 |
Description  |
|
|
FIELD OF THE INVENTION
The present invention relates to a cardiac arrhythmia detection system for
use in an implantable stimulation device, such as an implantable
pacemaker, cardioverter or a defibrillator. More particularly, the present
invention relates to a system for detecting and tracking cardiac events
and electrogram morphology so that the transition between rhythmic
activity to arrhythmic activity can be detected. The present invention
further includes an improved implantable cardiac event detection system
which eliminates double sensing of T-waves.
BACKGROUND OF THE INVENTION
The major pumping chambers in the human heart are the left and right
ventricles. The simultaneous physical contraction of the myocardial tissue
in these chambers expels blood into the aorta and the pulmonary artery.
Blood enters the ventricles from smaller antechambers called the left and
right atria which contract about 100 milliseconds (ms) before the
ventricles. This interval is known as the atrioventricular (AV) delay. The
physical contractions of the muscle tissue result from the depolarization
of such tissue, which depolarization is induced by a wave of spontaneous
electrical excitation which begins in the right atrium, spreads to the
left atrium and then enters the AV node which delays its passage to the
ventricles via the so-called bundle of His. The frequency of the waves of
excitation is normally regulated metabolically by the sinus node. The
atrial rate is thus referred to as the sinus rate or sinus rhythm of the
heart.
Electrical signals corresponding to the depolarization of the myocardial
muscle tissue appear in the patient's electrocardiogram. A brief low
amplitude signal known as the P-wave accompanies atrial depolarization
normally followed by a much larger amplitude signal, known as the QRS
complex, with a predominant R-wave signifying ventricular depolarization.
Repolarization prior to the next contraction is marked by a broad waveform
in the electrocardiogram known as the T-wave.
A typical implanted cardiac pacer (or pacemaker) operates by supplying
missing stimulation pulses through an electrode on a pacing lead in
contact with the atrial or ventricular muscle tissue. The electrical
stimulus independently initiates depolarization of the myocardial (atrial
or ventricular) tissue resulting in the desired contraction. The P-wave or
R-wave can be sensed through the same lead (i.e., the pacing lead) and
used as a timing signal to synchronize or inhibit stimulation pulses in
relation to spontaneous (natural or intrinsic) cardiac activity. The
sensed P-wave or R-wave signals are referred to as an atrial electrogram
or ventricular electrogram, respectively.
Note that the term electrogram lead is used herein to refer to the lead
that transmits the sensed electrogram signal from the heart, and the term
pacing lead is used to refer to the lead that transmits the stimulation
pulse to the heart. As mentioned above, however, these "leads" are
generally combined (i.e., the sensed electrogram signal is transmitted
from the heart by the same lead that transmits the stimulation pulse to
the heart). The separate terms "electrogram lead" and "pacing lead" are
used herein merely to indicate that the electrogram signal and the
stimulation pulse could be transmitted using separate leads.
Every modern-day implantable pacemaker includes a sensing circuit, whether
the activity of one or both chambers of the heart are sensed. A cardiac
event is sensed when an amplified electrogram signal exceeds a threshold
value. If the sensitivity level is too low (i.e., the gain is too low),
then some cardiac events will not be sensed because even peak signals may
not exceed the threshold level. If the sensitivity level is too high, on
the other hand, the high gain of the amplifier may cause noise or T-wave
signals to be sensed, giving rise to erroneous sensing of cardiac events.
Pacemakers provided with communications telemetry (e.g., noninvasive
programming capabilities) advantageously allow the physician to set the
sensitivity level.
There are at least two disadvantages to having the physician set the
sensitivity level. First, adjusting the sensitivity level is one more
thing that the physician must remember to do, and it would be advantageous
to relieve him or her of that task if it is possible to do so. Second, and
more important, the physician generally sees the patient only
occasionally, and weeks or months may go by without the sensitivity level
being changed. Problematically, the sensitivity level that will accurately
detect cardiac events at a given threshold level for a patient does not
stay static; both R-wave amplitude and frequency content can vary
considerably within a given patient. Changes in the sensitivity level are
needed to accommodate for physical and mental stress. In addition, the
sensitivity level needs to change as myocardial tissue (heart muscle
tissue) undergoes scarring or other physical responses to the implanted
electrogram lead(s). Other changes in the myocardium-electrogram lead
interface, e.g., shifting of the position of the electrogram lead, may
also cause changes in the proper sensitivity level.
