WO2000053089A1

WO2000053089A1 - Discrimination of rapid cardiac events

Info

Publication number: WO2000053089A1
Application number: PCT/US2000/006460
Authority: WO
Inventors: Alan F. Marcovecchio
Original assignee: Cardiac Pacemakers, Inc.
Priority date: 1999-03-12
Filing date: 2000-03-10
Publication date: 2000-09-14
Also published as: US20110034817A1; US8229552B2; US20020032469A1; US7991457B2; US6484055B1; US7953476B2; US6223078B1; US20040116972A1; US7039463B2; US20060122527A1; US6687540B2; AU3525600A; US20110288427A1

Abstract

A method and system for discrimination of supraventricular tachycardia and ventricular tachycardia events. Morphological features points are extracted from normal sinus rhythm (NSR) complexes and used to generate a NSR template. A numerical convolution is performed using the NSR template and the feature points for each sensed NSR to give a NSR filter output. Using a plurality of NSR complexes, a median NSR filter output template is determined, where the median NSR filter output template has a median value for each value in the NSR filter output. The median NSR filter output template is then used during a tachycardia event to distinguish tachycardia events as either ventricular tachycardia events or supraventricular tachycardia events.

Description

DISCRIMINATION OF RAPID CARDIAC EVENTS

Field of the Invention

This inv ention relates generally to medical devices, and more particularly to a system and method for discriminating supraventricular tachycardia from ventricular tachycardia during a tachycardia event

Background The heart is generally divided into four chambers, the left and right atπal chambers and the left and right ventricular chambers As the heart beats, the atπal chambers and the ventricular chambers go through a cardiac c> cle The cardiac cvcle consists of one complete sequence of contraction and relaxation of the chambers of the heart The terms systole and diastole are used to descπbe the contraction and relaxation phases the chambers of the heart experience duπng a cardiac cycle In systole, the ventπcular muscle cells contract to pump blood through the circulatory system. Dunng diastole, the ventπcular muscle cells relax, causing blood from the atπal chamber to fill the \ entncular chamber. After the period of diastohc filling, the systolic phase of a new cardiac cycle is initiated

Through the cardiac cycle, the heart pumps blood thiough the circulatory system Effective pumping of the heart depends upon fi\ e basic requirements First, the contractions of cardiac muscle must occur at regular interv als and be synchronized Second, the \ alves separating the chambers of the heart must fully open as blood passes through the chambers Third, the vah es must not leak Fourth, the contraction of the cardiac muscle must be forceful Fifth, the \entncles must fill adequately during diastole

When the contractions of the heart are not occurring at regular intervals or are unsynchronized the heart is said to be arrhythmic During an arrhythmia, the heart's ability to effectn ely and efficiently pump blood is compromised. Man different

of arrhythmias ha\ e been identified Arrhythmias can occur in either the atπal chambers or in the ventπcular chambers of the heart Ventπcular fibrillation is an arrhythmia that occurs in the \ entncle chambers of the heart In \ entncular fibπllation, v arious areas of the v entπcle are excited and contract asynchronous!) Duπng v entπcular fibrillation the heart fails to pump blood Since no blood is pumped duπng v entπcular fibrillation, the situation is fatal unless quickly corrected by cardiac conversion Ventπcular tachycardia is another arrhythmia that occurs in the ventricular chambers of the heart Ventricular tachycardia is a v ery seπous condition Ventπcular tachycardias are typified by ventπcular rates between 120-250 and are caused by disturbances (electπcal or mechanical) within the ventπcles of the heart Duπng a ventπcular tachycardia, the diastolic filling time is reduced and the ventπcular contractions are less synchronized and therefore less effective than normal Ventπcular tachycardias must be treated quickly in order to prevent the tachycardia from degrading into a life threatening ventricular fibπllation

Arrhythmias that occur in the atπal chambers of the heart are referred to generally as supraventricular tachycardias Supraventricular tachycardias include atπal tachycardias, atπal flutter and atπal fϊbπllation. Duπng certain supraventπcular tachycardias, aberrant cardiac signals from the atπa drive the ventπcles at a very rapid rate Such a situation occurs duπng paroxysmal atπal tachycardia This condition begins abruptly, lasts for a few minutes to a few hours, and then, just as abruptly, disappears and the heart rate reverts back to normal

Cardioverter-defϊbnllators, such as implantable cardioverter-defibπllators (ICDs), have been shown to be effectiv e in reducing the incidence of sudden cardiac death Sudden cardiac death is typically caused by either ventπcular tachycardia or ventπcular fibrillation Cardioverter-defibπllator systems operate by sensing and analyzing cardiac signals and applying electrical energy to the heart w hen either a v entπcular tachycardia or ventπcular fibrillation is detected One common way cardiov erter-defibπllators detect cardiac arrhythmias is to sense and analyze the rate of v entπcular contractions When the ventricular rate exceeds a programmed threshold v alue, the cardiov erter-defibπllator applies electrical energy in one or more specific patterns to treat either the v entπcular tachycardia or ventπcular fibrillation Rapid entricular rhythms, however, can occur in the presence of a supraventπcular tachycardia As previously mentioned, one example is during paroxysmal atπal tachycardia In this situation, treating the entπcles w ith electπcal energy is inappropriate as the treatment does not address the precipitating factor of the rapid v entπcular rate Therefore, a need exists for reliably assessing and determining the origin of a rapid v entπcular rate. By reliably discriminating the origin of the rapid v entπcular rate, more appropriate and effective therapies can be applied to treat the heart

