CA1299252C - Implantable cardiac stimulator for detection and treatment of ventricular arrhythmias - Google Patents

Abstract

ABSTRACT

A cardiac stimulator for detecting and treating ventricular tachyarrhythmias of a patient's heart includes means for selectively dividing the heart rate continuum into regions including at least two classes of tachycardia, con-tiguous to each other and of progressively higher heart rate ranges, the lowest and highest of the tachycardia classes being bounded respectively by a sinus rate region and a fib-rillation region of the continuum. Means is provided for sel-ectively adjusting the boundaries between the tachycardia classes and between the lowest and highest of those classes and the respective sinus rate and fibrillation regions, to correspondingly adjust the rate ranges of the classes in a desired manner. Means is also provided for selectively detecting cardiac events anywhere within the continuum and for distinguishing between normal and abnormal tachycardias among the detected events. The stimulator additionally in-cludes means for selectively treating a detected abnormal tachycardia with any of a multiplicity of therapy regimens of differing degrees of aggressiveness, toward terminating the detected tachycardia.

Description

IMPLANTABLE CARDIAC STIMULATOR_FOR_DETEGTION
AND_TREATMENT_OF_VENTRICULAR_ARRHYTHMIAS

BAGKGROUND OF_THE _NVENTION
The present invention pertains to an implantable cardiac stimulator with a hierarchical approach to the treatment of ventricular arrhythmias, utilizing combinations of pacing, cardioverting, and defibrillating therapies.
Many factors affect the rate at which the human heart beats, to vary the rate from what i9 termed the normal sinus rate range.
Adult heart rates below 60 beats per minute (bpm) are referred to as sinus bradycardia, while the rates generally ranging from 101 to 160 bpm are called sinus tachycardia. In healthy persons, such tachycardia arises, for example, from physical or emotional stress (exercise or excitement~, consumption of alcoholic or caffeinated beverages, cigarette smoking, or taking of certain drugs. Rates exceeding 200 bpm have been observed in younger persons during strenuous exercise.
Variation from normal sinus rate range is generally characterized as cardiac arrhythmia, and arrhythmia rates exceeding the upper end of the sinus rate range are termed tachyarrhythmias. Healthy persons usually experience a gradual return to the sinus rate after removal of the factor(s) giving rise to sinus tachycardia. On the other hand, abnormal arrhythmias require special treatment, and in -some instances require immediate emergency treatment toward preventing sudden death of the afflicted individual. Alteration or impairment of any of the electrophysiologic pr~perties of the heart may result in cardiac arrhythmias. For example, A-V junctional tachycardia is acceleration of ectopic rhythmicity despite generation of cardiac impulses at sinus rate by the sinus node of the heart.
Fibrillation is a tachyarrhythmia characterized by the commencement of completely uncoordinated random contractions by sections of conductive cardiac tissue of the affected chamber, quickly resulting in a complete loss of synchronous contraction of the overall mass of tissue and a consequent loss of the blood-pumping capability of that chamber.
Artificial cardiac pacemakers have been prescribed by cardiologists for many years to treat bradycardia. The pacemaker pulse generator is implanted in a pouch beneath the skin in the patient's chest and delivers electrical impulses to el~ctrodes positioned at the patient's heart via one or more catheter leads, to stimulate the heart to beat at a desired rate in the normal range. Over the past several years cardiac pacing has found increasing usage in the management of tachyarrhythmias. Anti-tachyarrhythmia pacemakers take advantage of the heartls own inhibitory mechanism which acts on the secondary natural pacemakers heart to prevent spontaneous rhythmicity, this mechanism being referred to a "postdrive inhibition" or "overdrive inhibition". In essence, the heart may be stimulated at a faster than normal pacing rate to suppress ectopic activity in the form of premature atrial or ventricular contractions (extrasystoles) that might otherwise initiate supraventricular or ventricular -~299252 tachycardia, flutter (typically, a tachyarrhy~hmia exceeding 200bpm), or fibrlllation; or to ~erminate an existing tachyarrhythmia. ~he pulses delivered to the heart for pacing therapy need only be of sufficient magnitude to stimulate the excitable myocardial tissue in the immediate vicinity of the pacing electrode. In contrast, another technique for terminating tachycardias, termed cardioversion, utilizes apparatus to shock the heart with one or more current or voltage pulses of generally considerably higher eneryy content than is delivered in pacing pulses. Whether pacing or cardioverting therapy is employed in an effort to terminate a tachycardia, a considerable risk is present that the treatment itself may precipitate fibrillation.
Defibrillation t"D~"), the method employed to terminate fibrillation, involves applying one or more high energy "counter-shoc~s" to the heart in an effort to overwhelm the chaotic con-tractions of individual tissue sections, allow re-establishment of an organized spreading of action potential from cell to cell of the myocardium, and thus restore the synchronized contraction of the mass of tissue. The term "cardioversion" is sometimes used broadly to include DF, but as used herein, a distinction is maintained between the two terms.
In general, atrial fibrillation ("AF") is hemodynamically tolerated and not life-threateniny because the atria provide only a relatively small portion (typically on the or~er of 15 to 20 percent) of the total cardiac output, i.e., the volume of blood pumped by the heart per unit time. Indeed, a technique lZ9925Z

frequently used in the past for terminating atrial flutter involves stimulating the a~rium with artificial pacing pulses delivered at a rate higher than the flutter rate to convert the flutter to AF. Within a relatively brief interval after such pacing, the heart usually reverts to normal sinus rhythm on its own. During this time, the tissue remains healthy because it is continuing to receive a fresh supply of oxygenated blood as a result of the continued pumping action of the ventricles.
Atrial tachycardia ("AT") is also hemodynamically tolerated because of the natural protective property of the A-V junctional tissue (referred to as "functional A-V block") attributable to its longer refractory period and slower conductivity than atrial tissue. This property renders the A-V junctional tissue tissue unable to fully respond to the more rapid atrial contractions.
As a result, the ventricle may miss every other or perhaps two of every three contractions in the high rate atrial sequence, resulting in 2:1 or 3:1 A-V conduction, and thus maintain relatively strong cardiac output and near-normal rhythm.
In cases where the patient is symptomatic or at high risk in events of AT or AF -- for example, instances where the patient suffers from ventricular heart disease and consequent reduction of ventricular pumping capability, with a correspondingly greater contribution by the atria to cardiac output -- special treatment is necessitated. Types of treatment commonly prescribed include medication, drugs, pacing therapy, cardiac shock therapy, and in some cases, surgically creating an A-V block and implan~ing a 12~9252 ventricular pacemaker.
During an episode of ventricular tachycardia ("VT"), cardiac output is diminished because the ventricles, which are the main pumping chambers of the heart are only partially filled between the rapid contractions. There is a high risk that the VT may accelerate into ventricular fibrillation ("VF"), spontaneously or in response to treatment of the VT. In that event, there is an instantaneous cessation of cardiac output as a result of the ineffectual quivering of the ventricles. Unless cardiac output is restored almost immediately, tissue begins to die for lack of oxygenated blood, and death will occur within minutes.
An object of the invention is to provide an improved medical device for treating ventricular tachyarrhythmias, including ventricular tachycardia, flutter, and fibrillation, with improved techniques for detecting the arrhythmia and distinguishing it from normal high rates, and with a hierarchical approach to the aggressiveness and delivery of therapies.
The pulse energy requirements for cardioversion and defibrillation overlap to an extent, ranging from as low as about 0.05 joule to approximately 10 joules for cardioversion and from about 5 ~oules to approximately 40 joules for DF. The energy level required differs from patient to patient, and depends on type of pulse waveform and electrode configuration used, as well as various other known factors.
Traditional practical defibrillators were rather bulky electrical apparatus for applying a high-energy pulse through the ~2~252 heart via paddles placed at predetermined locations on the patient's thorax. More recer~tly, implantable cardioverters and defibrillators have been proposed for use in detecting and treating VT and/or VF. In 1970, M. Mirowski et al. and J. C.
Schuder et al. separately reported in the scientific literature their independent proposals of a "standby automatic defibrilla-tor" and a "completely implanted defibrillator", respectively, and experimental results in dog tests. Since that time, a vast number of improvements in implantable cardioverters and defib-rillators has been reported in the scientific literature and patent publications. Following is a representative sampling.
U.S. Patent No. 3,805,~95 describes a defibrillator circuit with implanted electrodes for delivering defibrillating pulses to the heart only if separate signals respectively indicative of electrical and mechanical ac~ivity are both absent for a predetermined period of time, and in which the first pulse has lower energy content than succeeding pulses. U.S. Patent No.
4,114,628 discloses a demand pacemaker with an operating modé in which a difibrillating pulse is automatically applied to the patient ~5 heart in the absence of cardiac activity for a predetermined period of time. U.S. Patent No. Re.27,652 suggests an automatic implantable defibrillator in which a preset delay is imposed between successive shocks, and in which further shocks are inhibited following successful defibrillation. U.S. Patent No. 4,181,133 describes a programmable implantable pacemaker which provides the dual functions of demand pacing and standby ~299252 cardioversion. U.S. Patent No. 4,300,567 discloses an implantable automat1c defibrilla1:or a~ar)ted to ~eliver a high energy defibrillating pulse in one mode and lower energy cardioverting pulses in another mode.
Generally speaking, the implantable defibrillators of the prior art detect ECG changes and/or absence of a "mechanical"
function such as rhythmic contractions, pulsatile arterial pressure, or respiration, and, in response, deliver a fixed therapy typically consistîng of one or more shocking pulses of preset waveform and energy content. If any other cardiac therapy is available from the device, such as cardioversion for treatment of tachycardia, it too is delivered according to a fixed plan in response to conventional detection of the specific arrhythmia.
These proposed devices offer little or no flexibility of therapy regimen or capability to detect subtle chariges in the arrhythmia to be treated and to respond with appropriate therapy.
A more specific ob~ect of the present invention is to provide an improved implantable medical device adapted to detect ventricular tachycardias and other arrhythmias throughout the heart rate continuum, an~ automatically responsive to such detection to selectively deliver one or more of a plurality of predetermined therapies, including bradycardia and antitachy-cardia pacing-type therapies and cardioverting and DF shock-type therapies, in different regimens Or dlssimilar aggressiveness of treatment according to the degree of hemodynamic tolerance or intolerance of the detected arrhythmia.

~z~z~z Another object of the invention is to provide a medical device suitable for delivering antitachycardia pacing therapy for managing ventricular tachycardias, and having additional or backup capabilities for terminating VF in the event of acceleration, and thereby to lessen the risk associated with using pacing therapies to treat VT.