Unfortunately, these changes occur over a period of days and, in some
cases, even hours or minutes. Because the physician generally sees the
patient only every few weeks or months, the pacemaker sensing circuits can
erroneously detect, or not detect, cardiac events over large periods of
time. This erroneous detection/non-detection can cause under-pacing or
over-pacing of the heart. Unfortunately for the patient, such changes may
potentially leave him or her in a worse condition than he or she was in
before the pacemaker was implanted. At best, the pacemaker is not able to
operate efficiently--either by unnecessarily pacing and thereby draining
the battery and risking pacemaker-induced tachycardias; or by not pacing
as often as is needed by the patient. Thus, what is needed is a way to
adjust the sensitivity level of a cardiac event detector in response to
changing conditions in an electrogram signal over a short period of time.
One way of adjusting the sensitivity level of a cardiac event detector is
discussed in U.S. Pat. No. 4,708,144 issued to Hamilton et al. The
Hamilton et al. patent shows the use of an attenuator to attenuate an
amplified signal before such signal is digitized and rectified. After the
signal is digitized and rectified, the signal is connected to a digital
comparator. The digital comparator compares each digitized and rectified
sample to a threshold value. If the digitized and rectified sample exceeds
the threshold value, a cardiac event is detected. In response to the
detected cardiac event, a pacemaker control circuit takes appropriate
action.
Each digitized and rectified sample is also presented to a peak detector
which stores the maximum, or largest, digitized and rectified sample. The
stored maximum sample is coupled to the pacemaker control circuit. After
each cardiac event is sensed, the pacemaker control circuit averages the
stored maximum with any of the previously occurring maximums yielding an
average peak value. This average peak value is used to determine whether
or not the attenuator should be adjusted to increase or decrease the
attenuation provided by the attenuator. Specifically, if the average peak
value increases, the attenuation is increased; and if the average peak
value decreases, the attenuation is decreased, thereby adjusting the
amplitude of the signal before it is digitized and rectified.
Disadvantageously, even though the Hamilton et al. circuit provides one
technique for dynamically adjusting the sensitivity level of a cardiac
event detector, by attenuating the cardiac signals after the amplification
stage, it suffers from potentially clipping the input signal before even
reaching the attenuation stage or the peak detecting stage. Furthermore,
it completely lacks the ability to eliminate the sensing of high amplitude
T-waves, which could cause the peak detector to erroneously detect the
peak T-wave. Furthermore, if the amplitude of the T-wave is too high, or
if the gain of the amplifier is so high that the R-wave is clipped and the
T-wave is, by comparison, similar in amplitude to the clipped R-wave, it
can result in "double sensing". Double sensing, when it occurs, then
falsely indicates that a tachycardia is present. Thus, Hamilton et al. is
not suitable for important cardiac monitoring functions beyond merely
sensing a cardiac event. What is needed is a system which adjusts the gain
at the pre-amplification stage for an optimum signal (i.e., without
clipping the input signal) and then reliably eliminate sensing of the
T-wave.
In addition to the detection of cardiac events, it is desirable, in the
treatment of certain heart ailments, or for the detection of such
ailments, to continuously monitor the patient over a certain period of
time in order to determine the effectiveness of the treatment being
administered by a pacemaker, under different conditions of stress or
varying conditions of the heart. If the sensing circuit detects that the
pacemaker is administering a less than ideal, or optimum, treatment, the
treatment can be adjusted (e.g., by increasing or decreasing the rate at
which pacing pulses are delivered, by decreasing the threshold level of
the threshold detector, or by increasing the amplitude or duration of the
pacing pulse).