Summary of the Invention The present subject matter discloses a method and a system for discriminating, or classifying supraventπcular tachycardias (SVT) from malignant ventricular tachycardias (VT) during a tachycardia ev ent In one embodiment, the present subject matter is implemented in an implantable cardioverter defibrillator By using the method of the present subject matter, the implantable defibrillator assesses and determines the oπgm of a rapid ventricular rate, allowing the implantable defibπllator to reduce the number of inappropπate therapies dehv ered to the heart.

In one embodiment, QRS-complexes are sensed duπng normal sinus rhythm (NSR) A plurality of feature points are located on the sensed NSR QRS-complexes based on morphological features of the individual NSR QRS- complexes The plurality of feature points from the NSR QRS-complexes are then used to determine a NSR template. In one embodiment, a plurality of NSR QRS-complexes are used to determine the NSR template

In one embodiment, the NSR template includes a median value for each of the plurality of feature points taken along the NSR QRS-complex. A numerical conv olution is then preformed on the values of the NSR template. A numeπcal conv olution is also preformed on the plurality of feature points for each of the plurality of the NSR QRS-complexes This process gives a NSR filter output for each of the NSR QRS-complexes Using the NSR filter output for each NSR QRS-complex. a median NSR filter output template is determined In one embodiment, the median NSR filter output template includes the median values of the NSR filter output v alues for each NSR QRS-complex. Once the median NSR filter output template has been determined, the system senses for the occurrence of a tachycardia event. When a tachycardia event is detected, the system senses the tachycardia complexes. In one embodiment, QRS-complexes are extracted, or sampled, from the tachycardia complexes in the sensed signals. The plurality of feature points are then located in the QRS-complexes The feature points located in the QRS-complexes duπng the tachycardia event are based on morphological features of the QRS-complex. In one embodiment, the morphological features taken from the QRS-complex duπng the tachycardia episode are from the same relative position as the morphological features taken along the NSR QRS-complex.

A tachycardia complex output is then determined by performing a numerical convolution of the median NSR filter output template with the plurality of feature points from a QRS-complex of a tachycardia complex sensed duπng the tachycardia event. The differences between the v alues of the tachycardia complex output and the median NSR filter output template are summed to give a sum of residual value. The sum of residual (SOR) value is then compared to a sum of residual (SOR) threshold value, and when the SOR value is greater than or equal to the SOR threshold value the tachycardia complex is classified as a ventπcular tachycardia complex When the SOR value is less than the SOR threshold value the tachycardia complex is classified as a supraventricular tachycardia complex. When the number of tachycardia complexes classified as either ventricular tachycardia complexes or supraventπcular tachycardia complexes exceeds a predetermined threshold value, the tachycardia event is classified as the tachycardia event that exceeded the threshold value ( i.e., as either a ventπcular tachycardia or a supraventπcular tachycardia).

These and other features and advantages of the invention w ill become apparent from the following descπption of the preferred embodiments of the invention. Brief Description of the Drawings

Figure 1 is a schematic of an implantable medical device: Figure 2 is a flow chart illustrating one embodiment of the present subject matter;

Figure 3 is a flow chart illustrating one embodiment of the present subject matter: Figure 4 shows one embodiment of an electrocardiogram of a normal sinus rhythm complex;

Figure 5 is a flow chart illustrating one embodiment of the present subject matter;

Figures 6A, 6B and 6C show plots of sensed cardiac complexes as a function of time for three cardiac conditions;

Figures 6D, 6E and 6F show plots of SOR values as a function of sensed cardiac complex for the sensed cardiac complexes of Figures 6A, 6B and 6C; and

Figure 7 is one embodiment of a block diagram of an implantable medical device according to the present subject matter. Detailed Description

In the following detailed description, reference is made to the accompanying drawings which form a part hereof and in which is shown by way of illustration specific embodiments in which the invention can be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice and use the invention, and it is to be understood that other embodiments may be utilized and that electrical, logical, and structural changes may be made without departing from the spirit and scope of the present invention. The following detailed descπption is. therefore, not to be taken in a limiting sense and the scope of the present invention is defined by the appended claims and their equivalents.