SUMMARY OF THE_INVENTION
The present invention integrates the functions of brady-cardia and anti-tachycardia pacing-type therapies, and cardio-version and defibrillation shock-type therapies, to provide a coordinated approach to the management and treatment of ventricular arrhythmias, including VT and VF. A significant aspect of this approach is to provide flexible sequencing among the therapies, with appropriate regard to hemodynamic tolerance (or intolerance) of the patient to the detected arrhythmia, and sophisticated detection of arrhythmias together with means for distinguishing those episodes for which treatment is required (such as reentrant tachycardias) from those which are not associated with cardiac or other disease (such as exercise-generated sinus tachycardias). The coordinated approach of the invention further takes into account and deals effectively with the risk of acceleration of a tachycardia, as well as with considerations of longevity of the power source for an imp~antable device, and of patient acceptance of the device.
According to the invention, a multiplicity of hierarchical detection algorithms and hierarchical therapeutic modalities are selectively available to the physician and applicable to detect and treat classes of ventricular t:ac~y~ardia according to their respective positions in the heart rate continuum, and thus according to hemodynamic tolerance or intolerance of the patient to the tachycardia, with backup capabilities of defibrillation and bradycardia pacing for cardiac arrhythmias at the respective higher and lower regions of the rate continuum.
A feature of the invention is the application of the principle that aggressiveness of the therapy should be increased with elapsed time and with increasing abnormal heart rate. The invention provides the physician with complete control over the aggressiveness of the therapy for any particular patient and tachyarrhythmia, utilizing a hierarchical approach to treatment.
The methodology employed in developing the hierarchy is such that physician control is imparted over a wide variety of possible therapy regimens ranging from the basic to the highly complex, with relatively simple programming of the device.
In one embodiment, the cardiac stimulator permits selective partitioning of the heart rate continuum into a plurality of contiguous tachycardia classes of progressively higher rate ranges, the lowest and highest of these classes being bounded respectively by regions of the continuum denoting sinus rate and fibrillation. Each of the rate ranges and the latter regions may be arbitrarily designated by the physician, as may be necessary to meet the particular needs of the patient's disorder and the flexibility of the therapy regimens to be prescribed. The ~99252 stimulator includes a hierarchical detection system for detecting cardiac episodes lndicative of arrhythmia and for distinguishing between normal and abnormal tachycardias among the detected episodes, using criteria of greater or lesser stringency depending on the location of the episode in the rate continuum.
In response to detection of an arrhythmia within or outside any of the designated tachycardia classes, the stimulator will automatically deliver one or more therapies according to the physician's exact prescription (based on various factors including, for example, specific patient data, arrhythmia rate, episode longevity and acceleration or deceleration). In this particular embodiment, the available therapies include bradycardia pacing, anti~tachycardia pacing, cardioverting shocks, and DF shocks, which may be delivered separately or in any combination (according to the physician's prescription) to treat the detected arrhythmia, and more particularly, with an ascending order of aggressiveness of the therapy according to the degree of hemodynamic intolerance of the arrhythmia.
Another feature of the invention is that the hierarchy of algorithms developed for detecting arrhythmias in the various rate ranges may be assigned to make the criteria progressively less stringent for detecting episodes in progressively higher rate ranges, so that the detection criteria are relaxed with ncreasing hemodynamic intolerance of the arrhythmiaO
A further object of the invention is to provide such an arrhythmia detection and treatment device in which redetection to determine the presence or absence of the arrhythmia tincluding any acceleration or deceleration thereof) upon delivery of the prescribed therapy regimen in response to the earlier detection, may employ fewer than the tests used for the original detection.
Still another object of the present invention is to provide such an arrhythmia detection and treatment device in which the delivery of the therapy sequences for treatment of the arrhythmia may be altered while the episode is in progress, according to a preselected control option.
The above and still further obJects, features and advantages of the present invention will become apparent from the following detailed description of a presently preferred embodiment, taken in conjunction with the accompanying drawings, irl which:
FIGS. la and lb are diagrams illustrating the manner in which the heart rate continuum is divided (partitioned) into arrhythmia classes according to the invention;
FIGS. 2a and 2b are diagrams illustrating the assignment tPrescription) of exemplary therapy regimens to the arrhythmia classes;
FIGS. 3a and 3b are diagrams illustrating the designation of detection criteria algorithms to the arrhythmia classes;
FIG. 4 is a diagram indicating the characteristics of the fine structure of the burst pacing therapies;
FIG. 5 is a flow diagram illustrating timing aspects of the cardiac stimulator;
FIG. 6 is a diagrammatic representation of the electrical i~9925Z

circuit of a presently preferred embodiment of the implantable cardiac stimulator according to the invention;
FIG. 7 is a more detailed diagram of a portion of the circuit of FIG. 6; and FIGS. 8 though 18, inclusive, are more detailed circuit diagrams of various portions of the preferred embodiment.
Fig. 15 appears on the same drawing sheet as Fig. 13.

DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT
Nith reference now to FIG. la, according to one significant aspect of the invention, the heart rate continuum or spectrum is divided, or partitioned, into a multiplicity of region~ defining contiguous, successive heart rate ranges consistent respectively with sinus rhythm, designated SINUS
at the lower end of the continuum of interest, progressively higher rate ranges associated with VT, respectively labelled TACH-l, TACH-2, and TACH-3, and into and beyond the commencement of rates associated with VF, designated FIB at the upper end of the continuum of interest.
Preferably, the spectrum is partitioned such that the rate ranges (regions) so defined are representative of respective degrees of hemodynamic tolerance of the patient to cardiac rates in those regions. Thus in the illustrative example of FIG. la, the ascending order of the three VT
regions depicts well tolerated, moderately tolerated, and poorly tolerated classes of tachycardia, respectively. The actual number of such classes may be greater or fewer than three depending on the judgment of the physician regarding the management of arrhythmias and the prescription of therapy regimens for a particular patient. As ~2992S2 indicated in FIG. la, heart rates in the SINUS region are normal (at least that portion of the region considered to be in the normal resting sinus rate range), whereas rates in the FIB region are not tolerated at all.
The overall continuum of interest may be left unbounded and the rate ranges of the tachycardia classes selectively designated by assigning specific rate numbers to the boundaries of those regions with each other and with the 5INUS and FIB regions. For example, the lower and upper boundaries of the TACH-1 region may be set at 150 and 175 bpm, and of the TACH-3 region at 200 and 275 bpm, respectively, coincidentally defining the rate range of the TACH-2 region as well as the upper boundary of the SINUS
region and the lower boundary of the FIB region, as illustrated in FIG. lb.
Each boundary rate separating adjacent regions is selectively ad~ustable by the physician during the programming or reprogramming of the cardiac stimulator (using a programmer unit external to the stimulator), based on the particular patient data, including age, nature of the disorder, and any other factors the physician may deem pertinent. These programmed boundary rates are stored in a computer memory associated with a central microprocessor within the cardiac stimulator, via an implanted antenna and data transmission networ~ of the stimulator, as will be described presently.
The present invention also gives the physician the capability to prescribe any of a plurality of basic therapies for treatment of the arrhythmias, to specify the detailed nature (i.e., the fine structure) of each of those therapie~, to designate the sequence in which the therapies are to be delivered in response to a detected arrhythmia in any of the designated arrhythmia regions, and to select the algorithms for detecting arrhythmias in each region. For example, in the presently preferred embodiment of the stimulator any of four basic therapies may be selectively designated to treat respective detected events in each of the four arrhythmia classes TACH-1, TACH~2, TACH-3, and FIB. The number of basic therapies may be greater or fewer than the number of arrhythmia classes, and there is no particular significance to the common number of them in thi~ embodiment. Also, the number and complexity of the basic therapies, and of other stored and/or programmable data functions described herein, are limited only by memory type and capacity in the cardiac stimulator and associated programming unit.
In essence, the basic therapies may be defined in any desired manner from the group consisting of all of the potential therapies which may be delivered by the device, each of the deliverable therapies being alterable (again, within the practical limitations of the device) in terms of its fine structure. Thus, the device gives the physician the capability to alter the fine structures of the deliverable therapies, and to define the basic therapies. The latter may then be assigned ~eparately or in any combination as a plurality of regimens or sequences appropriate for treatment of arrhythmias in the respective designated rate ranges, tailored to the particular patient.
For example, one definition of the four basic therapies of the present embodiment might be as follows THERAPY-A: non-aggressive pacing bursts;
THERAPY-B aggressive pacing bursts;
THERAPY-C: cardioverting shocks; and THERAPY-~: defibrillating shocks.
However, THERAPIES A and B could equally well both be defined as aggressive pacing bursts, differing in fine structure only, or THERAPIES A, B and C could equally well be defined as cardioverting shocks, again differing only in fine structure, or each of the therapies could be defined in any other manner desired by the physician, limited only by the range of therapies deliverable by the stimulator.
Another important feature of the therapy control scheme of the invention i8 that there is no conceptually pre-defined association between the basic therapies and the designated arrhythmia classes. That is, from a conceptual standpoint any or all of the basic therapies may be prescribed for any of the arrhythmia classes. However, that does not dictate against the phy~ic{an specifying one therapy per arrhythmia class in the particular order listed above. For instance, as indicated in FIG~ 2a, a single extra-stimulus (this and other possible fine structures of the therapies will be described presently) THERAPY-A pacing burst could be prescribed (i.e., programmed) by the 129925;~

physician as suitable treatment to pace the patient out of a very low rate, hemodynamically well tolerated VT detected in segion TACH-1; a more aggressive autodecremental pacing burst of THERAPY-B could be prescribed as the appropriate treatment for a higher rate, moderately tolerated VT of class TACH-2; a 2-joule cardioverting shock of THERAPY-C for a poorly tolerated VT in class TACH-3; and a 30-joule defibrillating shock of THERAPY-D
in response to detection of VF.
While the association between therapies and arrhythmia classes is intended to be flexible, there are practical considerations that place certain limitations on the designation of a therapy vis-a-vis a particular class. For example, if VF is to be treated, it follows that a defibrillating shoc~ therapy should be specified for that class of arrhythmia, and that because time is of the essence in treating VF, an aggressive pacing burst is neither a viable alternative nor a suitable preliminary therapy to the defibrillating shock.
A basic principle of therapy control according to the invention is that the therapies delivered by the device should become more aggressive with increasing rate of the tachyarrhyth-mia and/or with passage of time for a cos~tinuing tachyarrhythmia.
In the present embodiment, for any specified combination of the basic therapies applicable to a particular arrhythmia class, the therapies are delivered in the order (sequence) of increasing aggressiveness. An illustration of such therapy sequencing will be described with reference to FIG. 2b. In this instance, the 1~9252 physician has prescribed the combination of T~ERAPY-A and THERAPY-B for delivery in resp~nse to detection of a TACH-1 event (i.e., an arrhythmia in the TACH-1 class); THERAPY-A, THERAPY-B, and THERAPY-C for a TACH-2 event; THERAPY-B and THERAPY-C for a TACH-3 event; and THERAPY-D for a FIB event (VF).
Assuming that the fine structures of these therapies are the same as those specified in the preceding example of FIG. 2a, these prescribed therapy regimens for the arrhythmia classes are manifested as follows. Upon detecting a class TACH-l VT, the device will initially apply a single extra-stimulus pacing pulse to the heart via a stimulating cathodic electrode positioned for ventricular stimulation. If that therapy fails to pace the patient out of the tachycardia, the continued presence of the VT
will be sensed by a redetection al~orithm (to be described presently) of the dev$ce upon completion of delivery of the therapy. In response to redetection of the VT, the device will promptly deliver an autodecremental burst of pacing pulses as the second attempt to break the VT. Unless certain therapy control options (to be described below) have been selected by the physician, the sequence THERAPY-A, THERAPY-B will be repeated while the VT remains in progress and in class TACH-1, up to the device's capability for repetitions (255 times in this particular instance, in the present embodiment).
If the VT accelerates to the TACH-2 class, regardless of whether this occur~ after delivery of a therapy or while either THERAPY-A or THERAPY-B i6 in progress, the device will proceed directly into the "default" sequence for TACH-2. The default sequence is the therapy sequence programmed by the physician for that arrhythmia class (in this example, THERAPY-A, followed by THERAPY-B, followed by TE~ERAPY-C), and always follows the rule (in the absence of particular ones of the therapy control options, to be described presently, having been selected) that if the arrhythmia accelerates or decelerates (i) the therapy or therapy sequence for the old class ceases (even if a therapy is in progress at the time of the transition), and (ii) the least aggressive therapy prescribed for the new class is commenced immediately. If two or more therapies have been prescribed for the new arrhythmia class, the therapy sequence is always delivered (again, in the absence of certain therapy control options having been selected) in the order from least aggressive to more aggressive. In this example, this scheme results in delivery of the therapy sequence in alphabetical order. It will be observed that in the present example, the default change does not result in an increase in therapy aggressiveness.
Continuing with the example of FIG. 2b, if the VT had accelerated to TACH-3 during the therapy delivery sequence for TACH-1, that new event will be sensed from application of the redetection algorithm, and the cardiac stimulator will commence the delivery of the therapy sequence prescribed for TACH-3. In that instance, the programming dictates the delivery of THERAPY-B, which could be an autodecremental pacing burst to the stimulating electrode as the first response, and, should that 1~:9~;~52 fail to terminate the tachycardia, prompt follow-up with THERAPY-C, which co~ld be a 2-~oule cardioverting shock to appropriately positioned electrodes distinct from the pacing electrode and suitable for handling a pulse of that energy level.
If a VT decelerates to a lower class, say, from TACH-3 to TACH-l in the example of FIG. 2b, the default order will be the delivery sequence THERAPY-A, THERAPY-B. Also, if the VT were to commence at a rate within, say, the TACH-3 region, the therapy control re~ponsive to the detection of that event would be the delivery se~uence specified for TACH-3 -- THERAPY-B followed by THERAPY-C in the present example. References throughout this specification to a therapy delivery sequence in response to detection of an arrhythmia are intended to mean delivery of each therapy in the order of aggressiveness, and repetitions of that sequence, only until the arrhythmia is terminated. That is to say, if a TACH-3 VT i8 terminated by THERAPY-B in the example of FIG. 2b, that result is detected by the stimulator which thereupon promp-tly discontinues the therapy regimen 50 that THERAPY-C will not be delivered.
If the patient has a tendency to accelerate spontaneously from VT to VF, the therapy regimen programmed into the device by the physician for that patient is likely to be more aggressive than the regimen selected for response to detection of an identical tachyarrhythmia in a patient whose cardiac history indicates no such predisposition.
The se~uence in which the selected therapies are delivered ~xg92S2 by the cardiac stimulator may be selectively modified according to a plurality of ~herapy control options which are programmable by the physician. In the present embodiment the following options are made available to be exercised by physician programming using the external-programmer unit:
1. Retry exact pacing therapy.

