 14 is utilized only by the R-wave detection circuitry and the waveform height or amplitude detector. The initial or common portion of band-pass filter 14 typically has a band-pass range of .8-8 Hz and the final section comprises a high-pass section which, in combination with the preliminary stages, provides an overall band-pass range to the filtex o 8-32 Hz for lX operation. This range of 8-32 Hz is intended to pass not o~ly nonmal QRS complexes but also those o~ an ectopic nature. The lower frequency T-wave component is attenuated to some extent by filter 14.
The output signal from multistage band-pass filter 14 is input to a full-wave rectifier 16 of conventional design.
Rectifier 16 ensures that all signal excursions to either side of the base line appear on only one side o~ the base line, eliminating need for positive and negative threshold detection as a result o~ the possibly reversed polarities to the input ECG
signal.
The output o~ full-wave rectiEier 16 is input to peak detec~ox 18 and as one o~ the two inputs to comparator 21.
Whereas ~ilter 10, rate limiter 12, band-pass filter 14 and full-wave recti~ier 16 comprise means for "cleaning up" the ECG signal for R-wave detection, peak detector 18, variable ~hreshold 20 ~23~
and comparator ~1 provide the actual means for detecting the ~-wave and one-shot 22 indicates detection. The combination of peak detector 18, variable threshold 20 and comparator 21 with feedhack of the output of compaxator 21 to variable threshold 20 via conduc~or 23 provides novel means for accurately detecting R-waves, even those of r~latively small magnitude as may be associated with ectopic beats, without also responding to T~waves, - particularly those of relatively large magnitude.
Additional detail to the combination of peak detector 18, variable threshold 20 and comparator 21 is illustrated in Figure 3. The output from rectifier 16 of Figure 2 is applied to the noninverting input of an operational amplifier Al via input resistor Rl. The output of amplifier Al is extended to the noninvertlng input of amplifier A2 through the series circuit of a low value resistor R2, a peak detecting diode Dl ancl an input resistor R3. A capacitor Cl and resistor R4 are connected in parallel to ground ~rom the cathode Qf diode Dl. The capacitor Cl is charged by the rectified voltage pulses corresponding with the R-wave as well as other portions of the ECG signal.
The value of resistor R4 in combination with the value of capacitor Cl are selected such that capacitor Cl discharges at a relatively slow rate, for instance, to two-thirds of its value in about five seconds or five heartbeats. The output o amplifier A2 is fed back to the inverting inputs of amplifiers Al and A2. A clamping diode D2 is connected between the :inverting input and the output of amplifier Al.
The voltage appearing across capacitor Cl and subsequently at the output of amplifier A2, and khus also the output of peak ~ 23~
detector 18 of Figure 2, reflects the maximum peak vol~age, or a~ least an average of the peak voltages associated with the QRS complexes, at the input over the preceding interval of several seconds (lX operation). Typically, the R-wave is responsible for the largest peak voltage in the ECG waveform during each heartbeat. Whether it is the largest peak over several heartbeats or a value more nearly to an average of the peaks over that interval appearing at the output, the output is representative of the peak value of one or more QRS
complexes during that inter~al. This output is applied to the input of variable threshold circuit 20 for variable scaling and subsequent application as the reference input to comparator 21.
Although described as a peak detector, circuit 18 may embody the circuits of U.S_ Patent 3,490,811 such that the output is - 15 clearly an avarage of the peaks.
Variable threshold 20 includes a voltage divider comprised of resistors R5 and R6 with one end of R5 being connected t~
the output of peak detector 18 and one end of R6 being connected to ground, the ~unction between R5 and R6 being connected through input resistor R7 to the inverting or reference input of amplifier A3 of comparator 21. The values of resistors R5 and R6 are scaled such that, at steady state, the threshold value applied to comparator 21 is normally about 40% if the peak value provided as the output from peak detector 18. The signal output ~rom rectifier 16 is also led directly to the other input of comparator 21 via lead 24. The signal amplitude on lea~ 24 is compared by amplifier A3 with that provided by variable threshold 20. The normally relatively low voltage at the output of _g _ ~3~3~S~
amplifier A3 goes relatively more positive whenever the rectified ~CG signal exceeds the instantaneous thre~hold value.
This positi~e transi~ion in the voltage at the output o~
ampli~ier A3 is developed across resistor R8 and serves as the S triqger input to non-retriggerable one-shot 22. When one~shot 22 is triggered, it provides a positive pulse of approximately 180 milliseconds duration for signifying the occuxr0nce of an R-wave and keying the operation of a digital sequencer 25.
In addition to triggering one-shot 22, the positive going voltage at the output of comparator 21 is fed back via lead 23 through diode D3 and base resistox R9 to the base of transistor Tl, A capacitor C2 is comlected in parallel across resistor R5 and the emitter and collector of transistor Tl are similarly oonnected ïn parallel across resistor R5 and capacitor C2. Immedia~ely prior to the occurrence of the R-wave i~ a particular ECG wav~form, capacitor C2 will be charged and the threshold voltage appearing at the ~unction of resistors R5 and R6 will be about 40% that of the output of peak detector 18.
~Owever, as soon as the ampli~ude of the present R-wave exceeds the threshold value and evokes a positive response from comparator 21, the positi~e going voltage on lead 23 operates to turn on transistor Tl, shoxt-circuiting capacitor C2 and resistor R5 and applying the value (i.e., 100%) appearing at the output of peak detector 18 as the threshold value to the reference input of comparator 21~ This has the e~fect o~
i~creasing the threshold for any relatively large T-waves which may ~ollow a QRS complex. Stated another way, the variable threshold enables increased sensitivity to small amplitude R-waves without requiring the threshold to remain so low as to additionally detect the following T-waves. As the threshold reference value rapidly ihcxeases to near lOQ% of the output of peak detector 18, it is no longer exceeded by the input appearing on line 24~ Accordingly, the output of compaxator 21 drops to its normal potential, thereby returning transistor Tl to nonconduction. The RC time constant of capacitor C2 and resistor R5 in parallel with resistor R6 is such that recharge of capacitor C2, and thus return of the threshold network to its normal 40% value, occurs continuously over an interval sufficiently long to exclude T-waves but sufficiently short to ant:icipate the next succeeding R-wave. A typical RC time constant is about 200-300 milliseconds at lX operation. Although more di~ficult techically, variable scaling might be applied to the rectifiPd EC~ signal from rectifier 16, except that the function would be inverted, i.e~, the ECG signal would be amplified by 2-1/2 times until R-wave detection whereupon it would return to normal amplitude to "miss" the T-wave then gradually increase to 2-1/2 times to detect the next R-wave.
Returning to the discussion of Figure 2, attention is given another portion of the ECG monitoring system comprising a principal aspect of the invention. The output of full-wave rectifier 16 is additionally extended via line 24' to the input o~ peak detector 26. Peak detector 26, in addition to conventional peak detecting circuitry of a type hereinbe~ore described, also includes hold circuitry ~or output of only the peak rectified signal CQmmensUrate with a QRS complex. A gating or reset signal is provided via line 27 from digital sequencer Z5 to ~.2~9 the reset input associated with the hold portion o peak detector 26. Digital sequencer 25 is of conventiona} design and is driven by a clock ~not shown). The R-wave indication provided by one-shot 22 serves to initiate certain predetermined sequences of control signals provided by se~uencer 25. Accordingly, the reset signal appeaxs on line 27 for only a brief inter~al immediately after an R-wave has been detected to reset t~e hold circuit for permitting detector and hold circuit 26 to respond to the present QRS complex. The output of peak detector 26 is proportional ~o the "height" o~ the QRS complex and is extended to the denominator input of a divider 28.
To obtain a measure of the area of the QRS complex o the input ECG wave~orm, ~he signal is led from an intermediate section of the multistage band-pass filter 14 via lead 0 to an addi~ional band-pass fllter 32. The characteristics of the pra-liminary stage of filter 14 and the band-pass fil~er 32 are - selected such that the impulse response thereof to the QRS complex will provide an accurate measure of the axea of the QRS complex.
Referring to the preliminary stages of multistage band-pass ~ilter 14 in Figure 4, the ECG signal passes through a first filter stage including ampli~ier A4 and a second filte,x stage including ampli-fier A5 to junction J, then lead 30 to ~ilter 32. The input to amplifier A4 is through a series-combination sf capacitor C3 and resistor R9. A parallel combination of capacitor C4 and resistor 2S R10 is connected between the output and the input o~ ampli~ier A4.
A "T" networ.k comprised of series res.istors Rll and R12, and capacitor C5 therebetween to ground, is connected between the output o~ amplifier A4 and the input of amplifier A5~ A parallel -~2-combination of capacitor C6 and resistor R13 is connected between the output and input of amplifier A5. The values of capacitors C3-C6 and resistors R9-R13 are selected to provide a band-pass range of about .8-8 Hz at lX operation, or may be selected to provide a 40 Hz-470Hz passband for 60X operation, etc.
Band-pass filter 32 includes amplifier A6, the series combination of capacitor C7 and resistor R14 in lead 30 to the input thereof, and the parallel combination of resistor R15 and capacitor C8 extending between the output: and input thereof.
The values of these resistances and capacitances are selected such that this filter individually has a passband range of about 2-20 Hz d~ring lX operation, but when considered in conjunction with the earlier sta~es of ~iltering in filter 14, the band-pass between the input to filter 14 and the output of ~r 15 filter 32 is in the range of about 2~8 Hz for lX operation.