The embodiments illustrated herein are demonstrated in an implantable cardiac defibrillator (ICD), which may include numerous defibnllation, pacing, and pulse generating modes known in the art. However, these embodiments are illustrative of some of the applications of the present system, and are not intended in an exhaustive or exclusive sense. The concepts descπbed herein can be used in a vaπety of applications which will be readily appreciated by those skilled in the art upon reading and understanding this descπption For example. the present system is suitable for implementation in a vaπetv ot implantable and external dev ices

As discussed abov e ICDs can deliv er inappropriate therapy to the v entπcles of the heart One example is vv hen a suprav entπcular tachvcardia causes a rapid entricular rate The ICD senses the rapid v entricular rate and responds by treating the heart (/ e , dehveπng electπcal energy to the ventπcles) as if a ventπcular tachycardia were occurring Howev er, when the cause of the rapid ventπcular rate is a supraventπcular tachycardia treating the ventπcles is inappropπate as the therapy is not directed at the source of the arrhythmia Looking at it another way, the ICD is attempting to treat the symptoms, not the cause of the rapid entπcular rate

The present subject matter allows for the oπgin, or the source, of a rapid ventricular rate to be distinguished In one embodiment, the present subject matter distinguishes the origin of a ventπcular tachycardia as either the result of a supraventricular tachycardia (/ e , an arrhythmia in the atria is dπving the ventπcles at a rapid rate) or the result of a disturbance in the ventπcles (- e , a ventricular tachycardia)

Rapid ventπcular rates can include ventπcular rates that are between 100- 250 beats per minute, as are typical duπng ventπcular tachycardia Ventπcular rates of greater than 250 beats per minute are also considered to be within the present subject matter

The present subject matter utilizes sensed cardiac complexes in determining, or distinguishing, the source of rapid v entπcular rates Cardiac complexes include the electrical excitatory impulses, or action potentials, that are sensed from the heart is it goes through the cardiac cycle In one embodiment, the sensed cardiac complexes are electrocardiogram signals sensed from the beating heart Major features of an electrocardiogram signal include the P-wave, the QRS-complex and the T-wave w hich are caused by the atπal depolaπzation, the v entπcular depolarization and the v entπcular repolaπzation. respectiv ely In one embodiment, cardiac complexes are sensed and analyzed using an

ICD system In one embodiment, the ICD system uses a transvenous lead system to allow sensing of the cardiac action potentials The transv enous lead system can include a rate-sensing electrode and at least one defibnllation electrode positioned on the transvenous lead Cardiac action potentials sensed using defibnllation electrodes are typically referred to as far-field signals (or morphology signals) and cardiac action potentials sensed using rate sensing, or pacing, electrodes are typically referred to as near-field signals (or rate signals) In one embodiment, the rale-sensing electrode is a pacing tip electrode positioned at the distal end of the transvenous lead system Other types of rate sensing electrodes are also considered appropriate to use with the present subject matter Examples of other types of rate sensing electrodes include ring electrodes, both annular and semi-annular, as are known in the art Rate sensing using the transv enous lead system can also be accomplished either through unipolar or bipolar sensing methods, as are known

In one embodiment, the trans enous lead system can have a single defibnllation electrode When a single defibnllation electrode is present, the ICD uses unipolar sensing and defibnllation, as is known In one embodiment, a transvenous catheter with a single defibnllation electrode is implanted with the rate sensing electrode and the defibnllation electrode positioned within the nght ventncular chamber of the heart In an alternative embodiment, the transvenous lead can have two, or more, defibnllation electrodes When two defibnllation electrodes are present, the ICD system can pieform bipolar sensing the cardiac action potentials In bipolar sensing, cardiac action potentials are sensed between the tw o defibnllation electrodes, and defibnllation electπcal shocks are delivered bet een the t o defibnllation electrodes The ICD system of the present subject matter can also use endocardial and patch electrodes as are known

In one embodiment, the ICD employs an single body lead catheter sold under the trademark ENDOTAK (Cardiac Pacemaker. Inc ' Guidant Corporation, St Paul, MN) having a pacing tip electrode and two defibnllation coil electrodes One example of such a system is shown in Figure 1 ICD 100 is coupled to catheter 1 10, hich is implanted to receiv e signals from heart 120 The catheter 1 10 also may be used for transmission of pacing and or defibnllation signals to the heart 120 In an alternativ e embodiment, a three defibnllation electrode system is employed, wherein the housing of the implantable sy stem is used as a third defibnllation electrode

In one example, the ICD 100 senses cardiac signals from the heart 120 When the ICD 100 detects the occurrence of an arrhythmic ev ent, the ICD 100 analyzes the sensed arrhythmic complexes (/ e , the cardiac signals) of the arrhythmic event Dunng the analysis, the ICD compares the sensed arrhythmic complexes to cardiac signals sensed and recorded dunng the patient's normal sinus rhythm Based on the companson, the ICD 100 is able to distinguish SVT events from VT events and. depending upon the ICD's classification of the arrhythmic event, to provide appropriate therapy to treat the heart 120

Figure 2 shows another embodiment of a system for distinguishing SVT events from VT events during a tachycardia ev ent At 200, the system senses cardiac signals representative of electπcal cardiac activity At 210, the system analyzes the sensed cardiac signals to detect the occurrence of a tachycardia event in the heart In one embodiment, the tachycardia event are ventncular arrhythmic events.

