Note: Descriptions are shown in the official language in which they were submitted.
1138531
CARDIAC ARRHYTHMIA
DETECTION AND RECORDING
The present invention is directe~ to monitoring and
recording apparatus and methods and, more particularly, to a
method and apparatus for monitoring the heart activity of a
cardiac patient, detecting cardiac arrhythmias and recording
such arrhythmias in real time for later analysis.
Prior art apparatus of the subject type are illus-
trated in the following United States Patents: 3,759,248
3,824,990; 3,832,994; 3,858,034; 3,861,387 and 4,023,564.
Reference may also be had to Kosowsky, "Holter Monitoring~,
Journal of Continuing Education in CardioloqY~ Vol. 14, No.
2, February 1978, pages 13-21; and Dreifus et al, "Newer Tech-
niques in Cardiac Monitoring", Heart and Lunq, July-August
1975, pages 568-572.
Desirability of providing electronic apparatus for
automatically detecting cardiac arrhythmias has heretofore
been recognized, as exemplified in the above-noted prior art.
Similarly, it has been recognized as desirable to provide
combined detection and recording apparatus as a single unit
which may be worn by a cardiac outpatient to monitor heart
activity while the patient follows his daily routine. Such
portable apparatus should be compact, rugged and lightweight,
and yet responsive to a wide variety of cardiac arrhythmia
types. Similarly, the apparatus should be responsive to
1. ~
1138531
critical arrhythmias indicative of cardiac pathology without
being sensitive to extraneous false positives caused by noise
or noncardiac muscular activity. The techniques and apparatus
proposed in the prior art are not considered to satisfy the
above-noted and other desirable, and in some cases critical,
features in apparatus of the subject type.
Moreover, some portable apparatus which have achieved
some commercial acceptance contemplate continuous recording of
heart activity for a specific time duration, su3ch as twenty-
four hours. Where the activity is recorded on a magnetic tapecassette, for example, the cassette must be played bac~ and
monitored on a CRT by a skilled technician for pathological
events. Such a technique is very expensive and does not
achieve optimum reliability. Other prior art apparatus con-
lS template only manually-activated recording when a wearer
thinks that he is experiencing unusual heart activity.
Accordingly, it is a general object of the present
invention to provide improved cardiac monitoring apparatus
which overcomes and satisfies the foregoing problems and dif-
ficulties.
More specific objects of the invention are toprovide cardiac monitoring apparatus which may be worn by
a cardiac outpatient without substantially interfering with
his daily routine, which is automatically responsive to a
wide variety of cardiac pathological events, which records
113~3531
such events in real time together with preceding and/or sub-
sequent ~normal~ heart activity for later analysis, which
simplifies arrhythmia detection techniques by recording only
clinically significant events while minimizing false spurious
S recording, which significantly improves the effectiveness and
efficiency of a battery-operated portable cardiac recorder,
which has an enhanced signal-to-noise ratio for reduced sen-
sitivity to extraneous or false positives, and/or which in-
cludes provision for manual activation of the recorded by
the wearer both for recording what he considers to be un-
usual heart activity and for recording a ~normal" cardiac
signal at desired intervals for purposes of comparison.
A further object of the invention is to provide
cardiac arrhythmia detection circuitry which is particular-
ly well adapted for use in a portable cardiac monitoringapparatus and yet may be used to advantage in a full-scale
hospital cardiac ~are unit.
The present invention, together with additional
objects, fea~ures and advantages thereof, will be best
understood from the following description, the appended
claims and the accompanying drawings in which:
1138S31
FIG. 1 is a functional block diagram of a presently
preferred embodiment of the cardiac arrhythmia detection and
recording apparatus provided by the invention; and
.FIGS. 2-9 are graphical waveforms useful in
understanding operation of the invention, FIGS. 3-9 being
drawn to scale.
FIG. 1 illustrates a presently preferred embodi-
ment of the detection and recording apparatus 20 provided
by the invention as comprising a differential input ampli-
fier 22 connected to conventional cardiac electrodes 24suitably positioned on a patient 26. The output of ampli-
fier 22 is eonneeted through a high pass filter 28 to the
input of a differentiating amplifier 30. The output of
differentiating amplifier 30 is eonneeted through respective
zero-erossing detection circuits 31,33 to the inputs of a
QRS duration eounter 32 and an R-R interval counter 34.