The upper end of this passband is lower than most of the instantaneous frequency ch~racteristics of portions of the QRS
complex. For this reason, the output from filter 32 does not accurately follow the QRS complex but instead provides a ringing impulse response therefor. The amplitude or magnitude of the impulse response is proportional to the energy content, and thus the area, of the QRS complex. By noting the magnitude of this impulse response. it is then possible to use that value as the measure of area of the QRS complex, which value is supplied to the numerator input o~ divider 28.
Accordingly, the output o band-pass filter 32 is extended to the input of a full-wave rectifier 33 (similar to full-wave rectifier l6). The output of full-wave rectifier 33, which will ~13-3~
comprise pulses of a single polarity, are input to peak detector 34 (similar to peak detector 26). Peak detector 34 detects the peak values of the recti~ied impulses resulting from the output of band-pass filter 32, and a sample-and-hold circuit associated therewith records those peak valuesO The sample-and-hold associated with peak detector 34 receives a gating or sampling pulse via line 27' from sequencer 25 which extends approxi.mately 180 milliseconds from detection of an R-wave. Both the "height" peak detector 26 and the "axea" peak detector 34 note the respective peak amplitudes of the input signals during the predetermined 180 millisecond interval mentioned above during which the respective signals have experienced peak magnitude~
Divider 28 is of conventional design and serves to divide the numerator signal~ value of the QRS "area" from peak detector 34 lS by the denominator signal value of the QRS "height" value from peak detector 26. The resulting output from divider 28 is a ; normalized measure of the area proportional to the widt~,of the QRS complex. This indirect measure of width is utilized to determine the existence of a VPB.
By dividing the area o~ the QRS complex by it~ height, variations to the area as a result of changes in the amplitude of the ECG waveform are "norm21ized". In other words, the output o~
divider 28 provides an indirect measure of the width of a QRS
complex without relying solely on the measure o~ the area for that result. ~e output o~ divider 28 is connected to the input of a sample-and-hold circuit 36 which, in response to a sampling pulse applied thereto on line 27'' from sequencer 25, records and holds the indirect measure of width. The sampling signal on line 27'' occurs almost immediately after the sampling pulse associated with peak detectors 26 and 34.
f ~ 3~ ll The output of sample-and-hold 36 is input to a ~irst or lower i~
l~vel threshold circuit 40 and a ~econd or higher level threshold ,1 circuit 41 and also, by way of selectively controllable attenuator ¦~
37, to the input of averager 38. Attenuator 37 i5 of conve~tional ~;
voltage divider~type design and nonmally introduces no atte~uation to the "width" signal sup~lied to the input of averager 38.
Averager 38 is of conventional design and averages the indication of "width" for several successive heartbeats, i.e., about five sèconds under lX operation. The averaged value of several successive "width" measures is output fr~ averager 38 to the respective reerence inputs of comparators 42 and 43. The present values of "width" are applied to threshold circuits 40 and 41, the respective outputs of which comprise the variable inputs to comparatorS 42 and 43~ respectively. Threshold circuits 40 and 41 ¦
- 15 each comprise conventional voltage divider circuits wi~h the respective voltage dividers being scaled such that the output from I
circuit 40 is about five-sixths or 83% of the input thereto ~nd the ¦
output of circuit 41 is about five-se~enths or 71% of the input thereto. Stated another way, the instant "width" signal seen ¦
at comparator 42 will exceed ~hat applied by averager 38 ~nly when it exceeds 120% of the average width signal before passing through threshold circuit 40 and the instant width signal seen at ~he input of comparator 43 will exceed that applied by averayer 38 only when it exceeds 14~/o 0~ the avarage value be~ore passing through threshold circuit 41. Accordingly, the output of comparator 42 shi~t,s to a more positive potential so long as the present width value exceeds~the avarage width by 120% or more, and the output o~ comparator 43 shi~t~ to its positlve L23~
level whenever and ~o long as the instant width exceeds 140% of the average value. These positive output levels of comparators 42 and 43 are cle ignated WIDE I and WID~ II. AS with the signal from one-shot 22 indicating the occurrence of an R-wave, the WIDE I and WIDE II signals are applied to logic unit 44 where they are utilized for determining ~he occurrence o VPB's, atrial premature beats (APB's), successive VPB's and the like, as will be described.
Because it may be generally undesirable to substantially distort the average "width" value provided by averager 38 when an abnormally wide QRS complex occurs, attenuator 37 is auto-matically operative to dilute or attenuate the effect of such "wide" value on the average provided by circuit 38. WIDE I and WIDE II are output via lines 45 and 46 to control inputs of atten-uator 37. The WIDE I signal introduces a 10% attenuation factor to the present width signal subsequently extended to the input of - avexager 38. The WIDE II signal introduces an additional amount of attenuation, for instance, 10-20%, to the instantaneous value supplied to the input of averager 38. Thus, the average value of . several successive "width 1I measures is not seriously distorted by occasional wider QRS complexes.
"Area" peak detector 34 is output additionally to averager 47 which is identical to averager 38. The output of averager 47 is input to comparator 48. The output of "area" peak detector 34 is input through threshold cixcuit 49 to the other or variable input of compaxator 48. Threshold circuit 49 comprises a voltage divider scaled to provide an output which is two-thirds or 66% of its input. In other words, the input to comparator 48 provided by . -16-~3~
, the instantaneous "area" signal will equal the average "area"
value provided by averager 47 when the former, prior to passage through threshold circuit 49, is 15~/o of the latter. Therefore, the output of comparator 48 will shift to its positive level whenever and so long as the instantaneous "area" value is 150%
or more of the average "axea" value. This output value from comparator 48 is designated LAR OE AREA and is an optional input to logic system 44 and may be used to detect VPB's in the classical manner.
; 10 Additionally, the LARGE AREA output of comparator 48 is fed back via line 50 to a control input of peak detector 34 in substantially the same manner and for substantially the same purpose as the WIDE I and WIDE II controls for attenuator 37.
Specifically, an attenuating or bleed resistor ~not shown) associated with peak detector-sample-and-hold circuit 34 is connected into ~hat circuit by an electronic switch (not shown) by the occurrence of a LARGE AREA signal on line 50 to attenuate the signal which subsequently appears at the output of detector 34. This action occurs only after the present actual "area"
value has been presented to divider 28 and the "width" measure stored in sample-and-hold 36. By thereafter attenuating the output of peak detector 34, the "area" input to averager 47 will deviate less ~rom the average than previousl~. Also, as the output of peak detector 34 declines, so will its input to comparator 48 via threshold circuit 49, until the apparent "area" value is no longer greater than 150% of the average, whereupon the attenuation control via line 50 is interrupted. The instantaneous "area"
value will continue then at the 150% o~ average level until the -3~9 next QRS complex.
Circuitry is provided for monitoring the intel~al between successive QRS complexes for indicating if a QRS complex occurs earlier or later than some average-interval range established by the interval between several preceding QRS complexes. A xamp generator 52 initiates a voltage ramp of predetermined slope or ramp rate in response to a reset and trigger signal provided by digital sequencer via line 27'''~ The reset a~d trigger to ramp generator 52 occurs a predetermined fixed interval after the recognition of an R-wave, i.e., 180 milliseconds. The output of ramp generator 52 is connected to the input of sample-and-hold circuit 53. A sampling signal is extended to sample-and-hold 53 from digital sequencer 25 via line 27'''' and is timed to store that output of ramp generator 52 existing a brief interval prior to reset of ramp generator 52. In this way the value stored in sample-and-hold 53 is substantially proportional to the interval between successive R-waves, being only slightly less than the actual value. The output of sample-and-hold 53 is extended to the respective inputs of threshold circuits 55 and 56 to the input of averager 54.
Averager 54 is substantially identical to averagexs 36 and 47. The average of the several (i.e., five) immediately-- preceding R-R intervals appearing at the output of averager 54 is applied to the respective reference inputs of comparators 58 and 59 respectively. Both threshold circuits 55 and 56 are scaled to provide outputs which are about nine-tenths or 90% of their respective inputs; howevex, the output of threshold 55 is connected to the noninverting input of comparator 58 whereas the output of Z39~9 threshold 56 i5 connected to the inverting input of co~paxator 59.
In this way, the output of comparator 58 moves to a rela~ively positive voltage level to provide an EARLY signal whenever the instantaneous R-R interval is less than about 90% of the average R-R interval, and the output of comparator 59 moves to a rela-tively positive voltage level to provide a 1ATE signal whenever the instantaneous R-R interval is greater than about 11~% of the average R-R interval. The outputs of comparatoxs 58 and 59 providing the respective EARLY and LATE signals are extended to logic system 44.
It may be desirable that a very early complex or a succession of early complexes should not be allowed to distort the average. Means may be employed to detect these condi~ions and selectively chan~e the time constant of averager 54.
Referring to logic 44, a detailed analysis thereof is not undertaken in the interest of brevity and because a statement of the relevant logic expressions is believed sufficient to enable one of ordinary skill in the art to practice that aspect of the invention. The R-wave pulse provided by one-shot 22, in addition to being a key input to digital seguencer 25 ~or initiating each cycle of the sequencer~ is also extended to and through logic 44 to comprise an output therefrom for utiLization by any ancillary circuitry. Logic 44 is structured to provide an output indication of the occurrence of a VPB under either of the following two logic conditions, (1) a WIDE II indication by comparator 43 or (2) the indication of WIDE I by comparator 42 and an EARL~ indication by comparator S8 and a subsequent indication by comparator 59 that the next QRS complex is late.