When tachycardia event is detected at 210, the system continues to sense the tachycardia complexes At 220, the sensed tachycardia complexes are then compared to a template in order to determine the classification of the sensed tachycardia complex In one embodiment, the template is determined from cardiac complexes sensed during normal sinus rhythm. For example, features from the cardiac complexes sensed during normal sinus rhythm are used to create a normal sinus rhythm template Features from the tachycardia complexes sensed during the tachycardia ev ent are then compared to the normal sinus rhythm template In one embodiment, the features from the sensed tachycardia complexes correspond to the features taken from the cardiac complexes sensed during normal sinus rhythm In other words, features taken from the tachycardia complexes are from the same relativ e location as those take from the normal sinus rhythm cardiac complexes In one embodiment, the companson of the sensed tachycardiac complexes and the normal sinus rhythm template at 220 determines how similar or dissimilar the sensed tachycardiac complexes are from the patient's normal sinus rhythm complexes. Based on the comparison, the system at 230 then classifies the sensed tachycardiac complexes as either ventricular tachycardiac signals or supraventricular tachycardia signals. As the tachycardiac complexes are classified, the system counts the occurrence and the classification of each tachycardiac complex. At 240, the system then determines whether the number of sensed ventricular tachycardia complexes has exceeded a predetermined threshold value. If the threshold has been exceeded, the system proceeds to 250 where a ventricular tachycardia event is declared and therapy to treat the ventricular tachycardia is delivered to the patient's heart. If the threshold has not been exceeded, the system returns to 210, where additional cardiac complexes are sensed and classified by the system.

Figure 3 shows an embodiment for determining a normal sinus rhythm (NSR) template. At 300, the system senses cardiac complexes during normal sinus rhythm (NSR). In one embodiment, the NSR complexes are sensed and recorded using the ICD under the supervision of the patient's attending physician. In one embodiment, the system samples, or senses, QRS-complexes of the NSR complexes.

As each QRS-complex is sensed it is isolated, or windowed, for analysis. At 310, a plurality of feature points are located on the QRS-complex. In one embodiment, the plurality of feature points are morphological features of the QRS-complex. For example, morphological features extracted from the QRS- complexes include the amplitude values of peaks and valleys (or maxima and minima) acquired by a process called feature extraction. During feature extraction, each NSR complex is isolated according to a known morphological template. In one embodiment, the morphological template operates to detect the activation of an heart beat (such as the occurrence of an R-wave), at which point the electronic control circuitry of the implantable medical device analyzes the complex associated with the signal indicating the activation of the heart beat. In one embodiment, a threshold value or a detection criterion is used to indicate the activation of the heart beat. Once a heart beat has been detected, the feature extraction derives the plurality of feature points from the morphological features of the NSR complex. The extracted feature points for each NSR complex create a vector In one embodiment, the vector includes a set of numbers, where each number is associated with a particular morphological point along the sensed NSR complex Other types of features know n in the art can also be extracted and used in dev eloping the NSR template. In one embodiment, tw o or more morphological features are extracted and used from the NSR complexes in developing the NSR template. For example, four features can be extracted from each NSR complex in developing the NSR template. In one embodiment, the features having a numeπcal value which are characteπstic of the morphological position along the cardiac complex So, at 310, as the NSR complexes are sensed, a set of four features are extracted, and a four feature NSR complex is created and stored for each of the NSR complexes Figure 4 shows one embodiment of a NSR complex 400. The NSR complex 400 is processed to determine the amplitudes of peaks 402 and valleys 404 in the QRS-complex 406. In one embodiment, the peaks 402 and valleys 404 are determined by determining major inflection points in the QRS- complex as represented in Figure 4. The resulting values of the peaks 402 and valleys 404 provides a four dimensional NSR complex vector, [1 A, IB, IC, ID], where the number "1 " represents the number of the sensed NSR complex. As previously mentioned, other features from cardiac complexes known in the art can be used in developing vectors By way of example, other features can include the start or ending of a cardiac complex as detected by a predetermined deviation from a baseline signal or by detecting a predetermined decrease or increase in the slope of the cardiac signal. Thus, the present subject matter is in no way limited to vectors developed from maximum and minimum deflection points along signals of cardiac complexes.