2. Retry last successful pacing therapy.

3. Ratchet.

4. Restart ~pacing portion) (1-255).

5. Disable pacing therapies on deceleration limit.
The "retry exact pacing therapy" ("REPT") therapy control option follows the rule that a particular pacing therapy previously delivered in response to detection of an arrhythmia within a specified class is to be redelivered exactly as on the preceding occasion as the first attempt to terminate the arrhythmia, upon the very next detected arrhythmia in that class -- but only if that therapy was successful in terminating the arrhythmia on the earlier occasion. Accordingly, if REPT has been selected (programmed) by the physician, the entire last successful pacing therapy sequence delivered for a VT of a given class is stored in memory for the subsequent redelivery.
The "retry last successful pacing therapy" ("RLSPT") option differs from REPT in that RLSPT follows the rule that the precise last pacing therapy in the sequence which was successful in terminating a VT within a particular class, is to be applied as the f~rst therapy upon the very next detected VT within that ~299252 class. For example, if the RLSPT option has been selected and the therapy sequence success~ul for the prece~iny episode was A-B-C, then THERAPY-C will be delivered first on the next detection of an arrhythmia in that class. If either the REPT or the RLSPT
option is selected and the respective pacing therapy called for by that option does not succeed in terminating the VT, the treatment will thereupon revert to the prescribed therapy delivery sequence for the particular arrhythmia class. If sufficient m~mory is available, the RLSPT option may be expanded, for example, to follow the corollary that the last successful pacing therapy will be redelivered first, and if that fails, the next-to-last successful pacing therapy will be retried, and so forth down the list of prescribed pacing therapies which have succeeded in breaking a VT in the past, for any arrhythmia class to which the physician has assigned a pacing therapy.
The ''RATCHETI' therapy control option implements the rule that acceleration of a VT to a higher class dictates that the initial therapy delivered for the new class be maintained at least at the level of aggressiveness of the therapy to which the therapy sequence for the old class had proyressed. That is, if the RATCHET option is selected, the cardiac stimulator will deliver prescribed therapy for the new class only in the direction of greater aggressiveness of therapy, and not bac~ard from the therapy level reached during treatment in the old class, nothwithstanding that the prescribed therapy sequence for the new class may include one or more less aggressive therapies.

~299252 As an illustration of the significance of RATCHET, in the example of FIG. 2b an acceleration of the VT from TACH-1 to TACH-2 at the time THERAPY-B was in progress resulted in a cessation of that therapy and a commencement of THERAPY-A, under the default regimen. This may have been acceptable if little time had elapsed in the delivery of THERAPY-A and through the point of progress of THERAPY-B, up to the moment of acceleration. But if both of those therapies had been physician-programmed as elabor-ate scanning bursts, for instance, each might take a clinically significant amount of time to deliver. In those circumstances it may be desirable to maintain the current level of aggressiveness rather than, as was the case in that example, revert to the less aggressive therapy specified for the new class. If the ratchet option had been selected in that example, upon detection of the acceleration of the VT to TACH-3, treatment would commence with a restart of THERAPY-B from the beginning, rather than a return to THERAPY-A. Further, if the therapy sequence had to be repeated, it would again commence with THERAPY-B.
It will be observed that each of the preceding options tends toward commencing the regimen with a therapy l~kely to terminate the VT either more rapidly than if the therapy control designated by the option had not been instituted, or with less discomfort to the patient. The underlying premise is that time is of the essence with movement of the arrhythmia toward hemodynamic intolerance, but if greater luxury of time is present (as where the arrhythmia is well tolerated and there is no significant rate 129932S;;~

advancement) treatment appropriate to arrest the VT without the discomfort attendant in shock therapy may be desirable.
The "RESTART" option is consistent with that premise, providing therapy control in .instances when the VT is well tolerated. In essence, RESTART is an appropriate therapy control option where the prescribed therapy sequence begins with one or more pacing therapies and concludes with a shock-type therapy.
RESTART calls for a predetermined number of repetitions of the pacing therapy(ies) before proceeding into the shock therapy.
Thus, if the therapy sequence specified for delivery in response to a TACH-2 event is ABC, and ~ESTART has been specified, THERAPY-A and THERAPY-B will be repeated the programmed number of times before THERAPY-C is initiated, up to the point of termination of the tachycardia. Selection of the RESTART option may be made such that the pacing therapies are restarted, or not, depending on the specific arrhythmia class.
Reasons for the physician's selection of the RESTART option may include (1) clinical observation that a VT of the class to be treated is hemodynamically well tolerated by the patient; (2) elapsed time from onset of the episode is therefore not as crucial as in instances of moderate or poor tolerance; (3) there is little or no patient discomfort from the pacing therapies, in contrast to a shock-type therapy; and (4) each application of a shock therapy may (and typically does) consume a considerable amount of energy in comparison to the pacing therapies.
Yet another option which may be selected in the presently 129~25~

preferred embodiment is the disabling of pacing therapies on deceleration limit ("DPTODL"). As noted earlier, treatment o~ a VT carries a significant risk of acceleration to VF. Of course, if that happens the device will promptly detect the VF and apply DF. It has been observed, however, that the application of a defibrillating shock sometimes causes a deceleration to VT rather than a return to sinus rhythm. Theoretically, this may occur repetitively, with the delivery of pacing therapy upon detection of the reemergent VT resulting in acceleration again to VF, and so forth in a pace-shock, pace-shock loop. DPTODL is adapted to break this loop by disabling the pacing therapies after a preset number of decelerations from VF to VT (preferably on the second occurrence), and proceeding directly to shock therapy.
Until the VT is terminated the stimulator will repeatedly deliver treatment according to the specified therapy delivery se~uence for that arrhythmia class, unless and to the extent modified by a selected therapy control option, with continuing redetection during each cardiac cycle according to predetermined criteria tto be described presently). It will be appreciated that the availability of the device's therapy control options provides the physician with a powerful tool to adjust the delivery of the therapies from the prescribed sequence for each class, according to the then-available treatment most likely to succeed under the constraints of time, hemodynamic tolerance, acceleration risk, patient discomfort, and energy capacity existing at the time the arrhythmia is detected.

~2~9252 Although the partitioning of the heart rate spectrum into arrhythmia classes in itself provides a foundation for an arrhythm1a detection technique, it is desirable to selectively develop additional information beyond the rate boundaries of these classes in order to more reliably classify the arrhythmia.
For example, the device might detect a single cardiac interval in the TACH-3 rate range, but tha~ may simply be attributable to an isolated premature ventricular contraction (PVC), which is often observed in individuals without heart disease, and not the start of a reentrant VT. Also, a detection scheme based solely on the rate ranges of the designated arrhythmia classes could experience difficulty in distinguishing between a sinus VT and a reentrant VT, which may have a con~iderable overlap in rates. The present invention avoids such difficulties and provides a highly reliable arrhythmia detection technique.
Because reentrant tachycardias are typically characterized by an abrupt onset (in contradistinction to a gradual ramping up in exercise-induced sinus tachycardias) and a stable high rate (in contradi~tinction to the rate fluctuation of exercise tachycardias), the following four basic detection criteria form the foundation of an arrhythmia detection system:
(1) high rate ("HR") (2) sudden onset ("SO") (3) rate stability ("RS" ) (4) sustained high rate ("SHR") The HR criterion specifies (through physician-programming) a high 12~ ZS2 rate run length of n consecutive intervals at a heart rate exceeding a selected base rate. For example, n may range from 1 to 255 intervals (beats) at a rate exceeding the boundary rate separating the SINUS and TACH-1 classes. Thus, if n is programmed at, say, 6 and the lower boundary rate for TACH-1 is specified to be, say, 100 bpm, the HR criterion is satisfied if the patient's heart rate exceeds 100 bpm over the course of at least 6 consecutive beats.
The SO criterion consists of a physician-specified step increase (delta change) in the heart rate. The SO criterion is satisfied if the patient's heart rate suddenly jumps by an amount exceeding this delta.
The RS criterion consists of two physician-specified factors, one of which is a run length of n consecutive intervals exceeding a selected base rate, and the other of which is a specified rate stability delta. The RS criterion is satisfied if the patient's heart rate exceeds the specified base rate (which typically would be identical to the minimum boundary rate for the TACH-1 region) over n consecutive beats, and that heart rate does not vary by more than the specified delta rate over those n consecutive beats.
The SHR criterion is analogous to the HR criterion except that the specified run lengths differ. In particular, SHR
specifies a considerably longer run length n than- that used for the HR criterion.
These four basic detection criteria may be combined by ~Z9925Z

Boolean logic into nine tachycardia detection algorithms, as follows tsymbolically in parentheses after each statement of the respective algorithm):
1. high rate (HR);
2. high rate and sudden onset (HR AND SO);
3. high rate and sudden onset, or sustained high rate (~HR AND SO] OR SHR);
4. high rate and rate stability (HR AND RS);
5. high rate and rate stability, or sustained high rate ([HR AND RS] OR SHR);

6. high rate and sudden onset and rate stability (HR AND SO AND RS);

7. high rate and sudden onset and rate stability, or sustained high rate ([HR AND SO AND RS] OR SHR);

8. high rate and either sudden onset or rate stability (HR AND [SO OR RS]);

9. high rate and either sudden onset or rate stability, or sustained high rate ((HR AND [SO OR RS]~ OR SHR).