Unfortunately, events that would indicate that the pacemaker may be
providing less than the optimum treatment may occur only infrequently.
Thus, a physician may not detect such abnormal events during a weekly,
biweekly or monthly examination, which may last only a few minutes and may
not be able to adjust the pacemaker accordingly. In an effort to solve
this problem, data acquisition systems have been developed that record
electrogram signals over a predetermined period of time, e.g., on the
order of days. The electrogram signals may then be analyzed by a physician
or, in more advanced system, by a microcontroller in the pacemaker in
accordance with a control program that is designed to react to various
conditions that are manifested by the electrogram signals. Such data
acquisition systems advantageously allow detailed analysis of the
electrogram signal over long periods of time thereby facilitating the
detection and accommodation of infrequent heart abnormalities or the early
detection of slowly developing heart ailments. Such long-term monitoring,
particularly where implemented in advanced programmed systems, makes
possible the purposeful and possibly automatic treatment of heart
abnormalities long before the actual failure of a pacemaker to properly
service the heart. In addition, with such automated systems, therapies,
such as antitachycardia pacing and defibrillation, can be performed on the
heart by pacing systems or dedicated defibrillators that would otherwise
not be able to be performed as quickly or automatically.
Unfortunately, implanted data acquisition systems have heretofore only been
operable over a limited sample of the electrogram signal. This is because
such systems store the electrogram signal in a memory. The memory is of a
limited size, and when the memory is full, either part of the previously
recorded electrogram signal must be discarded to make room for new
electrogram signal to be recorded, or the data acquisition system must
stop recording. In an effort to solve this problem, various high capacity
means of storing electrogram signals have been developed such as magnetic
tape recording systems. For example, U.S. Pat. No. 4,250,888 issued to
Grosskopf, suggests that when the memory is full, a warning message be
given that alerts the patient to the need to contact the physician or to
activate a tape recording system at home.
Disadvantageously, the Grosskopf approach may require that the patient
report to a potentially inconvenient location, (i.e., the physician's
office or the patient's home where the tape recording system is located).
Such inconvenience may encourage the patient to ignore the warning
message. In addition, the warning message can be intrusive and
embarrassing. Furthermore, such warning systems are not used with
implantable pacemakers for at least two reasons. First, implantable
pacemakers are implanted within the body and, as such, any warning means
are neither visible nor readily heard. Second, implantable pacemakers must
be compact and use little power. Generally, the warning message is
generated by a speaker or light source and thus draws a significant
current. It is thus apparent that what is needed is an implantable cardiac
event detection system that is not limited to operating on a small sample
of the cardiac signal over a limited period of time, and that does not
require the use of inconvenient and impractical storage devices such as
tape recording systems.
Some systems have been developed that store only anomalous portions of the
electrogram signal. See, e.g., Grosskopf. However, even these systems have
a limited capacity and when a sufficient number of anomalous portions are
stored, some data loss occurs. This data loss occurs when the memory is
full and either the new signal must be discarded or the previously stored
signal must be discarded. Problematically, the portion of the electrogram
signal that is discarded may be the portion of signal that is needed for
an accurate evaluation of the patient's heart condition. Thus, what is
needed is an implantable cardiac event detection system that is not
limited by the use of a finite capacity memory for storing the electrogram
signal, but that provides information sufficient for programmed evaluation
in a microcontroller and, if needed, automatic adjustment of a pacemaker
or activation of a defibrillator in response to such evaluation.
Another problem faced by the designers of automated cardiac pacing and/or
defibrillation systems is the need for analysis of the electrogram signal.
One approach to accurately analyzing the electrogram signal requires that
the stored electrogram signal be subjected to complex digital filtering
algorithms and statistical analysis. See e.g., U.S. Pat. No. 4,422,459
issued to Simson. In order to generate the digitally filtered and
statistically analyzed signals in Simson, a large computer system is
employed. Such computer system is immobile and inconveniently located at,
e.g., the physician's office, thus making implantation impossible.