Referring again to Figure 3, at 320, a NSR template is determined from a plurality of the NSR complex vectors. In one embodiment, the NSR template is a median value computed from the corresponding NSR complex v ectors. For example, the NSR template [TA, TB, TC, TD] is determined from a series of NSR complex vectors [ 1 A. IB, I C, I D], [2A. 2B, 2C, 2D], [3 A, 3B, 3C, 3D] . ., [iiA, nB, nC. nD], where TA equals the median value of ( 1 A, 2 A. 3 A, . nA). TB equals the median value of ( I B, 2B, 3B. nB). TC equals the median value of (I C, 2C, 3C. nC). and TD equals the median v alue ot ( I D, 2D, 3D, n-D) In an alternativ e embodiment, the NSR template feature v alue is an average value computed from the corresponding NSR complex vectors The number of NSR complexes used in determining TA, TB, TC, TD is a programmable value In one embodiment, the number of NSR complexes used in determining the NSR template is a programmable value in the range of 10 to 100 NSR complexes At 330, the NSR template is used to create the matched filter impulse response (h(t)) In one embodiment, the matched filter impulse response is defined to be h(t) = [TD, TC. TB. TA]

The h(t) is "matched" to the NSR template A matched filter is a specific type of filter designed to maximize the output signal-to-noise ratio The matched filter effectively correlates an input signal with a stored replica of a signal of interest The impulse response of the matched filter is the signal of interest time reversed and possibly time shifted When a signal of interest is detected by the matched filter, the output signal-to-noise ratio will be maximized.

In one embodiment, the h(t) is used to filter features from tachycardia complexes sensed duπng a tachycardia event In an additional embodiment, the system at 340 performs a numerical convolution of the NSR template and the NSR complexes to giv e a NSR filter output The system at 350 then determines a NSR filter output template from the NSR filter output In one embodiment, features are extracted from normal sinus rhythm complexes to create the NSR filter output template In one embodiment, the features extracted from the NSR complexes create a four element v ector as previously discussed The extracted features are taken from the same relative position along the tachycardia complexes as the features taken from the NSR complexes After extracting the features from the sensed tachycardia complexes, the feature v alues are stored for processing or filtering In one embodiment, as tachycardia complexes are sensed duπng a tachycardia e ent, the system extracts the values of the four features along the tachycardia complexes In one embodiment, the v alues of the four features are referred to as v ectors Λs the v ector is determined for each sensed tachycardia complex, the system filters the complex with respect to the four features of the normal sinus rhythm template The result is a sev en (7) element filter output for each of the sensed cardiac complexes In one embodiment, the v ector for each of the NSR cardiac complexes are convolv ed ith h(t) In one embodiment, this procedure produces the seven (7) element filter output for each of the sensed cardiac complexes The seven element filter output values are then used to determine the NSR filter output template In one embodiment, the NSR filter output template is a median value computed from a plurality of normal sinus rhythm complexes, where the median NSR filter output template has a median v alue for each v alue in the NSR filter output For example, the median NSR filter output template [TE, TF, TG, TH, TI, TJ, TK] is determined from a plurality of NSR complex v ectors [ I E. I F, 1G, 1H. I I, U, IK], [2E, 2F, 2G, 2H, 21, 2J, 2K], [3E, 3F, 3G, 3H, 31, 3J, 3K] , [nF, nG, nH, nl. nJ, nK], where TE equals the median value of ( IE, 2E, 3E, nE), TF equals the median value of ( IF, 2F, 3F, nF), TG equals the median value of (1G, 2G, 3G, nG), TH equals the median value of (1H, 2H, 3H, n-H), TI equals the median v alue of ( I I, 21, 31, nl), TJ equals the median value of ( U, 2J, 3J, nJ), and TK equals the median value of ( IK, 2K, 3K, nK) In an alternative embodiment, the NSR filter output template is an a erage value computed from the corresponding NSR complex ectors

In one embodiment, the method for determining a normal sinus rhythm (NSR) template is preformed w ith an implantable cardioverter defibrillator The ICD system discussed herein is one example of an appropriate system for determining a patient's NSR template In an alternativ e embodiment, other implantable medical devices, such as implantable defibπllators, or external defibπllators can be used to determine the NSR template and implement the present subject matter

Figure 5 show s an embodiment of a system for distinguishing SVT events from VT ev ents dunng a tachycardia event At 500, cardiac signals representativ e of electrical cardiac activ lty are sensed In one embodiment, the cardiac signals are sensed bv an endocardial lead system of an ICD as previouslv discussed The cardiac signals include cardiac complexes which are portions of the complete cardiac cycles In one embodiment, the sensed cardiac complexes include the QRS-complex of the cardiac cycle The system analyzes the sensed cardiac complexes to determine if a tachycardia event is occurring In one embodiment, the system determines the occurrence of a tachycardiac event by analyzing the sensed cardiac rate A cardiac rate that exceeds a predetermined threshold indicates the occurrence of a ventncular tachycardia In one embodiment, the predetermined threshold is for cardiac rates of between 150-250 beats per minute In an alternative embodiment, the predetermined threshold is a lower rate zone in an ICD that is configured with multiple rate-zones Other methods of determining the occurrence of tachycardia event which are known in the art may be used without departing from the present system

When a tachycardia event is detected, the system proceeds to 504. At 504, the system samples the tachycardia complexes during the tachycardia event Feature points along the tachycardia complexes are located on the QRS-complex of the tachycardia complexes at 508. In one embodiment, the feature points located along the sensed tachycardia complexes are at the same relative position as feature points located along QRS-complexes sensed dunng normal sinus rhythm For example, the QRS-complex is isolated, or windowed, for analysis and vector values denved for the complex as previously discussed.