Counters 32, 34 also have a counting input conneeted to-a
one-kilohertz clock oseillator 36. One output of counter
32 is connected to a QRS duration average computation cir-
euit 38 which preferably receives and stores a selectednumber of suceessive QRS duration signals from counter 32
and computes therefrom a running average Q~S duration. In
a preferred embodiment of the invention, averager 38 is re-
~ponslve to the QRS components of the three cardiac rhythm
signals immediately preceding the duration signal stored in
li38S31
counter 32. The outputs of ave.ager 38 and counter 32 are
connected to a comparator 40 which provides a first signal
to an OR gate 42 when the duration indicated in counter 32
is greater or less than the running average duration indica-
ted by averager 38 by an amount equal to or greater than
twenty percent of the running average.
Similarly, R-R interval counter 34 is connected
to an R-R interval averager 44 for computing a running
average R-R interval over a plurality of, prefera~ly three,
immediately preceding QRS cardiac rhythm signals. Counter
34 and averager 44 are connected to a comparator 46 which
provides a second signal to OR gate 42 when the R-R interval
indicated by counter 34 is greater or less than the running
average R-R interval indicated by averager 44 by an amount
equal to or greater than thirty percent of the average
interval. OR gate 42 provides control signal to a tape
drive circuit 48 in response to a signal from either compara-
tor 40 or comparator 46, or both.
The output of input amplifier 22 is also connected
20 through an analog-to-digital convertor 50 which provides
digital signals in real time to a shi~t register memory 52
indicative of the amplitude of the cardiac rhythm signal.
The gate input of memory 52 is connected to clocX oscillator
36~ Prefera~ly, memory 52 comprises a 3000-word serial
shift register memory which, when combined with a one-kilo-
hertz clock oscillator fre~uency, provides diqital signals
113~531
at the memory output as a sampled replica of the cardiac
rhythm input signals effectively delayed by a period of
three seconds. Clock oscillator 36 is also connected to the
sampling or gating input of a time and date memory circuit
54. The outputs of memories 52,54 are connected through
correspondingly respective digital-to-analog convertors
53,55 to two input channels of a three-channel tape re-
corder circuit 56 which is responsive to a control signal
from tape drive 48 to record the analog sig,nals from con-
vertors 53,55 on two channels of a cassette tape illustrated
schematically at 58. A manual record switch 60 which may
be activated by the patient 26 has one output connected to
a tape drive 48 and a second output connected to a third
channel o~ tape recorder 56.
The output of tape drive 48 is a retriggerable
pulsed signal 62 of predetermined duration, preferably on
the order of six seconds. Thus, when the tape drive is
activated by an control signal from OR gate 42, cardiac
rhythm signals are recorded on cassette 58 which effecti-ve-
2Q ly bracket the detected pathological event. Stated diff~r-
ently, the delayed output from memory 52 precedes the de-
tected cardiac arrhythmia by approximately three seconds,
while the tape drive signal 62 has a duration of about six
seconds, so that the recorded rhythm signals includes approx-
imately three seconds preceding and three seconds following
the pathological event. Time and date of the detected event
indicated by memory 54 are recorded on a second tape channel.
~138531
When tape drive 48 is activated ~y manual record switch 60,
a six-second cardiac rhythm signal is recorded, together
with time and date on the second tape channel and an indica-
tion on a third tape channel that the recording was initiated
by the manual record switch. If a second arrhythmia is de-
tected during the recording interval, pulsed signal 62 is
retriggered and the recording will continue for an additional
six seconds. The entire circuit, including the tape drive,
is battery opera~ed.