Algorithms 6 through 9, each containing the SHR criterion, are utiliæed because of the po~sibility that the individual criteria specified by algorithms 2 through 5 may be too stringent for the respective composite algorithm to be satisfied by a reentrant tachycardia. The SHR criterion acts as a "safety valve" to assure that persistent high rate activity will be detected as a reentrant VT by the device.

~29~S2 The arrhythmia detection algorithms are used for deciding that the detected evidence is sufficient to declare that a reentrant tachycardia is in progress. In essence, these algorithms serve to distinguish between arrhythmias which should be treated by the device and those which should not be treated.
As noted above, there may be a rate overlap between a sinus VT
and a reentrant VT, particularly at the lower rates, and hence, rate detection alone does not reliably distinguish between the two. At the higher rates, however, the HR criterion alone suffices as a relible indicator of a reentrant tachycardia.
The detection technique employed applies the principle that the algorithm stringency should decrease with increasing rate and thus with increasing hemodynamic intolerance of the arrhythmia.
In the present embodiment, the physician may specify three different detection algorithms, each for a different tachycardia class. Accordingly, the most stringent detection algorithm of those specified is appropriately assigned to the tachycardia class having the lowest rate range, and the progressively more relaxed detection algorithms are approprlately assigned to the successively higher rate range classes. It follows that if a highly stringent detection test applied to the TACH-1 region results in inconclusive evidence as to whether a reentrant tachycardia i8 in progress, and a moderately stringent test applied to the TACH-2 region is satisfied, the evidence is compelling that a reentrant VT is indeed in progress.
By way of example, FIG. 3a illustrates a suitable selection lZ~9Z52 of detection algorithms for the tachycardia classes TACH-1, TACH-2, and TACH-3 of FI~. la. The most stringent of the algorithms depicted in FIG. 3a, HR AND S~ AND RS, is assigned to the TACH-1 region; the most relaxed test, HR, i5 assigned to the TACH-3 region; and a moderate test intermediate the other two, HR AND
RS, suffices for and is assigned to the TACH-2 region.
Provisions are made for reducing the number of detection criteria to be used in redetection following initial screening of the VT, because of the need to deliver the next therapy as quickly as possible if the VT is still in progress and also because less stringent detection criteria will provide suitably compelling evidence. In particular, if any criterion is no longer applicable as a result of the continuation of the initially detected arrhythmia episode, that criterion is discarded from use in redetection, so long as that episode is still in progress.
For example, the SO criterion i5 not viable after initial detection of a VT inasmuch as the episode has now been detected and sudden onset no longer applies. Hence, that criterion is eliminated during any attempted redetection of the progress of that VT. Similarly, the SHR criterion is of no value for purposes of redetection, once having identified a VT in progress.
This eliminates algorithms 2, 3, and 5 through 9, inclusive, on the above list from use for purposes of redetection.
High rate ~algorithm 1), and high rate and rate stability (algorithm 4), are the only remaining tests suitable for redetection criteria; and since each of HR and RS is useful in 129~Z5Z

both initial screening and redetection, each is assigned two separately programmable n's, namely ni (for initial detection) and nr (for redetection). The reason for the separately selective ~ariables is that, for HR, although a rather long run of consecutive high rate intervals (ni) may be deemed by the physician as necessary for a reliable initial detection of a VT, a relatively shorter run length (nr) will suffice for purposes of redetection. Similarly, for RS, a shorter run of consecutive high rate intervals (nr) of relatively invariant rate may be deemed suitable for redetection, compared to the run length (ni) used for the initial screening.
As with the initial screening algorithms, the physician may spec~fy a different redetection algorithm for each of the VT
classes. In the scheme illustrated in FIG. 3b, the more stringent redetection algorithm HR AND RS is assigned to class TACH-l, and the more relaxed redetection test HR is assigned to each of the TACH-2 and TACH-3 classes.
A different redetection criterion that may be selected in the present embodiment of the invention is referred to as "HYSTERESIS". Redetections at or near the boundary between ad~acent arrhythmia classes may be inconclusive as to whether an acceleration of the arrhythmia has taken place to the next higher class, merely because of minor timing shifts. The "HYSTERESIS"
redetection option assigns a rate "delta" which is- automatically added to the detected tachycardia rate. If, on redetection under the ''HYSTERESISI' option, the redetected rate is below the 1299;~i2 boundary rate between the two classes, the redetection is conclusive that no acceleratlon has occurred. Similarly, if the redetected rate is above that boundary rate but still below the initially detected rate plus the hysteresis delta, there is no acceleration. However, if the redetected rate is both above that boundary rate and the initially detected rate plus the delta, an acceleration to the higher class i5 declared. Hence, the "HYSTERESIS" option is an important feature of the invention for distinguishing accelerations from non-accelerations for arrhythmias which are redetected at rates only slightly above a designated boundary between classes.
According to a further aspect of the present invention, a different set of bas1c criteria, numbering two in the presently preferred embodiment, is provided for purposes of detecting fibrillation. ~ne of the two is termed fibrillation rate ("FR"), and is somewhat analogous to the HR criterion for tachycardia detection. However, for the FR criterion to be satisfied, a physician-specified run length of n consecutive intervals must occur at a heart rate ~xceeding the rate at the upper boundary of the highest VT region of the rate continuum (e.g., a rate higher than the TACH-3 range, in the illustrative rate range partitioning of FIG. lb).
The second basic criterion employed for detection of fibrillation is termed "F x/y". The F x/y criterion specifies that x fibrillation rate intervals must occur within consecutive intervals (where both x and ~ are physician-12992~;2 programmable), as indication of VF. The latter criterion serves to detect VF despite the erratic heart ra~e and widely ~ariable signal amplitude which are characteristic of that arrhythmia.
The varying signal amplitude may result in failure to detect some signal peaks, which might falsely appear to be a VT or periods of sinus rhythm under the FR criterion, owing to restart of the count toward n. The F x/y criterion will recognize, in those circumstances, that x fib rate intervals have occurred within the y consecutive intervals.
It is pos~ible that an erratic, high rate cardiac arrhythmia may go undetected if the rates of the individual cycles swing back and forth between the TACH-3 and FIB regions. That is, the arrhythmia could exhibit a few fibrillation intervals which would reset the VT high rate run counter, but insufficient to trigger the FR or F x/y detection criteria. Next, the erratic arrhythmia could exhibit a few TACH-3 intervals which would reset the FR
rate counter and count as non-fibrillation intervals in the F x/y criterion. This situation could persist for an extended time, resulting in failure to detect a hemodynamically compromising arrhythmia. To insure that such an erratic, high rate, hemodynamically compromising arrhythmia is rapidly detected, and to bias the detection in favor of VF and away from a TACH-3 VT, the following additional rules may be applied to the VT and VF
detection criteria:
tl) A cardiac cycle in the TACH-3 region is completely disregarded by the FR and fib x/y criteria.

~2~39;;~52 (2) If the high rate count is greater than O, and a cardiac cycle i5 in the FIB region, then 1 is subtracted from the high rate counter.
The first rule declares that a cardiac cycle in the TACH-3 region cannot be used as evidence either for or against the detection of VF. The second rule states that an interval in the FIB region offers some (but inconclusive) evidence that a reentrant tachycardia is not in progress.
With these two additional rules in effect, when the erratic rate swings into the TACH-3 class, the detection criteria for VF
are unperturbed. Then when the rate swings back into the FIB
region, the VF detection criteria pick up as if there had been no intervening TACH-3 interval, allowing VF to be detected. While the rate is in the FIB region, the high rate counter is also being counted down, which tends to suppress the detection of TACH-3 when the rate swings back into that region. Therefore, the detection of VF is enhanced and the detection of a TACH-3 event is suppressed.
The two basic fibrillation detection criteria may be combined by Boolean logic into three VF detection algorithms, as follows (symbolically in parentheses following each respective statement of the algorithm):
1. fibrillation rate (FR) 2. fibrillation x out of y (F x/y) 3. fibrillation rate or fibrillation x out of y (FR OR F x/y) The foregoing algorithms provide multiple detection criteria for the region of the heart rate continuum exceeding the upper boundary of the highest tachycardia class.
After detection of an arrhythmia in one of the TACH or FIB
classes, and in response, the delivery of the prescribed therapy sequence (modified by any selected therapy control optionJ, the implanted stimulator must assess whether sinus rhythm has been reestablished. According to the present invention, the criterion of reversion to sinus rate specifies that there must be x intervals at sinus rhythm out of any ~ consecutive intervals.
Both x and ~ are physician-programmable. The sinus x out of y basic criterion ("sinus x/y") is also the only algorithm used for the purpose of this detection. It is preferable that relatively long counts be used for the programming of x and y; for example, the physician might set x at 18 and ~ at 20. The reason for this is that there i5 no clinical harm in a slight delay to provide a more reliable indicator of reversion to sinus rhythm, in contrast to the undesirability of delaying detection of a VT (particularly if poorly tolerated hemodynamically by the patient), or of VF.
Returning now to the therapy aspects of the present invention, an overall philosophy incorporated into the presently preferred embodiment is to provide the physician with virtually complete control over the aggressiveness of the therapy, within the availability of programmable individual therapies and therapy regimens. Consistent with that philosophy, each of the therapies is provided with a selectively modifiable fine structure. In the ~Z9~25Z

presently preferred embodiment, the fine structures that may be prescribed by the physician for shock-tyE~e therapies include:
(1) number of shocks to be delivered (i.e., the number of attempts to terminate t}~e detected arrhythmia);
(2) sense-to-shock delay (in milliseconds), to time the delivery of the shock relative to a particular portion of the ECG waveform (such as for synchronizing cardioverting shocks with that portion most likely to result in rapid termination of the tachycardia, while avoiding the vulnerable period);
(3) amplitude of shocking pulses for first attempt, and amplitude for subsequent attempts to terminate the arrhythmia (e.g., specifying a low level on the first attempt, and a relatively higher level if subsequent attempts are necessary, in keeping with increasing the aggressiveness of the therapy);
(4) definition of the waveform of the shock applied to the heart (e.g., specifying up to 5iX phases including pulse width, polarity, and presence (i.e., on) or absence (i.e., off). For example, the waveform could be specified as a tri-phasic waveform with a positive pulse, a negative pulse anc~ another positive pulse; or a sequence of two pulses separated by a gap as by specifying a negative pulse for a first pulse width, no pulse for a second pulse width, and a negative pulse for a third pulse width, and so forth.