In one embodiment, the set of features extracted from the QRS-complex of the cardiac complex are processed using the NSR template In one embodiment, the set of features extracted from the QRS-complex are convolved with the matched filter impulse response (h(t)) of the NSR template at 512 to give a tachycardia complex output. In one embodiment, the h(t) used at 512 is denved from features extracted from a patient's normal sinus rhythm complexes

In processing the tachycardiac complexes with h(t), the system extracts features from the morphology of the tachycardiac complexes In one embodiment, the number of features extracted from the morphology signal of the tachycardiac complexes is equal to the number of features extracted from the morphology signals of the sensed normal sinus rhythm complexes So, for example, four features are extracted from the morphology of the sensed normal sinus rhythm complexes and so four features are extracted from each of the sensed tachycardiac complexes The extracted features are taken from the same relative position along the sensed tachycardiac complexes as those taken from the complexes sensed duπng normal sinus ihythm After extracting the features from the sensed tachycardiac complexes, the feature values are stored for processing, or filtering (including the use of a matched filtering), w ith the normal sinus rhythm template at 512

In one embodiment, as the tachycardiac complexes are sensed during the tachycardia ev ent, the system measures values for each of the features along the tachycardiac complex In one embodiment, the values of the features are referred to as vectors As the vector values are determined for each sensed tachycardiac complex, the system filters the signal with respect to the features of the normal sinus rhythm template (or the matched filter impulse response) to give the tachycardia complex output For example, when four features are sensed from the normal sinus rhythm complexes and the tachycardiac complexes, the system filters the tachycardiac complexes with respect to the four features of the normal sinus rhythm template (or matched filter impulse response) When four features are used, the result of the filtering process is a seven (7) element filter output for each of the sensed tachycardiac complexes In this embodiment, the sev en element filter output is the tachycardia complex output for the sensed cardiac complex

In one embodiment, the filtering process is similar to a mathematical convolution of the normal sinus rhythm template with the features extracted from the sensed cardiac signals In one embodiment, the process of numerical convolution can be en lsioned as the interaction bet een two strips In one embodiment, the numencal convolution of y(t) = f(t) * h(t) can be described where the sequence of f[l ], f[2], f[3], , [samples for f(t)] are vvπtten on the upper stπp, and the sequence h[l ], h[2], h[3], . [samples for h(t)], are vvπtten on the lower strip The upper stip is fixed, and the lo er strip is folded (inverted) about t=l and then mov ed from left to right, one slot at a time, w ith the f[l] and h[l] slots coinciding at t=l To compute y(t), the lo er stnp is shifted bv t slots and multiply the sample v alues of f(t) by the samples alues ot h(t) lying in the adjacent slots The product of the multiplied samples are then summed together This procedure is then repeated for each time the lower strip is shifted along the upper stπp For example, in filtering the four features of the cardiac complex for the present system, the result is a seven member v ector (the tachycardia complex output for the sensed cardiac complex), where the seven member vector is determined from the calculation of

[ h(l )f(l), h(l)f(2) - h(2)f(l ), h(l )f(3) + h(2)f(2) + h(3)f(l ), h(l )f(4) T h(2)f(3) - h(3)f(2) -t- h(4)f( l ), h(2)f(4) + h(3)f(3) + h(4)f(2), h(3)f(4) - h(4)f(3), h(4)f(4)]

At 516, the system then sums the difference between the tachycardia complex output and the median NSR filter output template In one embodiment, the resulting difference is given the term sum of residuals (SOR) value In one embodiment, a SOR value is calculated for each of the sensed tachycardiac complexes One way to think of the SOR is as a differentiation value, where the SOR is used as a companson of magnitudes at specific points along morphological signals In one embodiment, the SOR is calculated from the seven-element output vectors from the sensed tachycardiac complexes and the normal smus rhythm output template

As the SOR can be thought of as the sum of differences between the tachycardia complex output and the median NSR filter output template, there are many ways in which the differences can be calculated For example, the SOR is determined by taking the absolute alue of the differences of corresponding elements from a sev en-element output vector and the single, sev en-element NSR output template These quantities are then added together to produce a scalar quantity for each complex This computation can also be descnbed mathematically as

7

Sum of Residuals = ∑ N, " 1 , _>

Where N = seven-element NSR output template T = seven-element tachycardia output v ector In an additional embodiment, the SOR v alue can also be calculated by determining the sum of the square of the differences between .V- and -T- Alternatively , the SOR value can be calculated by summing the difference between N and T_l Other mathematical methods of quantifying differences between .V and T_t exist and are considered within the scope of the present invention.