~23~g In providing the logic ~or the EARLY and next LATE determinations, it will, of course, be n2cessary to provide one stage of delay or storage, as by the use of a flip-flop as is well known in the art. It should be understood that the 120% and 140% threshold values suggested for establishing WIDE I and WIDE II respectively in the illustrated embodiment, while being generally preferred, are certainly not limiting and some range o~ variation is within the scope of the invention~
Logic 44 indicates an atrial premature beat (APB) with logic that recognizes the occurrence of the following conditions:
an EARLY QRS complex which is also not WIDE I and neither the previous nor the following QRS complexes are EARLY.
Logic ~4 also provides an indication of successive VPB's when one of the ~ollowing logic conditions is satisfied: (l) a QRS complex is EARLY and WIDE I and the next QRS complex is EARLY
- . and is also WIDE I, (2) a QRS complex is EARLY and WID~ I and the next complex is WIDE II, or (3~ a QRS complex is WIDE II and the next complex is WIDE II.
Although the invention has been described in the context of a hard-wired system including mostly analog circuitry with some digital sequencing and logic circuits, the invention may be implemented by means of a programmable computer or microprocessor.
While implementation entirely with digital circuitxy and a microprQcessor may be possible, it is preferable that khe input stages involving rate limiting and filtering be accomplished in the analog domain with the ~ollowing signal processing being provided by a progra~mable microprocessor as illustrated in Figure 5. In that ~igure, low-pass filter lO, slew-rate limiter .
3~
12, multistage band-pass filter 14 and band-pass filter 32 are identical to those illustrated in Figure 2, retaining the s~me reference numerals. The output o~ band-pass filter 14 is passed through an absolute value circuit 16 ' analogous to full-wave rectifiex 16. Similarly, the output of band-pass filter 32 is - passed through an absolute value circuit 33' analogous to full-wave rectifier 33. The output signals from absolute value circuits 16' and 33' respectively are extended to the inputs of analog-to-digital convertexs 70 and 71 respectively for conversion from analog to digital ~orm. The output signal ~rom A-to-D
converter 70 is designated HEIGHT and represents the digital absolute value representations of the signal passed by filter 14.
The output signal fro~ A-to-D converter 71 is designated AREA
and is a digital representation of the absolute value of the ~5 impulse response of filter 32. It should be understood that neither of these signals is here an actual measure of the height or the area of the QRS complex but are used to provide such. The digital "height" and "area" signals are respectively connected to the inputs of a suitable microprocessor 72 for subse~uent processing.
Referring to the flow chart of Figures 6A and 6B, to be considered as one, there is illustrated the sequence of control and decision making implementable by program on microprocessor 72.
Firstly, the peak value o~ the successive "height" signals axe successivel~ xtored. Then, using those stored peak values, a variable threshold is established, which threshold may approach 100% o~ the stored peak value immediatel~ after detection o~
an R-wave and declines therefrom to a ~alue o~ about 40% in the normal interval to the next R-wave. Successive values o* new "height" signals associated with the next QRS complex are then compared with the variable threshold value. If the value of a new "height" signal is less than the threshold vaIue, the control loop operates to permit continued reduction of the variable threshold level toward the 40% level. If no R-wave is detected after some time, i.e., two ox three normal R-R inter~als, then the threshold could be further reduced. If, on the other hand, the new "height" signal exceeds the threshold value, this signifies the ocurrence of an R-wave and an "R-W~VE" flag is set.
Upon indication of the occurrence of an R-wav~ the peak value of the AREA signal is stored as a measure of area. Subse-quently, the stored peak value of arQa is divided by the stored peak value of height to pxovide a value which is an indirect measure of the width of the respective Q~S complex, which measure is then stored. That value and several (i.e., four) immediately-preceding stored values of width are averaged to provide an average value of width. The stored value representing width is compared with the average value of widths to determine first if it is at least 140% wider than the average width and secondly if it is at least 120% wider than the average width. If at least 140%
wider, a "WIDE II" flag is set. If at least 120% wider, a "WIDE I" flag is set. Further, if at least 120% wider than average, the effect of the most recently stored value of width on the average is minimized by a first amount, and if at least 140%
widex, ~he effect on the average is reduced by an even greater amount.
I'he present stored peak value of area and the previous 3~
several stored peak values of area are averaged to provide an average value of area. ~he most recently stored peak value of area is compared with 150/~ of the average peak value of area and if the former is larger than the latter, a "LARGER ARE~"
flag is set. Also, if the instant value of area is more than 150% of the average value of area, the value o the former in computing the average is minimized. The interval between the settiny of successive R-WAVE flags is measured to provide the measure of R R interval. The interval between the present R-wave and the immediately-preceding R-wave, together with the intervals between the several immediately previous R-waves are averaged to provide an average R-R interval. The most recent R-R interval is compared with the avexage R-R interval and, if the former is less than 90% of the latter, an "EARLY" flag is set. If the most recent R-R interval is greater than 110%
of the average R-R interval, a "LATE" flag is set. If the most recent R-R interval is less than 90% of the average R-R interval, the effect of the former in determining the latter is minimized.
M~THOD AMD APPARATUS FOR MONITORING
E~ECTROCARDIOGRAPHIC WA~EFORMS
The invention relates to electrocardiographic monitoring systems which detect abnormal ECG waveforms.
In the field of cardiology, it is common to continuously monitor the ECG signal of a patient for analysis. Normally, it is impractical to continuously monitor the ECG waveform and systems have evolved for automatically monitoring and analyzing the waveform.
~he ECG waveform is normally comprised of a series of characteristic points conventionally designated by the letters P, Q, R, S and T. ~he Q, R a~d S portions of the wave when taken together are referred to as the QRS complex. Means for determining ~hat the QRS complex has occurred are commonly referred to as R-wave detectors, a representative example of which being disclosed in U.SO Patent No. 3,490,811 issued ~uly 6, lg71, to ~arris for Electrocardiographic R-Wave Detector. However, it i5 often desirable to further dif~erentiate between a QRS complex which corresponds to a normal heart action, and a QRS complex which corxesponds to abnormal heart action~
One such abnormal heart action of interest is that of ectopic beats. Ectopic bea~ are characterized by departure from a "~ormal" interbeat intexval and~or departure from a "~ormal"
wave~orm morphology. "~ormal" interbeat interval and/or waveæorm morphology for one patient may differ ~rom that which is "Normal" ~or another. Ectopic beats deserving o~ special , 3~
attention include ventxicular premature beats ~VPB), atrial premature beats (APB), and successive groupings of such premature beats.
Ven~ricular premature beats (VPB's) have been identified by monitoring the interval between successive QRS complexes and the width and/or area of such complexes, and signaling the occurrence of such VPB if those parameters differ by more than a predetermined amount from that which is "Normal" for the particular patient.
An example of these techniques is disclosed in some detail in U.S. Patent 3,616,791 issued November 2, 1971, to G. J. Harris for Electrocardiographic Morphology Recognition System. In that patent, certain QRS complexes are identified as VPB's if their width is greater than normal. That determination of width i~
made by measuring th`e area under the rectified ECG signal and comparing it with the average area That area determination is made by integrating the rectified waveform. Be~ore actually indicating that a "wide" QRS complex is actually a VPB, the interval between QRS complexes is normally required to exhibit a so-called compensatory pause. The QRS complex of a VPB will usually occur earlier than expected and the next succeeding QRS
complex will occur after a longer than no~mal interval or compensatory pause. This combination of a "wide" QRS complex occurring earlier than usual and followed by a compensatory pause is generally a reliable indicator of a VPB.
It would be more preferable to identify a VPB based on width alone without analysis of the interbeat interval. Howe~er, such analysis has not heretofore been relied on, in part because the area, and thus the width measure, of the QRS complex may increase as the height or amplitude of the waveform increases due to changes in patient respiration, drift in the signal baseline and/or other causes.
Accordingly, it is a pxincipal object of the present inven~ion to provide a method and apparatus for analyziny the ECG wave in a manner which simply and accurately identifies VPB's, and to provide improvad mea~s for accurately identifying the occurrence of the QRS complex in an ECG waveform whereby accurate identification of VPB's a~d other heart characteristics may be'made.
The present invention is bas,ed on the recognition of VPB's upon,the width of a particular QRS complex exceeding an average width for such complexes by a predetexmined amount. The de~ermination of wid~h for each QRS complex is done indirectly with a measurement of QRS area normalized to remove variations in waveform amplitude by dividing the area by the height thereof.
Although this width measure does not result in a signal which is identical to ~he width of the complex, it does result in a signal which is proportional to that width.
According to a preferred embodiment o~ the invention, a particula.r hear~beat is designated a premature ventricular contraction if the QRS complex is determined to be at least 40% wider than the average width of the immediately-preceding several QRS complexes. Supplementally, a VPB may also be indicated if the width of a QRS complex is early, exceeds the average'width by at least 20/~ and is additionally ~ollowed by a compensatory pause as determined by interbea~ interval measure-ments~
~3~
The l.CG signal is fed in~o two parallel filters, the outputs of which are rectified and peak de-tected. The bandwidth of one fil-ter is -that of a normal Q~S complex such that its output is proportional to height. The bandwidth of the other filter is below that of the first such tha-t -the filter's impulse response is in turn proportional to the area of the complex.