~Z99252 Referring now to the pacing therapies, a detected atrial reentraIIt tachycardia may be terminated with a properly timed burst of stimulation pulses. The number of attempts may be programmable, and a burst may be programmed with respect to number of pulses in the burst, sense-to-initial pulse delay interval, and burst cycle length (i.e., the pulse-to-pulse interval). Further, both the initial delay interval and initial burst cycle length may be programmed wit~l fixed values, or with adaptive values as a percentage of the interval between tachycardia beats; and may be changed in a scanning mode in which the values of either or both of those parameters are decremented, or alternately incremented and decremented, over successive bursts after the initial burst (the "burst scanning" mode). A
mode may be selected in which the interval between consecutive pulses in a burst is automatically decremented (hence, termed "auto-decremental" mode). Burst scanlling mode may also be programmable with respect to parameters such as step size (i.e., amount by which value of delay interval and/or burst cycle length is increased or decreased for each successive burst), number of steps (i.e., number of times those values will be increased or decreased by the selected step size), and number of sequences (i.e., number of times a scan sequence is to be repeated).
The present invention utilizes improvements in physician-programmable fine structures for the pacing therapies, as follows:
1. number of attempts (1-255);

2. decremental ~DEC) or search (SEARCH) scan;
3. coupling delay/fixe(l or adaptive;
. S1 burst cycle length (BCL)/fixed or adaptive; and S1 number of pulses (o-255);
5. S2 BCL/fixed or adaptive; and S2 number of pulses (0-255);
6. S3 BCL/fixed or adaptive; and S3 number of pulses (0-255);
7. S4 BCL/fixed or adaptive; and S4 number of pulses (o 255);
8. scanning step for coupling delay/fixed or adaptive;
9. scanning step for S1 BCL/fixed or adaptive;

10. scanning step for S2, S3, S4 BCLs/fixed or adaptive;

11. auto-decremental (AUT0-DEC) step for S1, S2, S3, S4 BCL/fixed or adaptive;

12. minimum interval/fixed or adaptive;

13. number of steps to scan (0-255);

14. add S1 pulse per attempt.
In the foregoing list of programmable fine stuctures for the therapies, the number of attempts refers to attempts to terminate a VT in any TACH class in which the pacing therapy is applied.
The fine structure of the system's burst pacing therapies provides two kinds of scanning behavior -- (1) decremental scan which starts with the initial burst definition and in which intervals become progressively shorter by the selected step amount; and (2) search scan which similarly starts with the initial burst definition and in which the intervals become progressively longer and shor~er as they are alternately incremented and decremented in a search for the termination zone of the tachycardia in the ECG.
Referring to FIG. 4, the coupliny interval is the time interval between last detection of the tachycardia (sense event) and delivery of the first pulse in the burst. The first pulse train in each scan is termed Sl, and sequential pulses in that train are separated by identical time intervals referred to as the S1 burst cycle length (BCL). ~'hus, S1 BCL is the pulse-to-pulse interval for train S1. The BCL may be physician-specified either as a fixed time interval or as an adaptive percentage of the measured tachycardia rate, and the number of pulses in the Sl train may be separately specified. Further, the physician may specify three additional stimuli in each scan, these being the successive pulse trains S2, S3 and S4 following the Sl train. As in the case of S1, the BCL (which will differ for each train) in either fixed or adaptive mode, and the number of pulses, may be programmed for each of S2, S3 and S4. The combination of the coupling interval and the succession of pulse trains following any given sense event is the entire burst.
Still further, separate scanning steps may be specified for each parameter within a burst, that is, for the coupling interval and the B~L for each of the pulse trains; and the coupling delay and, for any given train within the burst, the BCL may be selectively fixed or adaptive. If an auto-decremental step is 12~2~

.

selected, the BCL in each of the trains Withill the burst to which that selection applies becomes proyressively shorter (by a fixed or an adaptive amount, as specified by the physician at the time of programming ~he device) when and as tlle therapy is delivered.
In addition, the present embodiment allows an auto-decremental burst to be programmed for scannirlg (i.e., the burst may be scanned as it or selected parameters are decremented).
For the sake of clarity, scanning will be described with reference to an exemplary burst. In the case of decremental scanning the originally specified burst is delivered to the heart, and if that fails to terminate the tachycardia, the BCL of successive pulses (and coupling interval, if so programmed) in the next burst delivered is automatically decreased by the specified step amount, i.e., the yulses are closer together in the next burst. Successive bursts are similarly decremented relative to the immediately preceding burst by the designated step for the selected parameters, until successful arrest of the tachycardia or completion of the treatment sequence.
In the case of search scanning, if the first attempt (burst) is unsuccessful, the next burst is delivered with a longer BCL
and/or coupling interval (depending upon the specific parameter programming) by the specified step amount. Subsequent bursts, up to the point of successful termination or of completion of the treatment sequence, are alternately shorter and longer in the specified parameter(s) by the step amount, e.g., the ne~t longer BCL burst has a time interval between pulses which is greater by ~29~252 the step amount than that interval in the immediately preceding longer BCL burst (those two bursts being separated by a shorter BCL burst in the alternation cycle). However, in the case of auto-decrementing the BCL is decreased by the speci~ied step amount for each pulse in a single burst.
The minimum interval sets a lower limlt on the scanning, and on the step reductions; hence, that interval defines completion of the treatment sequence. When the BCL reaches 50 % of the time interval between beats of the VT, the risk becomes considerably greater that the VT will be accelerated to VF. Accordingly, it is desirable to set the minimum interval to remain safely above that critical level as determined for the particular patient. By appropriate selection of the number of steps to scan, coupled with setting of the minimum interval, the aggressiveness of the treatment is selectively adjusted. The capability is also provided to add a single pulse to the 51 train in each burst.
The premise here i8 that if the preceding burst was unsuccessful, the next attempt should be at least slightly more aggressive; and the more pulses in the train the more aggressive the therapy.
The timing of the sensing, arrhyth~ia detection, therapy preparation and therapy delivery is illustrated for each cardiac cycle by the flow diagram of FIG. 5.
Quiet time: Monitor input leads for high frequency noise.
If noise is detected, then go to idle time. If no noise is detected, go to sense time.
Sense time: Wait until a cardiac event is sensed, or until ~2992~

it has been longer than the bradycardia interval since the last sensed or paced cardiac event. Record the length of the cardiac cycle and save it. If a cardiac event was sensed, then go to therapy time; otherwise go to bradycardia pace time.
Idle time: Note that noise was sensed, then wait until it has been longer than the bradycardia interval since the last sensed or paced event. Go to brady pace time.
Brady pace time: If appropriate based on selected mode, generate a pacing pulse.
Therapy time: If a therapy was set up during common refractory time in the previous cycle, then deliver it now. If a therapy was delivered or a brady pacing pulse was emitted this cycle, then go to pace refrac-tory time; otherwise go to sense refractory time.
Sense refractory time: Update diagnostic counters assoc-iated with a sense event. Go to common refractory time.
Pace refractory time: Wait until the beginning of the T-wave window. Go to common refractory time.
Common refractory time: Use the interval just measured to update the arrhythmia and sinus detection criteria and algorithms. If an arrhythmia is detected or redetect-ed, prepare the appropriate therapy to be delivered synchronously during the next -therapy time. Wait until either the sense refractory time or the end of the T-~ Z99Z52 wave window. Go to update time.
Update time: Based on the sensiny/pacing history of thecycle up to this point, update the automatic gain control system. Perform miscellaneous initialization chores for the next cycle. Go to quiet time.
In the above sequence, if an arrhythmia is detected in the previous cycle, the therapy is delivered in the current cycle regardless of whether noise or a rate timeout was experienced.
External control of certain physician-designated functions may be given to the patient for selection by means of a magnet to be used in conjunction with a reed switch within the implanted cardiac stimulator. Any such functions will only be magnet-accessible to the patient if so programmed by the physician. In the stimulator of the invention, an elective replacement indica-tor (ERI), an end-of-service indicator (EOS), and a capture veri-fication test (i.e., to assure that pacing stimuli are producing the desired response in the excitable cardiac tissue), are among the functions which may be magnet controlled by the patient, for purposes of transtelephonic monitoring of the device. The ERI/EOS/ capture verification tests available in the device are:
1. Low battery detected by the ERI co~parator (this comparator detection is ignored by the microprocessor for a period of several hours after charging of the capacitors, during which the indication may be erroneous).
2. Post-charge battery voltage below preset limit.

~9925;~:

3. During high voltage charge-up, microprocessor-requested voltage level not reached within a specified short interval (e.g., 30 seconds), indicating low battery.
4. Total charging time for all shocks exceeded a preset limit.
The ERI/EOS and capture tests are performed by the device as follows:
1. If the magnet is applied and ERI/EOS is not detected, the device will generate a total of 12 pacing pulses in the VVI mode at a rate of 100 pulses per minute (ppm), with the last pulse width halved for purposes of the capture test.
2. If the magnet is applied and ERI/EOS is detected, the device wlll generate a total of 4 pacing pulses in the VVI mode at a rate of 100 ppm, with the last pulse width halved for capture test.
Other potential magnet functions available in the preferred embodiment, which are mutually exclusive, are:
1. Perform none of the other functions in this category.
2. Inhibit detection of arrhythmias and delivery of therapies.
3. Enable detection of arrhythmias and delivery of therapies.
4. Reduce tachycardia detection and redetection lZ~Z52 algorithms to "high rate" only.
5. Cause delivery of one of the four therapies;
the selected therapy being programmable. If a scanning burst is selected, the first burst in the scan will be delivered. For a shock, first energy will be delivered.
The presently preferred embodiment of the cardiac stimulator is structured to be implanted in the patient. A unit including the portion of the stimulator for detecting and distinguishing the significance of the patient's cardiac activity, and respon-sive to abnormal arrhythmias for generating and managing the delivery of pacing and shoc~ therapies, with self-contained power source, may be assembled and housed in a metal case inert to body tissue and fluids. That unit is sometimes referred to herein as a "multiple carcliac therapy generator", or more simply as the "therapy generator" (although its functions go beyond mere thera-py generation), and in that respect is somewhat akin to the pulse generator unit of a cardiac pacemaker. Lead/electrode assemblies for sensing cardiac activity and for delivering the respective pacing and shock impulses to the patient ' 9 heart may be separably connectable to the therapy generator, and in that respect are somewhat akin to the leads of a cardiac pacemaker. Together, the therapy generator and the lead/electrode assemblies constitute the cardiac stimulator.
The therapy generator includes a digital control section for storing and executing software instructions and for storing and ~99zs~

processing the data for all digital functions of the device (aside from those functions which, for purposes of conserving memory capacity, are readily consigned to an external programmer unit ("programmer"1 of conventional type available to the physician). The digital functions of the device include the previously described physician-programmable aspects, such as provision for programming the rate boundaries of the VT (TACH) classes to selectively partition the rate continuum, the therapies (including gross and fine structures) and therapy delivery sequences, and the detection and redetection algorithms, as well as various processing, timing, switching, control and other functions to be described presently.
The therapy generator also includes an analog portion for such functions as monitoring the patient's ECG signal information over each cardiac cycle, enhancing that signal information while eliminating noise and other interference through signal filtering and automatic gain control, developing the respective impulse waveforms to be delivered for the pacing and shock therapies, transmitting data between the device and external units such as the programmer and transtelephonic monitoring equipment, and protecting against overloads, at least some of these analog functions being controlled according to the programmed instructions. Also included are the battery cells, and voltage regulation and priority power sequencing section, for supplying power to the other sections of the overall generator.
The electrical circuit configuration of a present embodiment 12~9tZ5X