Overall operation of the preferred embodiment o~
the invention illustrated in FIG. 1 will be evident from the
foregoing discussion and need only be outlined briefly in
connection with a "typical" cardiac rhythm signal of the type
illustrated in FIG. 2. The rhythm signal of FIG. 2 includes
an atrial P component which has a small positive amplitude,
as on the order of 50 to 100 microvolts, and a relatively
short duration, as on the order of forty to eighty milli-
seconds. Thereafter, following a brief interval of quie-
scence on the order of 150 milliseconds, the signal cycles
through a QRS complex corresponding to depolarization of the
cardiac muscle in which the signal swings briefly negative
in the Q component, then a relatively sharp positive spike
of about one millivolt in the R component, and thereafter
through a brief negative swing in the S component. A nomi-
nal formal QRS duration of 100 milliseconds is typical~ Afteranother brief quiescent interval on the order or 200 milli-
seconds, a slight positive swing corresponding to the T com-
ponent indicates repolarization of the cardiac m~scle. The
li3~S31
interval between cardiac rhythm signals is the inverse of
the pulse rate and would be one second, for example, for
a typical cardiac rhythm at sixty beats per minute.
The frequency cutoff of high pass filter 28 is
preferably selected so as to block the low frequency P and
T signal components from amplifier 30, the arrhythmia detec-
tion circuitry thereby being responsive solely to the QRS
rhythm signal components. The input to differential am-
! plifier 30 is set to be responsive to signals above a mini-
mum threshold, and thereby cooperates with filter 28 to
block low frequency and/or low voltage high frequency noise
signals gene~ated by patient muscular activity. It will be
a~preciated that the arrhythmia detection circuitry is re-
sponsive to the derivative of the input signal via amplifier
3~, and is essentially independant of signal amplitude. Thus,
filter 28 and amplifier 30 cooperates to overcome problems
inherent in prior art techniques which are responsive in whole
or in part to the low frequency and low voltage P and T signal
components and/or to QRS signal amplitude.
Zero crossing detection circuit 31 is preferably
responsive to the "peak" of the first pulse co~ponent, i.e.
the Q component in FIG. 2 with the P component having been
blocked by filter 28 and amplifier 30. QRS duration counter
begins counting at this first "peak" and continues counting
until the differentiated signal stabalizes at zero, i,e.
at the end o~ the S signal co~ponent. Circuit 33 is respon-
sive to the second zero crossing of successive differentiated
113~531
rhythm signals, i.e. to the peak of the ~ signal component
illustrated in FIG. 2. The Q~S duration in counter 32 is
compared with the average duration in averager 38 and is
thereafter loaded into the averager between QRS rhythm sig-
nals for maintaining the running average. The R-R interval
counter 34 is responsive to detection circuit 33 sequentially
to stop the preceding R-R interval count, compare such count
with the running average interval in averager 44, load the
preceding count into averager 44 for maintaining a running -,
interval average and then beginning the succeeding R-R
interval count. All of such operations are, of course,
performed in microseconds and have no practical effect upon
the successive R-R interval count.
While arrhythmia monitoring is taking place as
described, all of the P, Q, ~, S and T components are being
continuously sampled and loaded into shift register memory
52 through convertor 50 for later recording if required.
When analysis of recorded information is desired, cassette
58 may be taken to a clinic and played bacX on a tape
reader connected to a conventional EKG strip chart recorder.
The r~sult will be a series of permanently recorded cardiac
rhythm signals exhibiting potential pathological events,
each preceded by a corresponding time signal, which may be
analyzed by a medic~l clinician. Th~s, the invention
both provides for automatic portable recording and
eliminates any re~uirement for scanning of several hours
recording searching for potentially pathological events.
~138531
It will be appreciated by persons skilled in the
analysis and treatment of cardiovascular disorders that
FIG. 2 illustrates a somewhat idealized rhythm signal for
a healthy individual. For patients exhibiting some form
of cardiac disorder the rhythm signal may vary substantial-
ly from that shown in FIG. 2. Indeed, for patients who
have suffered permanent heart damage, one or more Gf the
Q, R and S components may be exaggerated or obliterated. Thus,
for purposes of the present description and the appended
claims, the term "QRS" with reference to rhythm signal com-
ponents must be reaa in its broadest aspects as represent-
ing the signal which results from the patient's electrical
ventricular depolarization. Similarly, the term "R-R in-
terval" signifies the interval between successive rhythm
signals, and need not necessarily be measured ~etween
identifiable "R" signal components.