The peak-detected area is then divided by the peak-detected height to obtain a wid-th rneasure which is compared with an average of preceding width measures.
According to another aspect of the inventionr an improved R-wave detector is provided, which detector exhibits increased accuracy for controlling the peak detection of the height and area measures of a QRS complex. More particularly, a variable threshold is associated with generally conventional R-wave detection means whereby to increase sensitivity to ~ rela-tively small amplitude R-waves and decrease sensitivity to T-waves of relatively large amplitude. Both the VPB and the R-wave detection circuitry adapt to individual patients and to variations over time in the QRS complex of a single patient through use of averaging techniquesO
In summary of the above, therefore, the present invention may be seen as providing in an ECG monitoring system, the method of indirectly providing a measure of the width of a QRS complex appearing in an ECG waveform comprlsing the steps of: generating from the ECG waveform a siynal proportional to the area of the QRS complex; generating from the ECG waveform a slgnal propor-tional to the peak height o~ the QRS complex; and dividing the area signal by the peak amplitude signal to provide a normalized signal representative of the width of the QRS complex of the ECG signal.
B
sd~
: . .
The above method may be carried out by way oE an ECG
monitoring system for generating a signal indicatiny a ventricular premature beat comprising means responsive to an ECG waveform for providing a measure proportional to the area of the QRS
complex of the ECG waveform, means responsive to the ECG wave-form for providing a measure proportional to the peak height of the QRS complex of the ECG waveform, and means for dividing the area measure by the peak height measure thereby to provide a normalized measure proportional to the width of the QRS
complex of the ECG waveform.
Figure 1 depicts a typical ECG of an individual having a mixture of normal and abnormal heartbeats, the uppermost trace being the ECG waveform fox a single normal heartbeat;
Figure 2 is a detailed schematic block diagram of apparatus for practicing the invention;
Flgure 3, appearing on the same sheet as Figure 1, is a more detailed schematic diagram of certain portions of the .
R-wave de.e_tor forming part of the apparatus of Figure 2;
. .......................... ' .
. ' ' ,~
sd~ 4A-~3~3~L9 F'igure 4 is a more detailed schematic diagram of a band-pass filter utilized for its impulse response in ~he apparatus o~
Figure 2, Figure 5 is a flow diagram of a combined hard-wired a~d S computer-based hybrid of the circuitry of Figure 2; and Figuxe 6A and 6B in combination comprise the general flow diagram for the programmed computer-based portion of Figure 5.
In Figure 1 thPre is illustrated in the bottom trace an electrocardiogram ~ECG) ~f a heart patient characterized by a mixture of normal and abnormal beats. The ECG in the uppex trace of Figure 1 represents an enlarged single normal heartbeat of the patient. The different characteristic portions o~ the wave~orm are identified by the conventional signals P, Q, R, S
~ 1 and T. In the lower trace, two normal beats spaced at a proper time interval are followed by cne abnormal beat which is premature and which is followed by another normal beat~ The premature beat ollows the second normal beat more closely and is spaced from the final normal beat by more than the normal interval~ That increased interval is commonly referred to as the compensatory pause.
The ECG monitoring system illustrated in Figure 2 receives the ECG signal of Figure 1 as its principa} input and analyzes that waveform to detect VPB's as well as other types o~ ectopic beats. ~he circuitry of Figure 2 is particularly suited ~or analyzing ECG signals previously recorded, as during ambulatory monitoring; but is similarly applicable to on-line or real-time situations. The playback o~ prev~ously recorded ECG signals may 3~
occur at the same rate at which the si~nals were recorded (lX~, or, at multiples of the speed at which they were recorded,~e.g.
60X, 120X). Speed~up of playback expedites the analysis of the recorded signal. The following discussion of the circuitry of Figure 2 assumes playback at the same speed (lX~ as recorded.
The various time constants associated with the circuitry of Figure 2 will thus have a value for a lX playbacX speed, and alternate volume for 60X and 120Xo The capabi}ity of selectively processing the ECG signal at multiple playback speeds may be accomplished eit~er by having selective channels, each having different appropriate time constants or by a single channel having components with different time constants selectively insertable therein. Thi5 selectivity may be accomplished in a program-controlled, microprocessor-based system~
In Figure 2, the ECG signal, derived from the playback heads of a recorder is connected as an input to low-pass filter 10.
Low-pass filter 10, in addition to comprising the input to the overall analyzer system of Figure 2, is also the first staqe of an R-wave detector, and includes slew-rate limiter 12, band-pass filter 14, full-wave rectifier 16, peak detector 18, variable threshold 20, comparator 21, and non-xetriggerable one-shot 2?
(to be described in greater detail). Low-pass filter 10 is designed to remove 50/60 Hz, line noise (at lX) from the input ECG signal.
The output of low-pass filter 10 is connected to the input of slew-rate limiter 12 (of conventional design) and scaled to suppress pacemaker spikes, noise associated with tape dropout, musc~e artifact and the like. Normally, the signals passed ~z~
by slew-xate limiter 12 will include normal ECG waveforms, VPB's and other ectopic beats and possibly some muscle artifact.
The output of slew-rate lLmiter 12 is input to multistage band-pass filter 14. A first portion of filter 149 shown ~n greater detail as part of Figure 3, is shared by tha R-wave detection circuitry as well as the "area" detector (hereinafter described). The latter portion of filter 14 is utilized only by the R-wave detection circuitry and the waveform height or amplitude detector. The initial or common portion of band-pass filter 14 typically has a band-pass range of .8-8 Hz and the final section comprises a high-pass section which, in combination with the preliminary stages, provides an overall band-pass range to the filtex o 8-32 Hz for lX operation. This range of 8-32 Hz is intended to pass not o~ly nonmal QRS complexes but also those o~ an ectopic nature. The lower frequency T-wave component is attenuated to some extent by filter 14.
The output signal from multistage band-pass filter 14 is input to a full-wave rectifier 16 of conventional design.
Rectifier 16 ensures that all signal excursions to either side of the base line appear on only one side o~ the base line, eliminating need for positive and negative threshold detection as a result o~ the possibly reversed polarities to the input ECG
signal.
The output o~ full-wave rectiEier 16 is input to peak detec~ox 18 and as one o~ the two inputs to comparator 21.
Whereas ~ilter 10, rate limiter 12, band-pass filter 14 and full-wave recti~ier 16 comprise means for "cleaning up" the ECG signal for R-wave detection, peak detector 18, variable ~hreshold 20 ~23~
and comparator ~1 provide the actual means for detecting the ~-wave and one-shot 22 indicates detection. The combination of peak detector 18, variable threshold 20 and comparator 21 with feedhack of the output of compaxator 21 to variable threshold 20 via conduc~or 23 provides novel means for accurately detecting R-waves, even those of r~latively small magnitude as may be associated with ectopic beats, without also responding to T~waves, - particularly those of relatively large magnitude.
Additional detail to the combination of peak detector 18, variable threshold 20 and comparator 21 is illustrated in Figure 3. The output from rectifier 16 of Figure 2 is applied to the noninverting input of an operational amplifier Al via input resistor Rl. The output of amplifier Al is extended to the noninvertlng input of amplifier A2 through the series circuit of a low value resistor R2, a peak detecting diode Dl ancl an input resistor R3. A capacitor Cl and resistor R4 are connected in parallel to ground ~rom the cathode Qf diode Dl. The capacitor Cl is charged by the rectified voltage pulses corresponding with the R-wave as well as other portions of the ECG signal.
The value of resistor R4 in combination with the value of capacitor Cl are selected such that capacitor Cl discharges at a relatively slow rate, for instance, to two-thirds of its value in about five seconds or five heartbeats. The output o amplifier A2 is fed back to the inverting inputs of amplifiers Al and A2. A clamping diode D2 is connected between the :inverting input and the output of amplifier Al.
The voltage appearing across capacitor Cl and subsequently at the output of amplifier A2, and khus also the output of peak ~ 23~
detector 18 of Figure 2, reflects the maximum peak vol~age, or a~ least an average of the peak voltages associated with the QRS complexes, at the input over the preceding interval of several seconds (lX operation). Typically, the R-wave is responsible for the largest peak voltage in the ECG waveform during each heartbeat. Whether it is the largest peak over several heartbeats or a value more nearly to an average of the peaks over that interval appearing at the output, the output is representative of the peak value of one or more QRS
complexes during that inter~al. This output is applied to the input of variable threshold circuit 20 for variable scaling and subsequent application as the reference input to comparator 21.
Although described as a peak detector, circuit 18 may embody the circuits of U.S_ Patent 3,490,811 such that the output is - 15 clearly an avarage of the peaks.
Variable threshold 20 includes a voltage divider comprised of resistors R5 and R6 with one end of R5 being connected t~
the output of peak detector 18 and one end of R6 being connected to ground, the ~unction between R5 and R6 being connected through input resistor R7 to the inverting or reference input of amplifier A3 of comparator 21. The values of resistors R5 and R6 are scaled such that, at steady state, the threshold value applied to comparator 21 is normally about 40% if the peak value provided as the output from peak detector 18. The signal output ~rom rectifier 16 is also led directly to the other input of comparator 21 via lead 24. The signal amplitude on lea~ 24 is compared by amplifier A3 with that provided by variable threshold 20. The normally relatively low voltage at the output of _g _ ~3~3~S~
amplifier A3 goes relatively more positive whenever the rectified ~CG signal exceeds the instantaneous thre~hold value.