Disadvantageously, such algorithms and analysis require that many hundreds
of mathematical operations be performed before an accurate conclusion as
to whether the cardiac pacer and/or defibrillator are performing optimally
can be obtained and, thus, before needed adjustment of the therapies
provided by the cardiac pacer and/or defibrillator can be made.
Problematically, this requires not only the use of a memory to store the
incoming electrogram signal while the mathematical operations are being
completed, the disadvantages of which are discussed above, but requires
that many complicated computational steps be traversed by the
microcontroller. Such complicated computational steps are highly
power-consuming--which would require more frequent replacement of the
battery that powers the implantable cardiac pacer and/or
defibrillator--and thus, disadvantageous in implantable cardiac pacing
applications.
Another approach to accurately analyzing the electrogram signal has been to
allow the physician to analyze the electrogram signal stored in a memory
using conventional electrogram analysis techniques. Disadvantageously, in
order to obtain the stored electrogram signal, the physician must download
the stored electrogram signal via a telemetry circuit in the cardiac pacer
system and/or defibrillator system. Hence, because a memory is used, the
problems discussed above are also present in this approach. A further
disadvantage of this approach is that no automated adjustment of the
therapies provided by the cardiac pacer and/or defibrillator can be made
because such adjustment must wait until the patient has traveled to the
physician's office and until the physician has completed his or her
analysis. Thus, what is needed is an implantable cardiac data acquisition
and analysis system that does not require the use of complicated and
highly power-consuming mathematical computations.
SUMMARY OF THE INVENTION
The present invention advantageously addresses the above and other needs by
providing an improved implantable cardiac event detection system and
method usable with implantable cardiac pacemakers, cardioverters,
defibrillators, or the like.
One aspect of the present invention provides an implantable cardiac event
detection system which can eliminate the sensing of T-waves, reliably and
dynamically set the detection threshold, and optimize the dynamic range of
the incoming cardiac signal (through automatic gain control of a
pre-amplifier stage) for the purpose of automatically adapting to changing
cardiac signal morphology.
The system is coupled to the heart via an electrogram lead as is known in
the art of cardiac pacing. The system is also coupled to a therapy circuit
that provides pacing and/or defibrillation therapies to the heart. An
electrogram signal is sensed through the electrogram lead and is
transmitted to signal conditioning circuitry. The signal conditioning
circuitry includes a pre-amplifier having a plurality of programmable
gains, a narrow band filter, and digitizing circuitry. The resultant
digitized electrogram signal is then coupled to a threshold ,detector and
to a morphology detector, which are, in turn, coupled to a
microcontroller. The microcontroller controls the threshold value used by
the threshold detector, as well as controlling the gain of the
pre-amplifier, as a function of the various morphology parameters (e.g.,
the previous average peak R-wave value) sensed by the morphology detector.
The threshold detector detects a cardiac event (e.g., an R-wave) within the
electrogram signal whenever the electrogram signal exceeds a prescribed
initial threshold value. In the preferred embodiment, the threshold
detector is a digital threshold detector capable of digitally adjusting
the threshold value in a predetermined stepwise fashion. The threshold
detector eliminates the detection of high amplitude T-waves by increasing
the initial threshold value to a temporary threshold value for a
prescribed, or programmable, period of time following the detection of an
R-wave. The threshold value is then gradually ramped down, in a stepwise
fashion, to its initial value within a second prescribed period of time.
Advantageously, the second prescribed time period may be automatically
adjusted as a function of heart rate (e.g., a fast heart rate would
require a fast ramp down to the initial value). In this manner, the
detection of T-waves are eliminated. Thus, "double sensing" of the T-wave
and false indications that a tachycardia is present cannot occur.
In the preferred embodiment, the initial threshold value is automatically
determined by the microcontroller to be a percentage of the average peak
(or maximum) R-wave signals over at least a period of a few minutes (e.g.,
at 25% of the previous peak values). Preferably, the temporary threshold
value is also automatically set to a percentage of the average peak (or
maximum) R-wave signals over at least a period of a few minutes (e.g., at
100% of the previous average peak values). Alternately, the initial and
the temporary values could be a programmable value.