Referπng now to Figures 6A, 6B and 6C there are shown sensed cardiac complexes plotted as a function of time. Figure 6A shows cardiac complexes sensed duπng normal sinus rhythm; Figure 6B shows cardiac complexes sensed duπng ventricular tachycardia; and Figure 6B shows cardiac complexes sensed during atπal fibrillation.

Figures 6D, 6E and 6F show SOR values plotted as a function of sensed cardiac complexes. Figure 6D shows SOR values calculated for the cardiac complexes plotted in Figure 6A; Figure 6E shows SOR values calculated for the cardiac complexes plotted in Figure 6B; and Figure 6F shows SOR values calculated for the cardiac complexes plotted in Figure 6C.

Comparing the calculated SOR values for each of Figures 6D, 6E and 6F reveal differences in SOR v alues for the cardiac complexes sensed dunng the ventricular tachycardia event and the atrial fibrillation event as compared to cardiac complexes sensed during normal sinus rhythm. Additionally, differences in SOR values can be seen for the cardiac complexes sensed during the ventncular tachycardia event as compared to those of the atrial fibrillation event. Referring again to Figure 5, after the sum of the differences between the tachycardia complex output and the median ΝSR filter output template is computed at 516, the system compares the numerical difference to a threshold value at 524. In one embodiment, the SOR value for each tachycardiac complexes is compared to a SOR threshold value. In one embodiment, when the numencal difference is less than the threshold value, the cardiac complex is classified as a supraventπcular tachycardia (SVT) complex at 528. In one embodiment, when the numencal difference is greater than, or equal to. the threshold value, the cardiac complex is classified as a ventricular tachycardia (VT) complex at 532. As the sensed tachycardiac complexes are categoπzed, the svstem records the number of v entπcular tachvcardia complexes and SVT complexes that hav e been categorized during the tachv caidia ev ent at 536 In one embodiment, the predetermined threshold is an x out of the last y signals counter In one embodiment, the v alues for x and y are programmable In one embodiment x has programmable integer v alues of greater than 3, where the alues can be in the range of 3 to 10, where, in an additional embodiment, 5 is an acceptable value In one embodiment, y has a programmable integer v alue of greater than 8, where the v alues can be in the range of 8 to 30, where 10 is an acceptable v alue In an alternative embodiment, the system determines a percentage of VT complexes during the tachvcardia event When the percentage of the VT complexes exceeds a predetermined percentage threshold, the system declares the occurrence of a ventricular tachycardia In one embodiment, the predetermined percentage threshold is a programmable v alue in the range of 30 to 100 percent, w here 50 percent is an acceptable v alue

When the number of VT complexes exceeds the predetermined threshold, a VT event is declared A signal is then provided to the system (e g , an implantable medical device) to deliver v entπcular tachycardia therapy at 540 to a heart when a ventπcular tachycardia episode is declared In an alternative embodiment, if the number of SVT complexes exceeds the predetermined threshold, an SVT ev ent is declared In one embodiment, w hen an SVT ev ent is declared, a signal is provided to the system (e g an implantable medical dev ice) to deliver supraventπcular tachycardia therapy to the heart In an alternativ e embodiment, therapy is not delivered to the suprav entπcular region of the heart, but rather the system continues to monitor the cardiac condition and provides treatment only when a ventπcular tachycardia is determined If at 536 the number of classified cardiac complexes does not exceed the predetermined threshold, the svstem returns to 508 to sense and classify the next tachv cardiac complex Figure 7 shows one embodiment of an implantable cardiac defibπllator

(ICD) 700 which mav include numerous defibnllation, pacing, and pulse generating modes know n in the art A-n endocardial lead is phvsicallv and electπcally coupled to the ICD 700 The endocardial lead can include at least one pacing electrode and at least one defibnllation coil electrode as are known In one embodiment, the endocardial lead is an ENDOTAK lead as previously descnbed Figure 7 discloses ICD 700 which includes input circuitry 710 In one embodiment, input circuitry 710 includes a first amp 712 and a second amp 714 The first amp 712 receives rate-signals or near-field signals through the at least one pacing electrode In one embodiment, the rate-signals are sensed using a unipolar configuration, where the cardiac signals are sensed between the at least one pacing electrode and the housing 716 of the ICD 700 Alternatively, bipolar sensing is accomplished between two or more pacing electrodes on one or more endocardial leads The second amp 714 receiv es morphology-signals, or far- field signals, from at least two defibnllation coil electrodes located on the endocardial lead An R-wave detector 720 receives the rate-signals from the first amp 712

The R-wave detector 720 detects R-waves from the rate-signals being received by the first amp 712 and conveys information relating to the cardiac rate to a microprocessor 724 by a data bus 726 A morphology analyzer 730 receives morphology signals from the second amp 714 In one embodiment, the morphology analyzer ⁷30 extracts a plurality of feature points from sensed cardiac complexes Λ template generating circuit 734 is coupled to the signal morphology analyzing circuit 730 by the bus 726 The template generating circuit 734 receives the extracted plurality of feature points In one embodiment, the template generating circuit 734 generates a normal sinus rhythm template from sensed normal sinus rhythm complexes