of the overall cardiac stimulator will now be described by reference to the remaining FIGS. of the drawing. With initial reference to FIG. 6, the therapy generator includes a section 10 comprising a central microprocessor with associated memory capacity including random access memory (RAM) and read only memory (ROM), for processing and storing data necessary to provide the features described earlier herein. The micropro-cessor and memory circuits are preferably complemetary metal-oxide-semiconductor (CMOS) integrated circuit chips for low voltage, low power operation. Section 10 is bidirectionally coupled to a programming and data transmission section 14 which serves to transmit data to receiving and monitoring equipment ~e.g., transtelephonically) for analysis and assessment of the cardiac functions of the patient and the operating condition of the implanted device, and for receiving program instructions and data from the external programmer, via an implanted antenna 1~.
A crystal oscillator 20 electrically coupled to section 10 provides the necessary precise timing signals for system opera-tion. A reed switch 22 is also electrically connected to section 10 to permit limited external control by the patient of certain programmable functions, such as those previously described, by using an external magnet to control actuation of the switch.
A sense amplifier section 25, which includes automatic gain control and bandpass filtering, is coupled to section 10 for transmitting ECG signal information to the microprocessor and for receiving control signals from the microprocessor. The sense ~z99~z amplifier is also connected to data transmission section 14 so that the ECG telemetry siynal information may be supplied via the latter to external monitoring equipment. A quad comparator within section 25 provides the li~k for converting the ECG sense signal information obtained from the sensing electrode(s) and processed by the sense amplifier into digital information suit-able for use by the microprocessor. The microprocessor is within a feedback loop of the sense amplifier to provide improved auto-matic gain control, as will be explained in greater detail below.
The sense amplifier enhances the ECG signals to aid the tracking of signal content of rapidly varying amplitude, such as fibrillation signals. Preferably, the sense amplifier has a gain range on the order of 60:1. In addition, bandpass filtering is employed to provide the dual function of (1) reducing the amplitude of signals outside the frequency band of interest and ~2) further amp:lifying the low frequency (e.g., fibrillation) signals within that band in the absence of normal R-waves.
The power source section 28 of the overall stimulator system comprises high rate battery cells, a voltage regulator and a priority power sequencer. The high rate cells comprise any combination of cells capable of delivering sufficient energy to charge the capacitors in the output high voltage section (40) within a reasonable time (e.g., 20 seconds or less). The voltage regulator circuit has a voltage divider to provide a 3:1 reduc-tion if three cells are used in series, as is preferred, or a 2:1 reduction if only two cells are employed, and thereby improves ~2~9252 power source efficiency. The priority power sequencer assures adequate power is made available to essential circuit functions such as the control logic during periods when there would otherwise be high current drain on the cells, such as during charge up of the high voltage capa~itors in preparation for the delivery the defibrillating or cardioverting shock therapies.
The pacing section 31 of the system includes a voltage multiplier and output section, the former serving to scale up the regulated supply voltage from power source section 28 by multiples of one, two or three. The output section provides the output switching from this scaled voltage to deliver the pacing stimuli to the patient's heart via the pacemaker circuit including cathodic stimulating and anodic reference electrodes, under the control of the microprocessor.
An analog rate limit circuit 35 between microprocessor/
memory section 10 and pacing section 31 is employed to control-lably limit the pacing rate, and thereby safeguard against pace-maker runaway, :in the event of failure of the crystal oscillator circuit. However, the rate limiter is automatically disabled whenever an intentionally high rate of paciny pulses is required, such as during the generation of a burst pacing therapy.
The leads~ for the pacing and sensing electrodes are electrically monitored by the isolation/protection section 37 to protect low voltage, low power components of the stimulator from the high voltage of the defibrillating shocks generated by the stimulator (or applied from an external defibrillator that may be ~9~i2 used on the patient during emergency medical procedures).
The cardioverter/defibrillator shock therapy portion of the stimulator includes an isolated high voltage generator and output section 40. The voltage generator circuitry includes a high voltage oscillator coupled via an isolation transformer to output capacitors for charging the capacitors to the required voltage levels for the cardioverting and defibrillating shocking pulses, under the control of the microprocessor.
A low power analog-to-digital (A/D) converter in section 40 is utilized to monitor the voltage on the capacitors, to permit the microprocessor to set the desired high voltage output level in accordance with the physician-programmed fine structure energy content of the applicable shock therapy. Monitoring of the capacitor voltage also allows the microprocessor to measure the residual charge on the capacitors after delivery o~ each output pulse, and thereby to estimate the amount of energy consumed in the delivery for ongoing assessment of remaining capacity of the battery cells. In addition, the A/D converter input circuit may be switched by the microprocessor for connection to the power source section 28 to monitor the battery voltage, and thereby determine the present condition of the cells.
Output section 40 also contains level shifters and isolation transformers to convert the microprocessor-supplied low level logic control signals to the control signal levels required to drive the output switches of that section. The output switches themselves are of low "on" impedance and capable of handling the high voltages and currents being generated, to control the deliv-ery and polarity of each output pulse. A short circuit protec-tion circuit is provided in output section 40 to open the output circuit in the event that the current through that circuit rises above a predetermined level. This prevents a discharge of the capacitors into a very low impedance -- such as if the defibril-lator patch electrodes were shorted -- and thereby protects the output switches from overstress and potential destruction.
The sense amplifier with AGC and related bandpass filtering and quad comparator, and including the relationship with the microprocessor, is shown in greater detail in FIG. 7. The sensed ECG waveform components detected by the sensing electrode(s) are applied to the sense amplifier 60 via an input circuit 63. The gain of the sense amplifier is automatically controlled by a feedback loop 65 which includes the microprocessor 68 of section 10. The ECG signals processed by the sense amplifier are additionally enhanced by a filtering section 70 which includes a primary high gain bandpass amplifier 73 to reduce signal strength outside the selected band and to amplify low frequency signals within the band. The output of amplifier 73 is split and fed into separate bandpass amplifiers 75, 76, one of which (75) is digitally controlled by the microprocessor. The output of the gain/filtering control stages is applied to the quad comparator 80, which develops three inputs to the microprocessor in the feedback loop.
The AGC system of the present invention deals with the ~L2~ S~

difficult problem of sensing the low frequency, low signal amplitude characteristic of V~. Under ordinary circumstances, a loss of sensing may be indicative of VF, requiring that the gain of the sense amplif~er be increased to enable better detection.
If the loss of sensing is attributable to an intermittent heart block rather than fibrillation, a return of sense signal is likely to be overamplified with a consequent serious perturbation of the entire system. The manner in which the AGC system resolves this problem will be explained presently, in connection with the description of FIGS. lO and 11.
Sense amplifier section 25 comprises an AGC amplifier section (FIG. 8), a bandpass amplifier section (FIG. 9), and a quad comparator section (FIG. 10). Referring to FIG. 8, the AGC
amplifier section includes some initial bandpass filtering and a blanking circuit to block the large amplitude pacing and shock signals. This prevents amplifier saturation and thus decreases post-pace amplifier recovery time. The AGC amplifier gain is controlled by varyiny the gate voltage of an N-channel junction field effect transistor (JFET) 100 which acts as a voltage controlled input resistor to a non-inverting amplifier 101. The microprocessor controls the on/off duty cycle of switches 103 and 104, which set the gate voltage of JFET 100 by charging and discharging capacitor 106 to a voltage between V-~ and Vc. This technique is used to obtain a gain range of 60:1 as determined by the resistance of resistor 107 and the on impedance of JFET 100.
Switches 108 and 109 serve to prevent the large amplitude ~Z99252 pacing and shock signals from entering the amplifier input. To that end, immediately prior to delivering a pacing or shock output, switch 108 is switched from its normally closed state to an open state and switch lo9 is switched from its normally open state to a closed state. Consequently, the amplifier 101 input is disconnected from the lead system by switch 108 and grounded by switch 109. The two switches are returned to their normal states a few milliseconds after completion of a pace output and a few hundred milliseconds after completion of a shock output.
The initial bandpass filtering functiorl of the AGC amplifier section is provided by the circuit consisting of resistors 110 and 111 and capacitors 112 and 113. Resistor 115 is used for proper biasing of the JFET circuit. Although the switches are schematically depicted as mechanical devices in the AGC amplifier circuit of FIG. 8 and in some of the other circuit diagrams, it will be understood that in practice electronic switches (such as transistors) typically would be employed.
Referring now to FIG. 9, the bandpass amplifier section of the sense amplifier has a programmable sense margin feature and special bandpass characteristics which aid in tracking the variable amplitude fibrillation signal. The section includes two active bandpass filter amplifiers 130 and 131, a programmable gain DC amplifier 135, and a passive high pass filter comprising capacitor 13~ and resistor 138. This bandpass configuration has the advantage of reducing the amplitude of the signal components outside the frequency band of interest, and effectively increas-12~9~2 ing the sense margin for low frequency fibrillation signals(where higher frequency QRS signals are absent). In addition, the sense margin (i.e., the ratio between the inner and outer targets to be described below) is also selectable by the microprocessor in that the gain to the inner target comparators can be set separate from that of the outer target comparators, by setting the magnitude of resistance 140 around DC amplifier 135.
Alternatively, the effective ratio of the targets may be changed by varying the gain of bandpass filter amplifier 131 (by varying the value of resistor 142 around that amplifier), or by changing the target reference voltages themselves.
With reference to the circuit diagram of FIG. 10, and the exemplary input signal with logic outputs illustrated in FIG. 11 for the comparator circuit of FIG. 10, the quad comparator section of the sense amplifier consists of two comparator pairs, inner target comparators 150 and outer target comparators 151.
The logic outputs LIU and LIL of the inner target comparators are used by the microprocessor as valid sense input signals. The logical "OR" (Lo) of the output target comparators is used by the microprocessor to evaluate the need for increasing or decreasing the AGC amplifier gain. During sinus rhythm, the amplitude of the QRS complex dictates the gain setting of the AGC amplifier because of its relatively large amplitude and its frequency content. The same situation exists during sensing of a tachycardia. However, the presence of the lower frequency fibrillation signal causes an effective doubling of the sense 12~925~