This positi~e transi~ion in the voltage at the output o~
ampli~ier A3 is developed across resistor R8 and serves as the S triqger input to non-retriggerable one-shot 22. When one~shot 22 is triggered, it provides a positive pulse of approximately 180 milliseconds duration for signifying the occuxr0nce of an R-wave and keying the operation of a digital sequencer 25.
In addition to triggering one-shot 22, the positive going voltage at the output of comparator 21 is fed back via lead 23 through diode D3 and base resistox R9 to the base of transistor Tl, A capacitor C2 is comlected in parallel across resistor R5 and the emitter and collector of transistor Tl are similarly oonnected ïn parallel across resistor R5 and capacitor C2. Immedia~ely prior to the occurrence of the R-wave i~ a particular ECG wav~form, capacitor C2 will be charged and the threshold voltage appearing at the ~unction of resistors R5 and R6 will be about 40% that of the output of peak detector 18.
~Owever, as soon as the ampli~ude of the present R-wave exceeds the threshold value and evokes a positive response from comparator 21, the positi~e going voltage on lead 23 operates to turn on transistor Tl, shoxt-circuiting capacitor C2 and resistor R5 and applying the value (i.e., 100%) appearing at the output of peak detector 18 as the threshold value to the reference input of comparator 21~ This has the e~fect o~
i~creasing the threshold for any relatively large T-waves which may ~ollow a QRS complex. Stated another way, the variable threshold enables increased sensitivity to small amplitude R-waves without requiring the threshold to remain so low as to additionally detect the following T-waves. As the threshold reference value rapidly ihcxeases to near lOQ% of the output of peak detector 18, it is no longer exceeded by the input appearing on line 24~ Accordingly, the output of compaxator 21 drops to its normal potential, thereby returning transistor Tl to nonconduction. The RC time constant of capacitor C2 and resistor R5 in parallel with resistor R6 is such that recharge of capacitor C2, and thus return of the threshold network to its normal 40% value, occurs continuously over an interval sufficiently long to exclude T-waves but sufficiently short to ant:icipate the next succeeding R-wave. A typical RC time constant is about 200-300 milliseconds at lX operation. Although more di~ficult techically, variable scaling might be applied to the rectifiPd EC~ signal from rectifier 16, except that the function would be inverted, i.e~, the ECG signal would be amplified by 2-1/2 times until R-wave detection whereupon it would return to normal amplitude to "miss" the T-wave then gradually increase to 2-1/2 times to detect the next R-wave.
Returning to the discussion of Figure 2, attention is given another portion of the ECG monitoring system comprising a principal aspect of the invention. The output of full-wave rectifier 16 is additionally extended via line 24' to the input o~ peak detector 26. Peak detector 26, in addition to conventional peak detecting circuitry of a type hereinbe~ore described, also includes hold circuitry ~or output of only the peak rectified signal CQmmensUrate with a QRS complex. A gating or reset signal is provided via line 27 from digital sequencer Z5 to ~.2~9 the reset input associated with the hold portion o peak detector 26. Digital sequencer 25 is of conventiona} design and is driven by a clock ~not shown). The R-wave indication provided by one-shot 22 serves to initiate certain predetermined sequences of control signals provided by se~uencer 25. Accordingly, the reset signal appeaxs on line 27 for only a brief inter~al immediately after an R-wave has been detected to reset t~e hold circuit for permitting detector and hold circuit 26 to respond to the present QRS complex. The output of peak detector 26 is proportional ~o the "height" o~ the QRS complex and is extended to the denominator input of a divider 28.
To obtain a measure of the area of the QRS complex o the input ECG wave~orm, ~he signal is led from an intermediate section of the multistage band-pass filter 14 via lead 0 to an addi~ional band-pass fllter 32. The characteristics of the pra-liminary stage of filter 14 and the band-pass fil~er 32 are - selected such that the impulse response thereof to the QRS complex will provide an accurate measure of the axea of the QRS complex.
Referring to the preliminary stages of multistage band-pass ~ilter 14 in Figure 4, the ECG signal passes through a first filter stage including ampli~ier A4 and a second filte,x stage including ampli-fier A5 to junction J, then lead 30 to ~ilter 32. The input to amplifier A4 is through a series-combination sf capacitor C3 and resistor R9. A parallel combination of capacitor C4 and resistor 2S R10 is connected between the output and the input o~ ampli~ier A4.
A "T" networ.k comprised of series res.istors Rll and R12, and capacitor C5 therebetween to ground, is connected between the output o~ amplifier A4 and the input of amplifier A5~ A parallel -~2-combination of capacitor C6 and resistor R13 is connected between the output and input of amplifier A5. The values of capacitors C3-C6 and resistors R9-R13 are selected to provide a band-pass range of about .8-8 Hz at lX operation, or may be selected to provide a 40 Hz-470Hz passband for 60X operation, etc.
Band-pass filter 32 includes amplifier A6, the series combination of capacitor C7 and resistor R14 in lead 30 to the input thereof, and the parallel combination of resistor R15 and capacitor C8 extending between the output: and input thereof.
The values of these resistances and capacitances are selected such that this filter individually has a passband range of about 2-20 Hz d~ring lX operation, but when considered in conjunction with the earlier sta~es of ~iltering in filter 14, the band-pass between the input to filter 14 and the output of ~r 15 filter 32 is in the range of about 2~8 Hz for lX operation.
The upper end of this passband is lower than most of the instantaneous frequency ch~racteristics of portions of the QRS
complex. For this reason, the output from filter 32 does not accurately follow the QRS complex but instead provides a ringing impulse response therefor. The amplitude or magnitude of the impulse response is proportional to the energy content, and thus the area, of the QRS complex. By noting the magnitude of this impulse response. it is then possible to use that value as the measure of area of the QRS complex, which value is supplied to the numerator input o~ divider 28.
Accordingly, the output o band-pass filter 32 is extended to the input of a full-wave rectifier 33 (similar to full-wave rectifier l6). The output of full-wave rectifier 33, which will ~13-3~
comprise pulses of a single polarity, are input to peak detector 34 (similar to peak detector 26). Peak detector 34 detects the peak values of the recti~ied impulses resulting from the output of band-pass filter 32, and a sample-and-hold circuit associated therewith records those peak valuesO The sample-and-hold associated with peak detector 34 receives a gating or sampling pulse via line 27' from sequencer 25 which extends approxi.mately 180 milliseconds from detection of an R-wave. Both the "height" peak detector 26 and the "axea" peak detector 34 note the respective peak amplitudes of the input signals during the predetermined 180 millisecond interval mentioned above during which the respective signals have experienced peak magnitude~
Divider 28 is of conventional design and serves to divide the numerator signal~ value of the QRS "area" from peak detector 34 lS by the denominator signal value of the QRS "height" value from peak detector 26. The resulting output from divider 28 is a ; normalized measure of the area proportional to the widt~,of the QRS complex. This indirect measure of width is utilized to determine the existence of a VPB.
By dividing the area o~ the QRS complex by it~ height, variations to the area as a result of changes in the amplitude of the ECG waveform are "norm21ized". In other words, the output o~
divider 28 provides an indirect measure of the width of a QRS
complex without relying solely on the measure o~ the area for that result. ~e output o~ divider 28 is connected to the input of a sample-and-hold circuit 36 which, in response to a sampling pulse applied thereto on line 27'' from sequencer 25, records and holds the indirect measure of width. The sampling signal on line 27'' occurs almost immediately after the sampling pulse associated with peak detectors 26 and 34.
f ~ 3~ ll The output of sample-and-hold 36 is input to a ~irst or lower i~
l~vel threshold circuit 40 and a ~econd or higher level threshold ,1 circuit 41 and also, by way of selectively controllable attenuator ¦~
37, to the input of averager 38. Attenuator 37 i5 of conve~tional ~;
voltage divider~type design and nonmally introduces no atte~uation to the "width" signal sup~lied to the input of averager 38.
Averager 38 is of conventional design and averages the indication of "width" for several successive heartbeats, i.e., about five sèconds under lX operation. The averaged value of several successive "width" measures is output fr~ averager 38 to the respective reerence inputs of comparators 42 and 43. The present values of "width" are applied to threshold circuits 40 and 41, the respective outputs of which comprise the variable inputs to comparatorS 42 and 43~ respectively. Threshold circuits 40 and 41 ¦
- 15 each comprise conventional voltage divider circuits wi~h the respective voltage dividers being scaled such that the output from I
circuit 40 is about five-sixths or 83% of the input thereto ~nd the ¦
output of circuit 41 is about five-se~enths or 71% of the input thereto. Stated another way, the instant "width" signal seen ¦
at comparator 42 will exceed ~hat applied by averager 38 ~nly when it exceeds 120% of the average width signal before passing through threshold circuit 40 and the instant width signal seen at ~he input of comparator 43 will exceed that applied by averayer 38 only when it exceeds 14~/o 0~ the avarage value be~ore passing through threshold circuit 41. Accordingly, the output of comparator 42 shi~t,s to a more positive potential so long as the present width value exceeds~the avarage width by 120% or more, and the output o~ comparator 43 shi~t~ to its positlve L23~
level whenever and ~o long as the instant width exceeds 140% of the average value. These positive output levels of comparators 42 and 43 are cle ignated WIDE I and WID~ II. AS with the signal from one-shot 22 indicating the occurrence of an R-wave, the WIDE I and WIDE II signals are applied to logic unit 44 where they are utilized for determining ~he occurrence o VPB's, atrial premature beats (APB's), successive VPB's and the like, as will be described.