In the preferred embodiment, the gain of the pre-amplifier is automatically
adjusted for the optimum dynamic range so that the incoming cardiac signal
is not clipped. The control of the pre-amplifier is determined by the
microcontroller and the morphology detector, as described in more detail
below.
Another aspect of the present invention provides an implantable cardiac
arrhythmia detection system for detecting the transition between rhythmic
and arrhythmic cardiac activity in a heart using the morphology detector.
One feature of the present invention is the detection of a shift in the
average baseline of the rectified cardiac signal. For example, during
normal sinus rhythm, the average baseline of the cardiac electrogram is
approximately zero. When an arrhythmia occurs (such as, a tachycardia or
fibrillation), the average baseline of the cardiac electrogram increases
in magnitude. It is this detection of the shifting of the average baseline
which is used, in the present invention, to detect a change in the
patient's cardiac rhythm.
In one embodiment the "average baseline" may be thought of as an RMS value
of the electrogram signal, or the average of the rectified signal, in
that, only the positive values are considered. In the preferred
embodiment, the "average baseline" is the sum of the unsigned magnitudes
of a plurality of digitized samples during a prescribed interval.
Another feature of the present invention is the detection of a change in
the morphology of the R-wave as a way to indicate a change between
rhythmic and arrhythmic cardiac activity. For example, if the amplitude of
the R-wave increase or decreases by a prescribed amount, or changes
polarity, chances are that a new ectopic foci is generating R-waves from a
new location, thereby indicating a change in the patient's cardiac rhythm.
Furthermore, when using morphology changes in combination with the
shifting of the average baseline, an even higher confidence level is
achieved that the patient's cardiac rhythm has become arrhythmic.
Thus, in the present invention, the morphology detector detects various
parameters associated with the morphology of the electrogram signal, and
couples such parameters to the microcontroller. (Note, as used herein the
term "morphology" relates to the shape of the electrogram signal when
viewed as a signal waveform as a function of time.) The microcontroller
then indicates the presence of an arrhythmic cardiac condition in response
to a prescribed change in the morphology parameters. Such indication may
then be used by the desired therapy circuit, e.g., a pacemaker,
cardioverter or defibrillator, in order to deliver an appropriate therapy
to the heart.
The morphology detector, according to the present invention, includes one
or more of the following: a minimum detector, a maximum detector, a peak
detector, a baseline averager, a baseline sampler, an accumulator, and/or
an interval counter; the output signals of each being coupled to the
microcontroller for determining the presence of an arrhythmia as described
in more detail below.
The minimum detector, used in the morphology detector, generates a minimum
signal, indicative of the magnitude of the most negative value (i.e.,
below a baseline voltage) in the electrogram signal and records the
magnitude of such value as a minimum value.
The maximum detector generates a maximum signal, indicative of the
magnitude of the most positive value (i.e., above the baseline voltage) in
the electrogram signal and records the magnitude of such value as a
maximum value.
The peak detector, used in the morphology detector, generates a peak
signal, indicative of the largest value in the electrogram signal,
regardless of whether the largest value is negative or positive.
Alternatively, the peak signal may be generated by the microcontroller, in
which the microcontroller determines the peak signal to be the greater of
the minimum value and the maximum value. (Note that the minimum and
maximum values once determined, are both considered as positive
magnitudes).
The baseline averager generates an "average baseline signal", indicative of
the average magnitude of the electrogram signal over a predetermined time
period. The predetermined time period may be the time during which the
baseline is expected to be quiescent (e.g., as determined by the counter
or by the microcontroller). Alternatively, the predetermined time period
may be the entire cardiac cycle. The rationale for the latter is that
during normal sinus rhythm, the average baseline signal over the entire
cardiac cycle approximates the true (quiescent) baseline.