margin (because of the "special bandpass" filters 141 and 142 (FIG. 9)) which permits more reliable trackiny of the variable amplitude fibrillation signal.
A problem area that must be considered is the adjustment of gain during bradycardia pacing. In this case, the lack of sensed events could be attributable to slow rate or to an inadequate amplifier gain setting. To determine which of these is responsible, sensing of lower frequency post-pace T-waves is performed. If no T-wave is sensed at the inner target comparators in a preset time window following a pace event, the AGC gain is increased. This is continued until the T-waves are sensed or until the previously undetected rhythm i5 sensed.
The major difference between AGC gain corltrol using T-waves versus QRS and fibrillation signals is that the T-wave amplitudes are controlled about the irlner target comparator instead of the outer target comparator. This insures the amplifier gain will not be set too high should an intrinsic QRS signal be sensed.
A number of different lead connections are supported by the design of the embodiment described herein. Referring to FIG. 12, the output connections available from the therapy generator are designated "PACE", "REF", "HV1", "HV2" and "HV3". Among the more significant potential lead configurations that may be employed are the following:
(1) One myocardial screw-in or endocardial lead and two titanium (or other conventional material) mesh patches. The myocardial lead is placed on the apex of the left ventricle (LV) or the ~2~9~52 endocardial lead is placed transvenously at the apex of the right ventricle (RV) (in either instance, conrlected to "PACE"), one patch on the epicardium of the LV (connected to "REF" and "HVl") and one patch on the epicardium of the RV (connected to "HV2" and "HV3"). Pacing and sensing are from myocardial (anode) to LV
patch (cathode). Cardioversion and DF are from LV to RV (right ventricle) patch.
(2) Two myocardial screw-in leads or one bipolar endocardial lead, and two titanium mesh patches. Both myocardial leads are placed in the apex of the ventricles or the bipolar endocardial lead is placed transvenously at the apex of the RV (in either instance, one electrode connected to "PACE" and the other to "REF"), one patch on the epicardium of the left ventricle (connected to "HVl") and one patch on the right ventricle (connected to "HV2" and "HV3"). Pacing and sensing are from myocardial to myocardial lead. Cardioversion and DF are from LV
to RV patch.
(3) Same as (1) above, with the patches placed on the pericar-dium instead of the epicardium.
(4) Same as (2) above, with the patches placed on the pericar-dium instead of the epicardium.
(5) One two conductor transvenous lead and one titanium mesh patch. The tip electrode of the transvenous lead (connected to "PACE") is placed at the apex of RV, the ring electrode of the transvenous lead (connected to "REF" and "HVl") is placed at the upper part of RV, and the patch (connected to "HV2" and "HV3") is placed epicardially, extra-pericardially or subcutaneously such that shock current will flow through the ventricular septum.
Pacing and sensing are between the transvenous tip and ring.
Cardioversion and DF are from transvenous ring to patch.
(6) One two conductor transvenous lead and two titanium mesh patches. The tip electrode of the transvenous lead (connected to "PACE") is placed at the apex of the RV, the ring electrode of that lead (connected to "REF" and "HV1") is placed in the upper RV, and the two patches (one connected to "HV2" and the other to "HV3") are placed epicardiaily, extra-pericardially or subcutaneously on the right and left sides of the ventricles.
Pacing and sensing are between the transvenous tip and ring.
Cardioversion and DF are from the transvenous ring to each patch (separate waveforms).
(~) One three conductor transvenous lead and one titanium mesh patch. The tip electrode of the transvenous lead (connected to "PACE") is placed at the apex of the RV, distal ring electrode (connected to "HV1" and "REF") is placed in the upper RV, the proximal ring electrode (connected to "HV2") is placed in the fsuperior vena cava, and one patch (connected to "HV3") is placed epicardially, extra-pericardially or subcutaneously such that shock current will flow through the ventricular septum. Pacing and sensing are between the transvenous tip and the distal ring.
Cardioversion and DF are from transvenous distal ring to either the proximal ring or the patch (separate output waveforms).
The high voltage generation and output section is 12~252 illustrated in greater detail in the circuit diagram of FIG. 12.
That section comprises an isolated high voltage generation circuit 200, a transformer isolated switch driver circuit 203, two output capacitors 204 and 205 with bleeder resistors 206 and 20~, three protection diodes 210, 211 and 212, an output short circuit protection circuit 215, and two pairs of high voltage output switches 220, 221 and 224, 225. When the need for a shock therapy has been determined, the microprocessor enables the high voltage generation circuit 200 to charge output capacitors 204 and 205 to a preset value. ~fter the charging is completed, the prescribed output shock is delivered by closing the appropriate switch pair. Depending on how HV1, HV2 and HV3 are connected externally, a variety of output combinations can be accomplished.
It will be noted that if HV2 and HV3 are connected together, the closing of switch pair 220, 221 produces an output of opposite polarity to that produced by closing switch pair 224, 225. In addition, a hardw~re option 230 is included to enable selection of either a full or half amplitude for the output via switch pair 220, 221. This allows the generation of biphasic output waveforms of approximately half the amplitude in one direction compared to the other direction.
Referring now to FIG. 13, the transformer isolated switch driver circuit 203 comprises two transformers 240 and 241, each having one primary driver circuit, two secondary driver circuits and one secondary overvoltage protection circuit. An input circuit 244 assures non-overlap between the signals controlling 5~

~29~2S2 the pairs of output switches 220, 221 and 224, 225. Level shifted logic circuits from the microprocessor drive the logic input to the two primary drivers. The transformer isolation p~ovided by 240 and 241 protects the low voltaye, low power electronic components from the high voltage outputs.
FIG. 14 shows the primary and~secondary drivers of each transformer in greater detail, although for the sake of simplicity only the applicable portion of transformer 240 is depicted. The leading edge of a negative logic pulse at the input of the primary driver causes the negative end of capacitor 251 to be pulled from V+ to V-. This transfers energy from the capacitor to the primary coil winding and consequently to the secondary winding (through the ferromagnetic core on which the windings are wound). The result is a positive voltage across the secondary circuit which turns on P-channel transistor 254 and provides forward bias for the parasitic diode across P-channel transistor 255 charging capacitor 257 and the gate capacitance of N-channel transistor ~switch) 221 while holding PNP transistor 258 off. The positive voltage at the gate of transistor 221 turns that switch on, allowing current to flow from the output switches to the external load (i.e., the selected lead/electrode configuration and the patient's heart, as described with reference to FIG. 12).
On the trailing edge of the input logic pulse, the negative end of capacitor 251 (which is now fully charged) is pulled to V+, producing a negative voltage across the secondary circuit.

129~252 This turns on transistor 255 and provides forward bias for the parasitic diode across transistor 254, reverse charging capacitor 257, turning on transistor 258 and dumping the gate capacitance of transistor 221 which turns off that switch. This disconnects the high voltage capacitors 204, 205 (FIG. 12) from the external load, thus ending the shock output pulse. In addition, transis-tor 258 prevents the gate voltage of transistor 221 from being affected by capacitively coupled signals impressed on its drain.
The secondary overvoltage protection circuit 260 comprising zener diodes 263 and 264 guarantees that the voltage applied to the secondary driver of the respective transformer (240, in FIG.
14) is not large enough to break down any of the transistors in the switch driver circuit. This permits use of larger transformer winding ratios so that circuit performance is not diminished with partial battery depletion. It should be noted that the 4049 inverters 245, 246 are configured to provide the high current drive requirements of the emitter followers they are driving.
The output short circuit protection circuit 215 of the high voltage generation and output section (FIG. 12) is shown in greater detail in FIG. 1~. This protection circuit guards against excessive current flow through (and thus, against damage to) the output switches. Current from the high voltage output capacitors 204, 205 through the output switches to the external load must flow through low impedance resistor 266. Transistors 267 and 268 and resistor 269 form a discrete silicon controlled rectifier (SCR). If the current through resistor 266 increases 25;~

enough to turn on transistor 267 through the divider consisting of resistors 270, 271, the SCR latches on and pulls down on the gates of output switches 221 and 225 (through diodes 2~3 and 274), discharging their respective gate capacitances and turning them off. This causes the voltage across resistor 266 to fall because of the reduced current through that resistor, and once the current through the SCR returns to a very low value, the SCR
turns off and i5 ready to be triggered again.
Referring now to FIG. 16, the isolated high voltage generation circuit 200 provides the means to charge the high voltage output capacitors 204, 205 (FIG. 12~ to a preset value.
The capacitor voltage monitor 280 comprises a digital-to-analog (D/A) converter which is controlled by the microprocessor. The output of the D/A converter goes to one input of a voltage comparator. The other input to the comparator may be connected to a divided down version of the output capacitor voltage for controlling the charge voltage, or to the battery (V-) for battery condition monitoring. A high impedance referencing resistor 283 provides a reference to the V+ voltage.
When configured for capacitor charging, the microprocessor presets the desired voltage value and enables the high voltage oscillator 28~ to charge up the capacitors using the flyback circuit comprising transformer 288, N-channel transistor 290, battery cells 291, 292 and 293, filter capacitors 295, 296 and 29~, and high voltage oscillator circuit 287. To accomplish this, transistor 290 is turned on and current is allowed to flow ~.~992S2 through the primary winding of transformer 2B8. When this current has risen sufficlently, transistor 290 is abruptly turned off and a very large flyback voltage develops across the primary (and consequently across the secondaries) of the transformer.
The voltages across the secondaries are half wave rectified by diodes 301, 302, 303, 304, to provide a single direction charge transfer to the capacitors 204 and 205 (FIG. 12), forcing them to charge to a DC voltage.
When the voltage monitor comparator 280 signals the microprocessor that the requested voltage has been reached, the high voltage oscillator 287 is disabled and the output is delivered. It should be noted that the high voltage oscillator is also intermittently disabled by the low voltage regulator circuit to insure priority to the control circuitry power source.
Fuse 308 provides protection against overheating of the battery cells 291, 292 and 293 because of excessive current drain caused by some circuit failure. Diodes 310, 311 and 312 provide a low impedance path around a respective battery cell if it becomes depleted. This allows more efficient high voltage charging in the event that one cell has failed. The third secondary of transformer 288 provides a positive voltage source for the high voltage oscillator circuit.
High voltage oscillator circuit 287, illustrated in greater detail in the circuit diagram of FIG. 17, centers around a 7556 timer IC 325 which provides the output pulse train that drives the gate of N-channel transistor 290 (FIG. 16). It will be noted 12992~Z

that a high capacitance load driver comprising inverter 328 and transistors 330 and 331, similar to the driver used in the output switch driver circuit (FIG. 14), is used here to drive the gate capacitance of transistor 290 (FIG. 16). This circuit is needed to turn that transistor off fast, which improves the efficiency of the flyback operation. Transistor 290 is also selected to have a very low drain to source "on" impedance since switch drop losses may greatly diminish efficiency.
One-half of timer 325 is configured to run astable at a preset rate determined by associated resistance and capacitance, and to trigger the other half of the timer to produce an output pulse whose duration is determined by resistor 335 and capacitor 336 and which controls the switch driver. As the output capacitors charge up, energy is transferred more quickly out of the transformer core. This causes the negative transition of the flyback voltage (occurring at VTp) to occur sooner. If this transition occurs before the astable portion of timer 325 times out, transistor 338 is turned on, resetting this timer. In this stàte, the oscillator continues to speed up as the capacitors complete their charging, producing a more efficient operating condition (since the fixed rate would have resulted in wasted time).
A regulated voltage (generated by resistor 340, transistors 341 and 343, diode 345, resistor 347 and capacitor 348) is provided to the timer IC 325 to make circuit performance independent of battery voltage (over the usable range of the ~2992S~