Because it may be generally undesirable to substantially distort the average "width" value provided by averager 38 when an abnormally wide QRS complex occurs, attenuator 37 is auto-matically operative to dilute or attenuate the effect of such "wide" value on the average provided by circuit 38. WIDE I and WIDE II are output via lines 45 and 46 to control inputs of atten-uator 37. The WIDE I signal introduces a 10% attenuation factor to the present width signal subsequently extended to the input of - avexager 38. The WIDE II signal introduces an additional amount of attenuation, for instance, 10-20%, to the instantaneous value supplied to the input of averager 38. Thus, the average value of . several successive "width 1I measures is not seriously distorted by occasional wider QRS complexes.
"Area" peak detector 34 is output additionally to averager 47 which is identical to averager 38. The output of averager 47 is input to comparator 48. The output of "area" peak detector 34 is input through threshold cixcuit 49 to the other or variable input of compaxator 48. Threshold circuit 49 comprises a voltage divider scaled to provide an output which is two-thirds or 66% of its input. In other words, the input to comparator 48 provided by . -16-~3~
, the instantaneous "area" signal will equal the average "area"
value provided by averager 47 when the former, prior to passage through threshold circuit 49, is 15~/o of the latter. Therefore, the output of comparator 48 will shift to its positive level whenever and so long as the instantaneous "area" value is 150%
or more of the average "axea" value. This output value from comparator 48 is designated LAR OE AREA and is an optional input to logic system 44 and may be used to detect VPB's in the classical manner.
; 10 Additionally, the LARGE AREA output of comparator 48 is fed back via line 50 to a control input of peak detector 34 in substantially the same manner and for substantially the same purpose as the WIDE I and WIDE II controls for attenuator 37.
Specifically, an attenuating or bleed resistor ~not shown) associated with peak detector-sample-and-hold circuit 34 is connected into ~hat circuit by an electronic switch (not shown) by the occurrence of a LARGE AREA signal on line 50 to attenuate the signal which subsequently appears at the output of detector 34. This action occurs only after the present actual "area"
value has been presented to divider 28 and the "width" measure stored in sample-and-hold 36. By thereafter attenuating the output of peak detector 34, the "area" input to averager 47 will deviate less ~rom the average than previousl~. Also, as the output of peak detector 34 declines, so will its input to comparator 48 via threshold circuit 49, until the apparent "area" value is no longer greater than 150% of the average, whereupon the attenuation control via line 50 is interrupted. The instantaneous "area"
value will continue then at the 150% o~ average level until the -3~9 next QRS complex.
Circuitry is provided for monitoring the intel~al between successive QRS complexes for indicating if a QRS complex occurs earlier or later than some average-interval range established by the interval between several preceding QRS complexes. A xamp generator 52 initiates a voltage ramp of predetermined slope or ramp rate in response to a reset and trigger signal provided by digital sequencer via line 27'''~ The reset a~d trigger to ramp generator 52 occurs a predetermined fixed interval after the recognition of an R-wave, i.e., 180 milliseconds. The output of ramp generator 52 is connected to the input of sample-and-hold circuit 53. A sampling signal is extended to sample-and-hold 53 from digital sequencer 25 via line 27'''' and is timed to store that output of ramp generator 52 existing a brief interval prior to reset of ramp generator 52. In this way the value stored in sample-and-hold 53 is substantially proportional to the interval between successive R-waves, being only slightly less than the actual value. The output of sample-and-hold 53 is extended to the respective inputs of threshold circuits 55 and 56 to the input of averager 54.
Averager 54 is substantially identical to averagexs 36 and 47. The average of the several (i.e., five) immediately-- preceding R-R intervals appearing at the output of averager 54 is applied to the respective reference inputs of comparators 58 and 59 respectively. Both threshold circuits 55 and 56 are scaled to provide outputs which are about nine-tenths or 90% of their respective inputs; howevex, the output of threshold 55 is connected to the noninverting input of comparator 58 whereas the output of Z39~9 threshold 56 i5 connected to the inverting input of co~paxator 59.
In this way, the output of comparator 58 moves to a rela~ively positive voltage level to provide an EARLY signal whenever the instantaneous R-R interval is less than about 90% of the average R-R interval, and the output of comparator 59 moves to a rela-tively positive voltage level to provide a 1ATE signal whenever the instantaneous R-R interval is greater than about 11~% of the average R-R interval. The outputs of comparatoxs 58 and 59 providing the respective EARLY and LATE signals are extended to logic system 44.
It may be desirable that a very early complex or a succession of early complexes should not be allowed to distort the average. Means may be employed to detect these condi~ions and selectively chan~e the time constant of averager 54.
Referring to logic 44, a detailed analysis thereof is not undertaken in the interest of brevity and because a statement of the relevant logic expressions is believed sufficient to enable one of ordinary skill in the art to practice that aspect of the invention. The R-wave pulse provided by one-shot 22, in addition to being a key input to digital seguencer 25 ~or initiating each cycle of the sequencer~ is also extended to and through logic 44 to comprise an output therefrom for utiLization by any ancillary circuitry. Logic 44 is structured to provide an output indication of the occurrence of a VPB under either of the following two logic conditions, (1) a WIDE II indication by comparator 43 or (2) the indication of WIDE I by comparator 42 and an EARL~ indication by comparator S8 and a subsequent indication by comparator 59 that the next QRS complex is late.
~23~g In providing the logic ~or the EARLY and next LATE determinations, it will, of course, be n2cessary to provide one stage of delay or storage, as by the use of a flip-flop as is well known in the art. It should be understood that the 120% and 140% threshold values suggested for establishing WIDE I and WIDE II respectively in the illustrated embodiment, while being generally preferred, are certainly not limiting and some range o~ variation is within the scope of the invention~
Logic 44 indicates an atrial premature beat (APB) with logic that recognizes the occurrence of the following conditions:
an EARLY QRS complex which is also not WIDE I and neither the previous nor the following QRS complexes are EARLY.
Logic ~4 also provides an indication of successive VPB's when one of the ~ollowing logic conditions is satisfied: (l) a QRS complex is EARLY and WIDE I and the next QRS complex is EARLY
- . and is also WIDE I, (2) a QRS complex is EARLY and WID~ I and the next complex is WIDE II, or (3~ a QRS complex is WIDE II and the next complex is WIDE II.
Although the invention has been described in the context of a hard-wired system including mostly analog circuitry with some digital sequencing and logic circuits, the invention may be implemented by means of a programmable computer or microprocessor.
While implementation entirely with digital circuitxy and a microprQcessor may be possible, it is preferable that khe input stages involving rate limiting and filtering be accomplished in the analog domain with the ~ollowing signal processing being provided by a progra~mable microprocessor as illustrated in Figure 5. In that ~igure, low-pass filter lO, slew-rate limiter .
3~
12, multistage band-pass filter 14 and band-pass filter 32 are identical to those illustrated in Figure 2, retaining the s~me reference numerals. The output o~ band-pass filter 14 is passed through an absolute value circuit 16 ' analogous to full-wave rectifiex 16. Similarly, the output of band-pass filter 32 is - passed through an absolute value circuit 33' analogous to full-wave rectifier 33. The output signals from absolute value circuits 16' and 33' respectively are extended to the inputs of analog-to-digital convertexs 70 and 71 respectively for conversion from analog to digital ~orm. The output signal ~rom A-to-D
converter 70 is designated HEIGHT and represents the digital absolute value representations of the signal passed by filter 14.
The output signal fro~ A-to-D converter 71 is designated AREA
and is a digital representation of the absolute value of the ~5 impulse response of filter 32. It should be understood that neither of these signals is here an actual measure of the height or the area of the QRS complex but are used to provide such. The digital "height" and "area" signals are respectively connected to the inputs of a suitable microprocessor 72 for subse~uent processing.
Referring to the flow chart of Figures 6A and 6B, to be considered as one, there is illustrated the sequence of control and decision making implementable by program on microprocessor 72.
Firstly, the peak value o~ the successive "height" signals axe successivel~ xtored. Then, using those stored peak values, a variable threshold is established, which threshold may approach 100% o~ the stored peak value immediatel~ after detection o~
an R-wave and declines therefrom to a ~alue o~ about 40% in the normal interval to the next R-wave. Successive values o* new "height" signals associated with the next QRS complex are then compared with the variable threshold value. If the value of a new "height" signal is less than the threshold vaIue, the control loop operates to permit continued reduction of the variable threshold level toward the 40% level. If no R-wave is detected after some time, i.e., two ox three normal R-R inter~als, then the threshold could be further reduced. If, on the other hand, the new "height" signal exceeds the threshold value, this signifies the ocurrence of an R-wave and an "R-W~VE" flag is set.
Upon indication of the occurrence of an R-wav~ the peak value of the AREA signal is stored as a measure of area. Subse-quently, the stored peak value of arQa is divided by the stored peak value of height to pxovide a value which is an indirect measure of the width of the respective Q~S complex, which measure is then stored. That value and several (i.e., four) immediately-preceding stored values of width are averaged to provide an average value of width. The stored value representing width is compared with the average value of widths to determine first if it is at least 140% wider than the average width and secondly if it is at least 120% wider than the average width. If at least 140%
wider, a "WIDE II" flag is set. If at least 120% wider, a "WIDE I" flag is set. Further, if at least 120% wider than average, the effect of the most recently stored value of width on the average is minimized by a first amount, and if at least 140%
widex, ~he effect on the average is reduced by an even greater amount.