In accordance with an important feature of the
invention, the arrhythmia detection circuitry, as distin-
guished from the recording circuitry, is resp~nsive only
to the ventricular depolarization or QRS signal and blocks
or ignores the atrial depolarization or P component and
the ventricular repolarization or T component. ~The at-
rial repolarization signal is masked by the Q~ signal
and, in any event, is of sufficiently low frequency as to
~e blocked by filter 28~) An important eature of the
present invention lies in recognition of the fact that all
10 .
1138531
clinically significant arrhythmias may be detected using only
the ventricular depolarization or QRS signal components, namely
QRS duration and R-R interval. In this connection, arrhythmia
detection for initiating recording must be distinguished from
arrhythmia diagnosis, the latter requiring analysis of all
signal components including P and T. All of the P, Q, R, S
and T components are recorded for diagnosis on cassette 58
via convertors 50,53 and memory 52 whenever an arrhythmia
is detect,ed. According to one important aspect of the present
invention, it has been recognized that all clinically signi-
ficant cardiac events which must be diagnosed by reference
to the P and/or T signal components vary sufficiently in QRS
duration and/or R-R interval to be detected by the present
invention without monitoring the P and T components per se,
and thereby eliminating the source of problems inherent in
prior art systems that attempt to detect P and T components
directly. Thus, the present invention not only simplifies
detection techniques but also improves reliability problems
inherent in prior art techniques which attempt to detect
arrhythmias by, in e~fect, diagnosing various arrhythmia
types.
In accordance with another important feature of
the invention, it has ~een recognized that it is not so
much the absolute value of the duration of or interval
between QRS signals for a particular patient that is im-
portant as measured against fixed standards, as is the
relative duration or interval for the particular patient
1138531
as compared to what is "normal" for him. Indeed, a patient
who has suffered heart damage may have a "normal" QRS dura-
tion and R-R interval which departs significantly from the
above-noted times for a nominally healthy p~orson. Compari-
son of a patient's QRS duration and R-R interval in the
present invention to corresponding running average measure-
ments effectively provides arrhythmia detection for each in-
dividual patient by comparing each cardiac event to what is
"normal" for him.
Although it may be desirable to measure the actual
duration of the QRS signal from the beginning of the Q com-
ponent, it is convenient and pre3ently preferred to begin
measurement at the "peak" of the Q signal utilizing a dif-
ferentiated signal and conventional zero crossing detection
circuitry. As noted above, it is not the absolute duration
of the QRS signal that is important hut comp~rison of each
duration to past durations. Thus, the first half of the Q
component may be ignored so long as ~RS duration begins and
ends in the same manner for each successive rhythm signal.
Similarly, R-R interval may be measured even wnen there is
no discernable R signal component so long as measurement
takes place in a similar manner for each successive rhythm
signal. In some cases, "R-R interval" may in fact be an
S-S interval, for example.
Tlle use of a running average for co~parison of QRS
duration and R-R interval has the advantage of tracking rela-
tively slow changes in cardiac rnythm associated with changes
1138531
in patient activities without giving false alarm indications.
Average computation using three duration counts and three in-
terval counts has been empirically selecte~ in the present in-
vention as providing reliable results of desired sensitivity
at low cost. However, greater or possibly even fe-~r counts
could be utilized in co~puting average Q~S duration and/or
R-R interval without departing from the invention, although
it is presently believed that at least three counts are re-
quired. The allo~able range of plus or minus twenty percent
in QRS duration comparator 40 is recognized in the art as a
convenient cutoff point for detection of p~tholo~ical events.
The plus or minus thirty percent range in R-R interval com-
parator 46 has been empirically selected as providing reli-
able indications of actual pathological events without being
overly sensitive to nor~nal variations in cardiac rhythm.
other comparator ranges for QRS duration and R-R interval may
be utilized where desired.
The foregoing and other functional features of the
invention may be further appreciated with reference to FIGS.