As an alternative to the baseline averager, in accordance with one
embodiment of the invention, a baseline sampler may be used by the
morphology detector. The baseline sampler generates an accumulated
baseline signal, indicative of the accumulated magnitude (i.e., the sum of
all baseline values) of the electrogram signal during the predetermined
period of time. The microcontroller then counts the number of discrete
sample values in the processed electrogram signal that are accumulated by
the baseline sampler during the predetermined time period. In this way, a
count value is generated. The total baseline value, from the baseline
sampler, is then divided by such count value in order to generate the
average baseline value.
Typically, the baseline sampler operates digitally. That is, the
electrogram signal is sampled and digitized. The microcontroller then
simply reads the magnitude of the digitized samples occurring during the
predetermined time period and divides it by the total number of samples,
as determined by the interval counter.
In accordance with another embodiment of the invention, the electrogram
signal is not digitized (i.e. the processed electrogram signal is analog).
The baseline sampler may analogically and rapidly sample the magnitudes of
the processed electrogram signal over the predetermined time period. The
magnitude of each of the rapid samplings (discrete values) occurring
during the predetermined time period are added together. In this way, the
accumulated baseline signal can be analogically generated. The accumulated
baseline signal is coupled to the microcontroller, which then divides it
by the time interval, as determined by, e.g., an interval counter.
The accumulator, when used within the morphology detector, generates an
accumulated magnitude signal indicative of the accumulated magnitude of
the electrogram signal during a cardiac cycle. The cardiac cycle begins
when an R-wave is detected by the threshold detector, as described above,
and ends when a subsequent R-wave is detected. The accumulator may operate
digitally or analogically in the same or similar manner as the baseline
sampler described above. The accumulated magnitude signal may then be
divided by the total number of samples over the entire cardiac cycle to
produce an average baseline signal over the entire cardiac cycle.
In one embodiment, the microcontroller provides a count signal
corresponding to the number of samples in the predetermined time period
(or cardiac cycle). Alternatively, the count signal is determined using a
counter (as opposed to counting with the microcontroller). Such count
signal is then coupled to the microcontroller, where the accumulated
baseline value from the baseline sampler is divided by such count value to
generate the average baseline value.
An arrhythmia flag signal is generated by the microcontroller when the
quotient of the peak (or max) R-wave value divided by the average baseline
value exceeds an arrhythmia threshold value (stored, e.g., in the
microcontroller). The arrhythmia flag signal may then be used to engage a
therapy circuit, e.g., a circuit that issues stimulation and/or
defibrillation pulses in a prescribed pattern.
In addition, the microcontroller may generate one or more other morphology
change signals, such as signals indicating: a change in polarity, the
amplitude of the R-waves or the gain of the pre-amplifier (which is
related to the envelope of the incoming cardiac signal). A control program
executed by the microcontroller controls the generation of such output
signals. The control program is executed in response to the detection of
an R-wave, and/or a time-out signal (i.e., the time-out signal is
generated in the absence of cardiac signals in order to limit the number
of samples that will be acquired by the event detector).
In the discussion below, it is generally assumed that the maximum value of
the cardiac signal is greater than the minimum value; however, it is to be
understood that in some instances the minimum value may be larger than the
maximum value. Thus, while the following discussion is directed to setting
various morphology change values based on changes in the maximum value, in
the event that the minimum value is greater than the maximum value, the
morphology change values would instead be set in response to changes in
the minimum value.
During normal sinus cardiac rhythm, the relationship between the magnitude
of the maximum value and the magnitude of the minimum value does not
typically change (i.e., the maximum value generally remains larger than
the minimum value). In the event that the magnitude of the minimum value
suddenly becomes larger than the magnitude of the maximum value, a radical
change in the morphology of the electrogram signal is indicated. The
microcontroller sets the morphology change value to "POLARITY" in response
to such radical change. Furthermore, because it is likely that the signals
thereafter generated by the morphology detector and/or the microcontroller
are no longer indicative of the morphology of the previous electrogram
signal, the microcontroller resets the morphology detector and/or the
output signals (e.g., the minimum value, the maximum value, the peak
value, and the baseline value, etc.) generated by the microcontroller in
response to a change in the morphology change value, e.g., when the
polarity changes.