I'he present stored peak value of area and the previous 3~
several stored peak values of area are averaged to provide an average value of area. ~he most recently stored peak value of area is compared with 150/~ of the average peak value of area and if the former is larger than the latter, a "LARGER ARE~"
flag is set. Also, if the instant value of area is more than 150% of the average value of area, the value o the former in computing the average is minimized. The interval between the settiny of successive R-WAVE flags is measured to provide the measure of R R interval. The interval between the present R-wave and the immediately-preceding R-wave, together with the intervals between the several immediately previous R-waves are averaged to provide an average R-R interval. The most recent R-R interval is compared with the avexage R-R interval and, if the former is less than 90% of the latter, an "EARLY" flag is set. If the most recent R-R interval is greater than 110%
of the average R-R interval, a "LATE" flag is set. If the most recent R-R interval is less than 90% of the average R-R interval, the effect of the former in determining the latter is minimized.
Claims (13)
1. An ECG monitoring system for generating a signal indicating a ventricular premature beat comprising means responsive to an ECG waveform for providing a measure proportional to the area of the QRS complex of said ECG waveform, means responsive to the ECG waveform for providing a measure proportional to the peak height of the QRS complex of said ECG waveform, and means for dividing said area measure by said peak height measure thereby to provide a normalized measure proportional to the width of the QRS complex of said ECG waveform.
2. The ECG monitoring system of claim 1 further including means for deriving the average width measure over a plurality of successive ECG waveforms and means for comparing a said width measure with said average width measure derived from previous width measures and in response to said width measure exceeding said average width measure by a predetermined amount for registering a particular wide pulse condition for the respective QRS complex.
3. The ECG monitoring system of claim 2 wherein said predetermined amount by which said width measure must exceed said average width measure for registering said particular wide pulse is substantially greater than 20%.
4. The ECG monitoring system of claim 1 including filter means principally responsive to frequencies below those characteristically present in the QRS complex of an ECG waveform, whereby the impulse response of said filter means to the QRS
complex is proportional to the energy and thus the area of the respective QRS complex, and means responsive to the peak magnitude of said filter impulse response for the respective QRS complex for providing said area measure.
complex is proportional to the energy and thus the area of the respective QRS complex, and means responsive to the peak magnitude of said filter impulse response for the respective QRS complex for providing said area measure.
5. The ECG monitoring system of claim 4 wherein said peak height measure is provided by peak detecting means responsive to the QRS complex of an ECG waveform for storing the maximum amplitude thereof.
6. The ECG monitoring system of claim 3 further including means for comparing a said width measure with said average width measure derived from previous width measures and in response to said width measure exceeding said average width measure by an other amount less than said predetermined amount for registering a different particular wide pulse condition, means for providing a measure of the interval between successive QRS
complexes, means for comparing said interval measure with an average of said interval measures over a plurality of immediately preceding QRS complexes and generating an "early" signal if said interval is less than about 90% of said average interval measure or a "late" signal if said interval is greater than about 110% of said average interval measure, and logic means responsive to said different particular wide pulse condition and a commensurate "early" signal and a "late" signal for the next occurring QRS
complex for generating a ventricular premature beat signal.
complexes, means for comparing said interval measure with an average of said interval measures over a plurality of immediately preceding QRS complexes and generating an "early" signal if said interval is less than about 90% of said average interval measure or a "late" signal if said interval is greater than about 110% of said average interval measure, and logic means responsive to said different particular wide pulse condition and a commensurate "early" signal and a "late" signal for the next occurring QRS
complex for generating a ventricular premature beat signal.
7. The ECG monitoring system of claim 2 wherein said means for determining said average width measure includes averaging means receiving said width measures of successive ECG waveforms for averaging, and attenuating means connected intermediate said dividing means and said averaging means, said attenuating means being responsive to the occurrence of a said particular wide pulse condition for attenuating the amplitude of the respective said width measure by a predetermined amount, thereby to reduce its effect on the average.
8. The ECG monitoring system of claim 7 wherein said means for determining said average width measure includes averaging means receiving said width measures of successive ECG waveforms for averaging, and attenuating means connected intermediate said dividing means and said averaging means, said attenuating means being responsive to the occurrence of a said particular wide pulse condition for attenuating the amplitude of the respective said width measure by a first predetermined amount and to the occurrence of a said different particular wide pulse condition for attenuating the amplitude of the respective said width measure by a second predetermined amount, thereby to reduce the different wide pulse effect on the average.
9. In an ECG monitoring system, the method of indirectly providing a measure of the width of a QRS complex appearing in an ECG waveform comprising the steps of:
generating from the ECG waveform a signal proportional to the area of the QRS complex;
generating from the ECG waveform a signal proportional to the peak height of the QRS complex; and dividing said area signal by said peak amplitude signal to provide a normalized signal representative of the width of the QRS complex of said ECG signal.
generating from the ECG waveform a signal proportional to the area of the QRS complex;
generating from the ECG waveform a signal proportional to the peak height of the QRS complex; and dividing said area signal by said peak amplitude signal to provide a normalized signal representative of the width of the QRS complex of said ECG signal.
10. The ECG monitoring system of claim 1 wherein said means for providing a measure proportional to the peak height of the QRS complex includes means for providing a value represen-tative of the peak height of at least one R-wave in an immediately preceding limited plurality of R-waves, said system further including variable scaling means for scaling one of the instant said ECG waveform and seak peak height representing value to be a controllably variable percentage of the other, means for com-paring said scaled one of the instant said ECG waveform and said peak height representing value with the other and indicating occurrence of an R-wave when the instant ECG waveform-based signal exceeds the peak height representing value based-signal;
and means response to a said indication of an R-wave for controlling variation of said scaling percentage.
and means response to a said indication of an R-wave for controlling variation of said scaling percentage.
11. The ECG monitoring system of claim 10 wherein said variable scaling means operates to scale said peak height representing value to provide a threshold signal, said instant ECG waveform being compared with said threshold signal to provide said R-wave indication when the former exceeds the latter.
12. The ECG monitoring system of claim 11 wherein said means for varying the percentage that the threshold signal is of the peak height representing value from which it is derived operates to rapidly increase said percentage to a predetermined maximum value selected to place said threshold substantially above substantially all T-waves in the ECG signal and, following peaking of the instant R-wave, operates to decrease said per-centage at a rate slower than said increase toward a predeter-mined minimum value selected to detect R-waves having peak amplitudes substantially less than said peak height representing value.