3-9 which illustrates various "typical~' pathological e-~ents,
all of which may be detected and recorded by the preferred
embodiment of the invention illustrated in F~G. 1. Each
set of cardiac rhythm signals in FIGS. 3-9 is six seconds in
duration and is drawn to scale. FIG. 3 illustrates a "norm~l
sinus rhythm" at a relatively slow rate of fifty-four beats
per minute interrupted by a single premature ventricular con-
traction (PVC) at 62. The normal R-R interval in FIG. 3 is
1138S3~
1.36 seconds. The premature ventricular contraction at 62
occurs 560 milliseconds following the preceding R signal
- component, which is well below seventy percent of the average
R-R interval (950 milliseconds) to which comparator 46 is
responsive. Additionally, the QRS duration of the prei~ature
ventricular contraction is 170 milliseconds, which is sub-
stantially greater than one hundred twe~ty percont of the
average Q~S duration of eighty milliseconds. Thus, in this
example, OR gate 42 is energized by both'comparators 40
and 46.
This combination of a thirty percent R-R interval
range with use of a running average interval for comparison
purposes has a p~rticularly synergistic effect in detection
of PVC's. More specifically, a PVC may be just within the
thirty percent range and therefore not be detected upon
occurrence. However, the same PVC will so effect the run-
ning average of three preceding R-R intervals that gate 42
will be activated by the next QRS co,nplex. It is felt that
the R-R interval range of thirty percent will alone detect
ninety percent of all clinically significant PVC's, and
that the use of the running average f or comp~rison purposes
will pick up the other ten percent.
FIG. 4 illustrates an electrocardiographic abnor-
mality co;nprising fre~uent unifocal premature ventricular
contractions with a configuration of so-called "ventricular
trigeminy". The normal R-R interval in this cardiac rhythm
is 600 milliseconds, w~ile the premature R--R interval is 500
1138531
milliseconds. Thus comparator 46 is not activated. However,
the QRS duration of the premature ventricular contractions
at 64 is 120 milliseconds, which is substantially greater
than 120 percent of the normal QRS duration fifty milliseconds.
Thus OR gate 42 is activated by comparator 40.
FIG. 5 illustrates a cardiac rhythm abnormality
called "atrial fibrillation~. In this relatively common
abnormality, the patient's atrial activity indicated by the
P signal components are chaoti~ and difficult to discern.
The R-R intervals are constantly changing in the illustrated
ventricular rhythm from 155 to 170 beats per minute, with an
average of about 160 beats per minute. Although the R-R
intervals are irregular, they do not depart from the av-
erage by plus or minus thirty percent and therefore do not
indicate an alarm condition. Thus, the present invention
may be used in a patient with atrial fibrillation but reason-
ably consistent ventricular response without constantly in-
dicating an alarm condition. A premature ventricular con-
traction as illustrated at 66 in FIG. 5 having a duration of
120 milliseconds, as compared with a seventy millisecond
average QRS duration, provides an alarm signal through com-
parator 40 and gate 42.
FIG. 6 illustrates a pattern of normal sinus rhythm
interrupted by frequent premature atrial contractions 68. The
QRS durat-ion of the premature atrial contractions are usually
well within normal and relatively constant. ~lowever, each
premature atrial contraction occurs in the range of forty
~13~531
to two hundred forty milliseconds after the preceding signal
complex, which is substantially less than the running average
760 milliseconds R-R interval. Each premature atrial contrac-
tion, therefore, activates comparator 46.
FIGS. 7 and 8 illustrate examples of seco~d degree
heart blockage in which the ventricular cardiac muscle re-
sponds inconsistently to electrical atrial activity, which
is to say that each P wave is not followed by a QRS complex.
In sec,ond degree heart blockage of the Wenckebach variety
illustrated in FIG. 7, the P-R intervals progressively
increase until a QRS complex is dropped or "blocked", where-
upon the P-R interval again shortens. FIG. 7 illustrates
a four-to-three block, i.e., one wherein four P wave com-
ponents result in only three QRS complexes. The QRS com-
plexes are not distinguished by abnormally prolonged dur-
ation, but the rate change is significant in that the R-R
interval containing the missed QRS complex is almost twice
as long as the average preceding intervals. Thus as illus-
trated in FIGS. 6 and 7 combined, the present invention is
2~ adapted to be responsive both to short and long R-R intervals
indicative, respectively, of premature and blocked heart
rhythms. FIG. 7 also illustrates the aforementioned feature
of the invention whereby a clinically significant arrhythmia
diagnostically related to the P singal compone~t i5 detected,
and recorded for later diagnosis, without attempting to moni-
tor the low voltage and frequency P component per se. ~t will
16.