In addition to setting the morphology change value to "POLARITY," the
microcontroller may set the morphology change value to "INCREASE,"
"DECREASE," and/or "NONE." For example, in the event that the present
maximum value (or minimum value) is much greater than the average of the
preceding maximum values (or minimum values), the microcontroller sets the
morphology change value to "INCREASE." When the morphology change value is
set to "INCREASE," there is a high probability that a cardiac arrhythmia
has begun. The therapy circuit may then respond to such detected
arrhythmia as is known in the art of implantable pacemakers.
In the event that the present maximum value (or minimum value) is much
lower than the average of the preceding maximum values (or minimum
values), but is still greater than the minimum value (or maximum value),
the microcontroller sets the morphology change value to "DECREASE." When
the morphology change value is set to "DECREASE," there is a likewise high
probability that a cardiac arrhythmia has begun, and the therapy circuit
may thus respond accordingly.
In the event that the present maximum value (or minimum value) is not much
greater or much less than the average of the preceding maximum values (or
minimum values), and the present maximum value (or minimum value) remains
greater than the present minimum value (or maximum value), the morphology
change value is set to "NONE." This indicates to the therapy circuit that
the heart is experiencing normal sinus rhythm. In this way, changes in the
morphology of the electrogram signal, particularly the onset of
arrhythmias, can be detected without the need for complicated and
power-consuming computation by the microcontroller.
Note that in the event the peak detector is used instead of the minimum and
maximum detectors, the morphology change value is set in response to
changes in the peak value in a manner similar to that described above.
However, it is important to note that the morphology change value is
generally not set to "POLARITY" when the peak detector is used instead of
the minimum and maximum detectors.
The gain change value output signal is generated by the microcontroller in
response to variations in an average of the previous maximum values (or
minimum values). That is, in the event that the average of the previous
maximum values (or minimum values) increases substantially (e.g., as such
average is updated over a period of time), the microcontroller sets the
gain change value to "DECREASE" and decreases the gain of the
pre-amplifier (or other sense amplifier) used as a part of the signal
conditioning circuitry. Similarly, if the average of the previous maximum
values (or minimum values) decreases substantially, the microcontroller
sets the gain change value to "INCREASE" and increases the gain of the
pre-amplifier. If the average of the previous maximum values (or minimum
values) does not change substantially, the gain change value is set to
"NONE" and the gain of the pre-amplifier is not changed.
The gain change value and the control of the gain of the pre-amplifier
serve two functions. First, a change in the gain change value is
indicative of a change in the envelope of the cardiac signal, and thus, a
change in the morphology. That is, a change in gain indicates that R-waves
are now being generated from a new ectopic foci. Therefore, the gain
change value may be used to indicate an arrhythmia is present. Secondly,
control of the gain of the amplifier helps to minimize the sensing of
T-waves. That is, by maximizing the R-wave signal (by preventing
clipping), the effect is to minimize the T-wave amplitude, and further
enables a larger range of thresholds which can detect the R-wave without
detecting the T-wave. In this way, the gain of the pre-amplifier may be
adjusted in response to varying amplitudes of the electrogram signal.
In accordance with one aspect of the invention, the microcontroller sets
the sensitivity of the pre-amplifier based on the gain change value. In
this way, the sensitivity of the pre-amplifiers can be set to an optimum
dynamic range (i.e., without clipping the cardiac signal) so that cardiac
events are accurately sensed, while noise and the like are not mistaken
for cardiac events.
The output signals of the microcontroller representing the average baseline
value, the peak value, the arrhythmia flag value (i.e., the detected shift
in the average baseline signal or the peak value divided by the average
baseline), the morphology change value, and the gain change value are used
to generate the average baseline signal, an average peak signal, an
arrhythmia flag signal, a morphology change signal, and a gain change
signal, respectively. Such output signals are coupled to the appropriate
therapy circuit, and are used by the implanted therapy circuit to adjust
| | |