13. The ECG monitoring system of claim 12 wherein said means for establishing said percentage of said threshold signal comprises a resistive voltage divider having said peak height representing value operatively applied to one end thereof and said threshold value being tapped from said divider, and said percentage varying means include capacitor means and electronic switching means each connected in parallel with the one end and the tap of said divider, said switching means normally being non-conductive and being responsive to said R-wave indication to become conducting throughout its duration, said R-wave indication existing so long as the instant ECG signal exceeds said threshold.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US05/883,096 US4181135A (en) | 1978-03-03 | 1978-03-03 | Method and apparatus for monitoring electrocardiographic waveforms |
| US883,096 | 1978-03-03 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1123919A true CA1123919A (en) | 1982-05-18 |
Family
ID=25381975
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA322,460A Expired CA1123919A (en) | 1978-03-03 | 1979-02-28 | Method and apparatus for monitoring electrocardiographic waveforms |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US4181135A (en) |
| JP (1) | JPS54124591A (en) |
| CA (1) | CA1123919A (en) |
| DE (1) | DE2905407A1 (en) |
| GB (1) | GB2016152B (en) |
Families Citing this family (39)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2425231A1 (en) * | 1978-05-12 | 1979-12-07 | Cm Ind | PROCESS AND DEVICE FOR PROCESSING AND DISPLAYING A READING SIGNAL OF AN ELECTROCARDIOGRAM RECORDING READ AT A SPEED HIGHER THAN THE RECORDING SPEED |
| US4356825A (en) * | 1978-08-21 | 1982-11-02 | United States Surgical Corporation | Method and system for measuring rate of occurrence of a physiological parameter |
| US4367752A (en) * | 1980-04-30 | 1983-01-11 | Biotechnology, Inc. | Apparatus for testing physical condition of a subject |
| US4510945A (en) * | 1982-07-12 | 1985-04-16 | Cordis Corporation | P Wave detection system |
| US4630204A (en) * | 1984-02-21 | 1986-12-16 | Mortara Instrument Inc. | High resolution ECG waveform processor |
| JPS6399840A (en) * | 1986-10-17 | 1988-05-02 | テルモ株式会社 | Bio-signal measuring apparatus |
| US4784153A (en) * | 1986-11-12 | 1988-11-15 | Marks Lloyd A | Method of and apparatus for detecting cardiac rhythm disturbance |
| US4998535A (en) * | 1989-09-05 | 1991-03-12 | Univ. of Washington New England Medical Center Hospitals, Inc. | Thrombolysis predictive instrument |
| US5276612A (en) * | 1990-09-21 | 1994-01-04 | New England Medical Center Hospitals, Inc. | Risk management system for use with cardiac patients |
| FR2685624B1 (en) * | 1991-12-31 | 1994-03-11 | Ela Medical | CARDIAC ACTIVITY ANALYSIS SYSTEM FOR AN IMPLANTABLE DEVICE FOR TREATING TACHYCARDIAS. |
| US5312441A (en) * | 1992-04-13 | 1994-05-17 | Medtronic, Inc. | Method and apparatus for discrimination of ventricular tachycardia from supraventricular tachycardia and for treatment thereof |
| JPH06327644A (en) * | 1993-05-18 | 1994-11-29 | Fukuda Denshi Co Ltd | Analysis device for electrocardiogram information |
| FR2705576B1 (en) * | 1993-05-28 | 1995-07-07 | Ela Medical Sa | A method of analyzing cardiac activity to determine whether a tachyarrhythmia is likely to be stopped by stimulation. |
| US5724983A (en) * | 1994-08-01 | 1998-03-10 | New England Center Hospitals, Inc. | Continuous monitoring using a predictive instrument |
| US5501229A (en) * | 1994-08-01 | 1996-03-26 | New England Medical Center Hospital | Continuous monitoring using a predictive instrument |
| JP3465133B2 (en) * | 1997-08-05 | 2003-11-10 | 日本光電工業株式会社 | Patient monitoring device |
| US5876349A (en) * | 1997-08-08 | 1999-03-02 | Hewlett-Packard Company | Method and apparatus for ventricular fibrillation detection |
| US5957857A (en) * | 1998-05-07 | 1999-09-28 | Cardiac Pacemakers, Inc. | Apparatus and method for automatic sensing threshold determination in cardiac pacemakers |
| US6070097A (en) * | 1998-12-30 | 2000-05-30 | General Electric Company | Method for generating a gating signal for cardiac MRI |
| US6766189B2 (en) * | 2001-03-30 | 2004-07-20 | Cardiac Pacemakers, Inc. | Method and apparatus for predicting acute response to cardiac resynchronization therapy |
| US6993389B2 (en) * | 2001-03-30 | 2006-01-31 | Cardiac Pacemakers, Inc. | Identifying heart failure patients suitable for resynchronization therapy using QRS complex width from an intracardiac electrogram |
| US6705999B2 (en) * | 2001-03-30 | 2004-03-16 | Guidant Corporation | Method and apparatus for determining the coronary sinus vein branch accessed by a coronary sinus lead |
| US20030032871A1 (en) * | 2001-07-18 | 2003-02-13 | New England Medical Center Hospitals, Inc. | Adjustable coefficients to customize predictive instruments |
| US20050059880A1 (en) * | 2003-09-11 | 2005-03-17 | Mathias Sanjay George | ECG driven image reconstruction for cardiac imaging |
| US8668653B2 (en) | 2004-03-24 | 2014-03-11 | Nihon Kohden Corporation | Biological information measuring garment having sensor, biological information measuring system and equipment, and control method of equipment |
| JP4788915B2 (en) | 2004-03-24 | 2011-10-05 | 日本光電工業株式会社 | Biological information measuring garment having electrodes and biological information measuring system |
| US7415304B2 (en) * | 2004-04-15 | 2008-08-19 | Ge Medical Systems Information Technologies, Inc. | System and method for correlating implant and non-implant data |
| US7072709B2 (en) * | 2004-04-15 | 2006-07-04 | Ge Medical Information Technologies, Inc. | Method and apparatus for determining alternans data of an ECG signal |
| US7509159B2 (en) * | 2004-04-15 | 2009-03-24 | Ge Medical Systems Information Technologies, Inc. | Method and apparatus for detecting cardiac repolarization abnormality |
| US7187966B2 (en) * | 2004-04-15 | 2007-03-06 | Ge Medical Systems Information Technologies, Inc. | Method and apparatus for displaying alternans data |
| US7162294B2 (en) | 2004-04-15 | 2007-01-09 | Ge Medical Systems Information Technologies, Inc. | System and method for correlating sleep apnea and sudden cardiac death |
| US7272435B2 (en) * | 2004-04-15 | 2007-09-18 | Ge Medical Information Technologies, Inc. | System and method for sudden cardiac death prediction |
| US20050234353A1 (en) * | 2004-04-15 | 2005-10-20 | Ge Medical Systems Information Technologies, Inc. | Method and apparatus for analysis of non-invasive cardiac parameters |
| EP1814451B1 (en) * | 2004-10-25 | 2018-01-10 | Tai Chuan Alfred Kwek | A system and apparatus for detecting a cardiac event |
| US7283864B2 (en) * | 2005-02-10 | 2007-10-16 | Cardiac Pacemakers, Inc. | Method and apparatus for identifying patients with wide QRS complexes |
| US7751876B2 (en) * | 2005-09-23 | 2010-07-06 | Hewlett-Packard Development Company, L.P. | Method and system for detecting premature ventricular contraction from a surface electrocardiogram |
| US20090072866A1 (en) * | 2007-09-19 | 2009-03-19 | Neil Edward Walker | Method and system for controlling amplified signals reflecting physiological characteristics |
| US20100030061A1 (en) * | 2008-07-31 | 2010-02-04 | Canfield Monte R | Navigation system for cardiac therapies using gating |
| US9675270B2 (en) | 2015-04-23 | 2017-06-13 | Medtronic, Inc. | Method and apparatus for determining a premature ventricular contraction in a medical monitoring device |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3267934A (en) * | 1962-09-20 | 1966-08-23 | Avionics Res Products Corp | Electrocardiac computer |
| US3352300A (en) * | 1964-10-28 | 1967-11-14 | Fred A Rose | Cardiac monitor |
| US3518983A (en) * | 1967-10-03 | 1970-07-07 | Humetrics Corp | Arrhythmia detector and method of operation |
| US3524442A (en) * | 1967-12-01 | 1970-08-18 | Hewlett Packard Co | Arrhythmia detector and method |
| JPS4944513B1 (en) * | 1968-05-20 | 1974-11-28 | ||
| GB1282051A (en) * | 1969-03-14 | 1972-07-19 | James Mcewan Mcintyre Neilson | Apparatus for monitoring recurrent waveforms |
| US3616790A (en) * | 1970-01-21 | 1971-11-02 | American Optical Corp | Multiform ventricular premature beat detector |
| US3699946A (en) * | 1970-08-11 | 1972-10-24 | Walter A Michel | Waveform abnormality monitor |
| US3903874A (en) * | 1973-08-27 | 1975-09-09 | Mediscience Technology Corp | Cardiographic signal processing means and method |
| US3939824A (en) * | 1973-10-09 | 1976-02-24 | General Electric Company | Physiological waveform detector |
| US3998214A (en) * | 1975-05-19 | 1976-12-21 | Brondy Laboratories, Inc. | Premature ventricular contraction detector and method |
-
1978
- 1978-03-03 US US05/883,096 patent/US4181135A/en not_active Expired - Lifetime
-
1979
- 1979-02-09 DE DE2905407A patent/DE2905407A1/en not_active Withdrawn
- 1979-02-28 CA CA322,460A patent/CA1123919A/en not_active Expired
- 1979-03-01 JP JP2259679A patent/JPS54124591A/en active Pending
- 1979-03-02 GB GB7907464A patent/GB2016152B/en not_active Expired
Also Published As
| Publication number | Publication date |
|---|---|
| JPS54124591A (en) | 1979-09-27 |
| GB2016152A (en) | 1979-09-19 |
| GB2016152B (en) | 1982-09-02 |
| US4181135A (en) | 1980-01-01 |
| DE2905407A1 (en) | 1979-09-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA1123919A (en) | Method and apparatus for monitoring electrocardiographic waveforms | |
| US4240442A (en) | Variable threshold R-wave detector | |
| US4211237A (en) | Method and apparatus for identifying recurring signal patterns | |
| US3654916A (en) | Apparatus for monitoring recurrent waveforms | |
| US3940692A (en) | Apparatus for monitoring recurrent waveforms | |
| US6438411B1 (en) | Digital ECG detection system | |
| US3828768A (en) | Method and apparatus for detecting cardiac arrhythmias | |
| US4000461A (en) | R-wave detector | |
| US4379460A (en) | Method and apparatus for removing cardiac artifact in impedance plethysmographic respiration monitoring | |
| US5170794A (en) | Method and apparatus for deriving a respiration signal and/or artifact signal from a physiological signal | |
| US4478224A (en) | Artifact detector for heartbeat rate measuring system | |
| US5117833A (en) | Bi-spectral filtering of electrocardiogram signals to determine selected QRS potentials | |
| US5660184A (en) | Pacemaker pulse detection and artifact rejection | |
| US4459993A (en) | Continuity detector for heartbeat rate measuring system | |
| JP3319140B2 (en) | Heart rate variability waveform analysis method and apparatus | |
| WO1989001312A1 (en) | Diastolic clamp for bioimpedance measuring device | |
| DE3249490T1 (en) | Heart rate monitor | |
| WO2018082190A1 (en) | Ecg signal processing method and apparatus | |
| JP3171257B2 (en) | Method and apparatus for removing baseline fluctuations from an electrocardiogram | |
| US4446868A (en) | Cardiac arrhythmia analysis system | |
| US4237903A (en) | QRS detector for EKG signals | |
| JP4121611B2 (en) | Cardiac response detection system | |
| JP3314521B2 (en) | Heart rate variability waveform analysis method and apparatus | |
| US6115629A (en) | Two electrode heart rate monitor measuring power spectrum for use with exercise equipment | |
| EP0971627A1 (en) | Method of analysing a cardiac signal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MKEX | Expiry |