~38531
also be recognized that the present invention is adapted to
respond to Q, R and S signals of "normal" polarity, as well
as complexes of opposite polarity of the type illustrated in
FIG. 7-
FIG. 8 illustrates a two-to-one second degree heart block
with a period in mid-strip which displays a one-to-one
response. Since the patient's normal rhythm displays a
two-to-one heart block, the brief one-to-one response is,
in fact, double the patient's average rate and an alarm
signal is generated.
FIG. 9 displays a commonly encoùntered problem
with a patient wearing an intermittently functioning or
demand-type electronic pacemaker. The QRS duration follow-
ing a pacemaker "blip" is normally prolonged because of the
position of the pacemaker at the tip of the right ventricle.
The pacemaker spike, when first activated, provokes a
prolonged QRS duration at the onset of pacing, therefore
generating an alarm condition and recording the three-second
interval preceding pacing and the first three seconds after
the onset of pacing. Similarly, absence of the pacemaker spike
following turnoff or failure of the pacemaker, or failure of a
pacemaker spike to provoke a subsequent QRS results in a sensed
short QRS duration which would result in an alarm signal and
recording event. Filter 28 (FIG. 1) passes the high frequency
pacemaker spike to the arrhythmia detector circuitry. Where
use with a pacemaker is not contemplated, high pass filter 28
may conveniently be replaced by a band pass filter to bloc~
1138531
both the low frequency P and T slgnal components and high fre-
quency noise.
Counters 32,34, averagers 38,44, comparators 40,46,
gate 42, tape drive 48 and memories 52,54 may be provided in
the form of conventional low-cost digital integrated circuitry,
while the remainder, including amplifiers 22,30, filter 28
and convertors 50,53,55 may be provided in whole or in part
by readily available integrated linear circuit components.
Indeed, miniature cassette recorders which include both~tape
drive circuitry 48 and recording circuitry 56 are commercially
available. The integrated circuit components and the tape
drive and recording head may be readily provided in a compact
and rugged portable to be worn by a cardiac outpatient as in
the shirt pocket, for example. It will also be appreciated
that the recording fidelity and speed, typically on the order
of one-eighth inch per second, of a conventio~al cassette
recorder inherently performs a smoothing function between the
channel input and the corresponding recording head, thereby
eliminating any requirement for demodulation circuitry between
the sampled output of memory 52 and the channel input.
It will also be appreciated that the arrhythmia
detection circuitry provided by the invention and previously
described is responsive to the derivative of the cardiac
rhythm signal and is therefore substantially independent
of the amplitudes of the signal components. Thus the 5ig-
nal-to-noise ratio is substantially enhanced as compared
18.
1138S31
with prior art monitoring techniques which are responsive
in whole or in part to signal amplitude.
It will be appreciated that the present invention
offers particular advantages when utilized in a portable
battery operated unit since it is maintained in a non-record-
ing state until an arrhythmia is detected, at which time re-
cording begins but continues only for a short duration bracket-
ing the detected arrhythmia. The recorder then returns to
non-recording conditions unti,l the next detected arrhythmia.
This minimizes battering drain and, very importantly, allows
monitoring over long intervals, greater than 24 hours, since
the amount of tape used depends only upon the number and
frequenc~y of detected arrhythmias. These highly desirable
attributes are optimized by the present invention because,
as previously described, the present invention is based
in large part on the recognition that clinically significant
arrhythmias can be detected using QRS components (only), namely
Q~S duration and R-R interval, based on a differentiated wave-
form, with preselected ranges of permissable deviations from
a running average standard, in order to separate important
abnormal events. I~owever according to the present invention,
~S duration and ~-to-R interval are preferably, if not nec-
essarily, detected from this differentiated waveform for ex-
ample, in the case o~ QRS interval ~etection using techniques
of the type disclosed in U.S. Patents 3,552,386, 3,593,110,
3,616,791 and 3,903,873.