cells). The positive secondary voltage (VTs) provides the regulator circuit with ample voltage overhead to make this possible. To enable the high vGltage logic circuit, a positive logic signal is applied to the gate of N-channel transistor 351.
This turns that transistor on, providing power to the circuit and removing the reset condition (at RESET1 of timer 325). If the low voltage regulator requests that the high voltage oscillator be temporarily disabled, the logic signal at the gate of transistor 351 is brought low which immediately resets the timer 325 (via transistor 354) but does not remove the power to this circuit until capacitor 356 charges up to shut transistor 358 off. This allows short interrupts of the oscillator without disturbing power, which also improves the circuit efficiency.
Referring now to FIG. 18, the voltage regulator/power priority sequencer section comprises ~ive switches 371-375 (controlled by low power logic) to charge up a capacitor 377 to a preset value (Vreg)~ Capacitors 378 and 3~9 are much smaller than capacitor 3'~7 and are used in con~unction with capacitor 377 to divide the supply voltage by three, and to transfer charge to the latter capacitor a small amount at a time. When capacitor 377 is being charged, switches 371-375 are switching between their respective A and B states. When the regulated voltage monitor senses that capacitor 377 is at the appropriate voltage level, it signals the switch control logic to go to the standby mode where switches 371, 372 and 3~4 stay in state B and switches 373 and 375 go to their C state to stop the charging of capacitor ~L299~S2 377. The regu~ated voltage monitor then watches Vreg and, via the switch control logic, assures that Vrey is maintained at this preset value by enabling the charge mode whenever it is needed.
If the supply voltage drops below a level which will support this division by three (i.e., Vreg could not be properly regulated), the supply voltage monitor signals the switch control logic section 382 which changes to the direct regulation mode. In this mode, switches 372 and 3~4 are in state B, switches 373 and 375 are in state C, and switch 371 switches between states B and C, charging capacitor 377 directly from the supply voltage. This mode is much more inefficient but is designed to assure that Vreg generation has high priority (even if more power is required).
In addition, if the regulated voltage monitor is requesting that capacitor 3~7 be charged, switch control logic 382 generates a logic inhibit signal and sends it to the high voltage oscillator circuit (a circuit that causes major drain on the supply when active) and disables its operation until the Vreg voltage is back to the desired level. This assures that Vreg (which provides power to all of the very important control logic for the cardiac stimulator, including the logic that controls the high voltage operation) is given priority, for safe operation.
If it were desirable to configure the regulator circuit to divide the supply voltage by two instead of three (l.e., use two cells rather than three), capacitor 378 may be left out and a jumper connected between nodes 386 and 387.
The present state of the art pertaining to high voltage ~Z99252 components, such as the battery cells, capacitors, transformers, and so forth, dictates a relatively large size therapy generator for the cardiac stimulator compared to the small present-day pacema~er models. This tends to necessitate implantation of the device in the patient's abdomen, rather than in the pectoral region as is customary for pacemakers.
Although a presently preferred embodiment of the invention has been described, it will be apparent from to those skilled in the field to which the invention pertains, that variations and modifications may be made without departing from the principles of the invention. Further, as technological advances are made, for example, in developing practical small-size, low-cost high voltage components, similar to the advances in the semiconductor field, the principles of the invention may be applied directly to a "universal" implantable device for performing an all-purpose cardiac treatment function. Accordingly, i~ i5 intended that the invention be limited only to the extent required by the appended claims.

Claims (36)

1. A cardiac stimulator for detecting and treating ventricular tachyarrhythmias of a patient's heart, characterized in that the stimulator includes means for selectively dividing the heart rate continuum into regions including at least two classes of tachycardia, contiguous to each other and of progres-sively higher heart rate ranges, the lowest and highest of the tachycardia classes being bounded respectively by a sinus rate region and a fibrillation region of the continuum; means for selectively adjusting the boundaries between the tachycardia classes and between the lowest and highest of those classes and the respective sinus rate and fibrillation regions, to corres-pondingly adjust the rate ranges of the classes in a desired manner; means for selectively detecting cardiac events anywhere within the continuum and for distinguishing between normal and abnormal tachycardias among the detected events; and means for selectively treating a detected abnormal tachycardia with any of a multiplicity of therapy regimens of differing degrees of aggressiveness, toward terminating the detected tachycardia.

2. A cardiac stimulator according to Claim 1, character-ized in that the treating means is responsive to acceleration of a detected tachycardia into the fibrillation region to automatic-ally apply defibrillation therapy to the patient's heart.

3. A cardiac stimulator according to claim 1, characterized in that the detecting means also senses a loss of electrical activity of the patient's heart, and in response the treating means applies stimulating pulses to pace the heart until it resumes its normal electrical activity.

4. A cardiac stimulator according to claim 1, characterized in that the treating means also selectively controls the sequence in which therapy regimens are applied in response to detection of an abnormal tachycardia within any of the classes.

5. A cardiac stimulator according to claim 1, characterized in that the detecting means also selectively redetects the presence or absence of an abnormal tachycardia in any of the classes after each application of a therapy regimen by the treating means, and in doing so applies less than the number of criteria used for the initial detection of the tachycardia, to expedite the redetection.

6. A cardiac stimulator according to claim 1, characterized in that the detecting means also senses a transition of a detected abnormal tachycardia to a higher or a lower class, and the treating means is responsive to the transition to apply a therapy regimen preselected for a tachycardia of the new class.

7. A cardiac stimulator according to claim 1, characterized in that the treating means treats any abnormal tachyarrhythmia detected within the continuum, and the therapies available from the treating means include pacing and shock therapies which differ at least in the level of electrical energy applied to the heart.

8. A cardiac stimulator according to claim 7, characterized in that each of the pacing and shock therapies is represented by a respective electrical waveform having at least one parameter which is selectively variable by the treating means.

9. A cardiac stimulator according to claim 1, characterized in that the detecting means also senses reversion of the patient's heart rate to sinus rhythm.

10. A cardiac stimulator according to claim 1, characterized in that the detecting means applies different criteria for detecting tachyarrhythmias in different regions of the heart rate continuum, the detection criteria being most stringent for the lowest tachycardia class and progressively less stringent with each higher rate range.

11. A cardiac stimulator according to claim 5, characterized in that the criteria used for initial detection are selectively modifiable for purposes of redetection.

12. A cardiac stimulator according to claim 6, characterized in that the detecting means also distinguishes false indication of acceleration from actual acceleration of the detected tachyarrhythmia to a higher rate range region of the continuum.

13. A cardiac stimulator according to claim 1, characterized in that the treating means is programmable to repetitively deliver the same therapy regimen in response to repeated redetection of a tachyarrhythmia within the same region of the heart rate continuum.

14. A cardiac stimulator according to claim 1, characterized in that the treating means is programmable to deliver a therapy regimen upon redetection of an abnormal tachyarrhythmia as having undergone a change of rate to a new region of the continuum, which differs from the therapy regimen that was delivered in response to the previous detection.

15. A cardiac stimulator according to claim 1, characterized in that the treating means is programmable to modify the aggressiveness of the selected therapy regimen during treatment of a detected abnormal tachyarrhythmia.

16. A cardiac stimulator according to claim 1, characterized in that the treating means is programmable to preclude return to a therapy less aggressive than that last delivered, upon acceleration of the abnormal tachycardia to a higher class.

17. A cardiac stimulator according to claim 1, characterized in that the treating means is programmable to reapply a therapy identical to that previously applied in response to detection of an abnormal tachyarrhythmia, upon the very next incidence of an abnormal tachyarrhythmia.

18. A cardiac stimulator according to claim 17, characterized in that the reapplied therapy is identical to the last therapy of the overall therapy regimen last successful to terminate an abnormal tachycardia.

19. A cardiac stimulator according to claim 17, characterized in that the reapplied therapy is identical to the entire last therapy regimen successful to terminate an abnormal tachycardia.

20. A cardiac stimulator according to claim 1, characterized in that the treating means is programmable to repeat the least aggressive therapies within a therapy regimen before redelivering the entire therapy regimen.

21. A cardiac stimulator according to claim 1, characterized in that the detecting means selectively establishes criteria to detect each of the cardiac events of bradycardia, tachycardia, fibrillation, and reversion to sinus rate.

22. A cardiac stimulator according to claim 21, characterized in that the criteria selectively established by the detecting means to detect tachycardias includes high heart rate, sudden onset of the high rate, stability of the high rate, and sustained high rate.

23. A cardiac stimulator according to claim 21, characterized in that the criteria selectively established by the detecting means to detect fibrillation includes high heart rate exceeding the highest rate for detecting tachycardias, and the occurrence of as least x fibrillation intervals out of y consecutive intervals, where x and y are selectible.

24. A cardiac stimulator according to claim 1, characterized in that the treating means is controllable to selectively alter any therapy consisting of either cardioverting shocks or defibrillating shocks by altering any of the features in the group consisting of (i) the number of shocks to be delivered, (ii) the delay time interval between detection of the cardiac event and delivery of the shock, (iii) the amplitude of the shocks on the first attempt and on each subsequent attempt to arrest the tachyarrhythmia, and (iv) the phasic structure of the shock waveform including pulse width, polarity, presence and absence of each phase.

25. A cardiac stimulator according to claim 1, characterized in that the treating means is controllable to selectively alter any pacing therapy by altering any of the features in the group consisting of (i) number of attempts to arrest the detected tachyarrhythmia, (ii) number of pulses in a pacing stimuli burst, (iii) delay time between detection and initial pulse in the pacing stimuli burst, (iv) interval between pulses in a pacing stimuli burst, (v) automatic decrementing or alternate incrementing and decrementing of either of the delay time or the interval between pulses, (vi) variation of each of those intervals as a percentage of the tachyarrhythmia rate, (vii) amount by which each of the intervals is increased or decreased for each successive burst, and (viii) number of times any sequence defined by (v), (vi) or (vii) is to be repeated.

26. A cardiac stimulator according to claim 1, characterized in that the stimulator includes a microprocessor for processing and storing digital data to develop control signals for operation of the heart rate continuum dividing means, the detecting means, and treating means; and the detecting means includes a sense amplifier to receive analog signal information representative of the patient's ECG and convert same to digital data for transmission to the microprocessor.

27. A cardiac stimulator according to claim 26, characterized in that the sense amplifier enhances the ECG
signal information to track the rapidly varying amplitude of signals indicative of fibrillation of the patient's heart.

28. A cardiac stimulator according to claim 27, characterized in that the sense amplifier has automatic gain control, and bandpass filtering of the ECG signal information.

29. A cardiac stimulator according to claim 28, characterized in that the bandpass filtering reduces the amplitude of signals outside the tachyarrhythmia frequency band in the absence of normal R-waves, indicative of fibrillation of the patient's heart.

30. A cardiac stimulator according to claim 28, characterized in that the automatic gain control is provided by a feedback circuit which includes the microprocessor, to adjust the gain of the sense amplifier to sense rapidly varying low amplitude signals indicative of ventricular fibrillation.

31. A cardiac stimulator according to claim 30, characterized in that the feedback circuit defines inner and outer voltage levels, selectively sets the ratio between the inner and outer voltage levels, compares the sense signal levels which have been processed by the amplification and filtering with the inner and outer voltage levels, responds to the processed sense signal levels exceeding the inner voltage levels by generating logic signals indicative of valid sense signals, and responds to the processed sense signal levels exceeding the outer voltage levels by applying further logic signals to the microprocessor to adjust the gain of the sense amplifier; and in that the microprocessor responds to the further logic signals and to components of the ECG signal information to decrease the amplifier gain in the presence of a QRS complex indicative of either sinus rate or a tachycardia, and to increase the amplifier gain in the absence of a T-wave within a predetermined time interval following application of a pacing stimulus to the patient's heart.

32. A cardiac stimulator according to claim 1, characterized in that the treating means also generates high voltage levels for developing cardioverting shocks and defibrillating shocks to be applied to the patient's heart.

33. A cardiac stimulator according to claim 32, characterized in that the high voltage levels are generated by the treating means using high voltage capacitors, a high voltage oscillator, and an isolation circuit coupling the high voltage oscillator to the capacitors to charge them to the high voltage levels suitable for the cardioverting shocks and defibrillating shocks.

34. A cardiac stimulator according to claim 33, characterized in that the treating means also includes an output circuit to deliver the high voltage shocks to the patient's heart, and switches selectively operated under the control of the microprocessor to apply the high voltages on the capacitors as pulses of desired amplitude and polarity to the output circuit.

35. A cardiac stimulator according to claim 34, characterized in that a short circuit protection circuit is provided to open the output circuit in response to current flow in the output circuit exceeding a predetermined level.

36. A cardiac stimulator according to claim 1, characterized in that the stimulator is battery powered with all components except electrodes and associated leads housed in a biocompatible case of sufficiently small size for implantation in the patient, the electrodes and associated leads being part of the detecting means and the treating means for respectively sensing electrical activity of the patient's heart and delivering the selected therapy regimens to the patient's heart.