1~
1138S31
Although the present invention is particularly well
adapted for use in a compact and portable cardiac monitoring
units, the principles thereof, particularly the arrhythmia
detection circuitry embodied in filter 2~ to OR gate 42, may
readily be incorporated into full scale hospital coronary care
units. The invention is intended to embrace the above-noted
and all other alternatives, modifications and variations as
fall within the spirit and broad scope of the appended claims
The inv~.ntion claimed is:
li38531
SuPplemental Disclosure
FIG. 10 is a pictorial illustration of a
portable version of the invention worn by an ambulatory
patient; and
FI&. 11 is an enlarged front elevational
view of the portable apparatus in FIG. 10 with front
cover removed.
Referring to FIG. 10, portable monitor 80
in accordance with the invention comprises a case 82
of molded plastic or the like having a suitable clip
(not shown) for belt-attachment. Jacks 84 (FIG . 11),
such as 3.5 mm connectors, are provided in a side
wall of case 82 for connecting the electrode leads
23 to internal circuitry 20. The front cover of case
82 includes a latching window 86 (FIG. 10) for ob-
servation and removal of cassette tape 58. FI~. 11
illustrates portable monitor 80 with the front cover
removed. Electronic circuitry 20 (FIG. 1) is em~odied
in a printed circuit board assembly 81 mounted in-
ternally of case 82. Circuit assembly 81 is connected
by suitable leads (not shown) to recording head 58,
by leads 88 to jac~s 84 and by leads 90 to a battery
92, such as a nine volt transistor battery. In
accordance with an important feature of the inventîon,
intermittent recording only in response to arrhythmia
detection provides extended battery life as compared
~ ~3'
113853~
with continuously recording portable units of the
prior art. Additionally, digital shift register memory
52 provides the required continuous signal delaying
function with little power drain, particularly as
compared with apparatus embodying continuously operative
closed loop magnetic tape recording.
The invention is designed to be particularly
sensitive to perhaps the most dangerous of all arrhythmia
types, particularly in ambulatory patients, the pre-
mature ventricular contraction or PVC. FIG. 3 illustrates
a normal sinus rhythm at a relatively slow rate of
fifty-four beats per minute interrupted by a single
PVC 62. It i~ a characteristic of common PVCS that
it takes place with a shorter-than-average R-R interval
followed by a longer R-R interval called a compensatory
pause. The use of three R-R intervals in accordance
with a preferred embodiment of the invention, in
combination with incorporation of the PVC R-R interval
into the running average, insures that the PVC will
he detected either at the shortened R-R interval pre-
ceding the PVC or at the longer compensatory pause.
For example, it is possible that the R-R interval
preceding the PVC will only be 25% less than average
and, thus, would not trigger recording. However,
that shortened R-R interval is then incorporated
into the running average so that the succeeding com-
1138531
pensatory pause will be substantially greater than3~/O of average and recording will be triggered. The
use of four intervals for average computation purposes
would render the detection circuitry less sensitive,
while two intervals would render the circuitry overly
sensitive. For these rea~ons, three R-R intervals
in computation of the running average, in combination
with a _ 30% deviation for detection purposes and
incorporation of each rhythm signal whether normal
or arrhythmic, into the average, is preferred.
The use of three QRS durations for computation
of a running average duration is preferred because
the QRS duration circuitry thereby monitors the same
portion of the continuing rhythm ~ignal as does the
R-R interval circuitry. It will also be noted, that,
for normal sinus rhythms of 60 beats per minute or
greater, at least three R-R intervals and three QRS
durations preceding the detected arrhythmia will
be recorded in the preferred embodiment of the invention
wherein shift register 52 effectively delays the
rhythm signals for three seconds. An allowable range
of plus 20% of QRS duration is recognized in the art
as a convenient cut-off for detection of pat~ological
events~ particularly PVCs. Thus, PVC 62 in FIG. 3
will be detected both because the QRS duration of the
PVC is more than 20% above normal, and the compensatory
a c~? 3 -
~q
li38S31
pause following the PVC is more than 30/0 above the
PVC-modified running average R-R interval. Minus
20% in QRS duration permits the preferred embodiment
of the invention to detect unanswered pacemaker spikes,
as will be discussed in connection with FIG. 9.
O ~