A filter output response circuit 740 is coLipled to the template generating circuit 734 by bus 726 The filter output response circuit 740 creates a normal sinus rhythm filter output by performing a numencal convolution on the NSR template and the plurality of feature points for each of the plurality of the NSR complexes The filter output response circuit 740 also determines a median normal sinus rhythm output template from the plurality of normal sinus rhythm complexes, where the median NSR filter output template has a median v alue for each value in the NSR filter output

In one embodiment, when, dunng a tachycardia ev ent, the input circuitry 710 receive a QRS-wav e signal from a tachycardia complex, the signal morphology analyzing circuit 730 locates the plurality of feature points on the QRS-complex based on morphological features of the QRS-complex The filter output response circuit 740 then performs a numencal convolution of the NSR template with the plurality of feature points on the QRS-complex to give a tachycardia complex output The filter output response circuit 740 then sums a numerical difference between the values of the tachycardia complex output and the median NSR filter output template

The summed difference is then received by the microprocessor 724 where the calculated value is compared to the predetermined sum of residual threshold value Dunng a tachycardia episode, as the system senses each cardiac complex the ICD 700 classifies each sensed cardiac complex as either being a ventncular tachycardia complex or a supraventricular complex The ICD 700 then determines w hether the number of ventricular tachycardia complexes exceeded a predetermined threshold v alue In one embodiment, when the number of ventπcular tachycardia complexes exceeded a predetermined threshold v alue, the ICD 700 declares a ventncular tachycardia event When a ventncular tachycardiac event is declared, the microprocessor 724 provides a signal to a cardio ersion/defibnllation output circuit 744 to deliver ventπcular tachycardia therapy to a heart

In an alternative embodiment, the ICD 700 determines w hether the number of suprav entricular tachycardia complexes exceeded the predetermined threshold value In one embodiment, when the number of supraventricular tachycardia complexes exceeded a predetermined threshold value, the ICD 700 declares a suprav entncular tachycardia event When a supraventricular tachycardiac ev ent is declared, the microprocessor 724 provides a signal to a cardiov ersion defibnllation output circuit 744 to deliver suprav entncular tachycardia therapy to a heart Power to operate the ICD 700 is supplied by a battery 748. Memory 750 is also provided in the ICD 700, and is connected w ith the microprocessor 724. The ICD 700 further includes a transmitterreceiver 754, which can be used to communicate w ith the microprocessor 724 through a programmer 760 as is known

The embodiments provided herein are intended to demonstrate only some of the embodiments of the present system. Other embodiments utilizing the present subject matter are can be appreciated by those skilled in the art. For example, the concepts of the present subject matter are expressly described in terms of cardiac complexes sensed for the QRS-wave of the heart, however, applications to other cardiac complexes, including P-wave complexes or a combination of QRS-wave and P-wave complexes, can be readily appreciated by those skilled in the art without departing from the present invention.

Also, a dual chamber implantable cardiac defibrillator can be used to distinguish SVT events from SVT events based on sensed cardiac signals. In one embodiment, the dual chamber implantable cardiac defibrillator includes an ENDOTAK single body lead catheter implanted in the ventπcular region of the heart and an atπal catheter implanted in a supraventπcular region of the heart. This embodiment allows for ventricular near-field signals and ventricular far- field signals, along with atrial near-field signals to be sensed and analyzed by the implantable cardiac defibrillator

Other cardiac defibrillator systems and catheter configurations may also be used without departing from the present system. In addition to ICD systems, the present system may be utilized in external defibnllation systems and in external cardiac monitoring systems. In addition to employing endocardial leads, the present system can also utilize body surface leads.

Additionally, ev en though ventπcular tachycardia events were discussed herein, other arrhythmic events can also be analyzed to determine the nature, or origin, of the cardiac arrhythmia using the teachings provided herein, and therefore, the express teachings of this disclosure are not intended in an exclusive or limiting sense.

Claims

What is claimed is

1 A system, compnsing a signal moφhology analyzing circuit coupled to input circuitry where the signal moφhology analyzing circuitry extracts a plurality of feature points from a sensed complex, and a filter output response circuit coupled to the signal moφhology analyzing circuit, where the filter output response circuit performs a numerical convolution of a normal template with the plurality of feature points to give a complex output and sums a numerical difference between the values of the complex output and a median normal filter output template

2. The system of claim 1 , w here the signal moφhology analyzing circuit extracts a plurality of feature points from normal complexes, and where the system includes a template generating circuit coupled to the signal moφhology analyzing circuit, where the template generating circuit generates a normal filter output from a numencal convolution of the normal template and the plurality of feature points from the normal complexes, and determines the median normal filter output template from the normal complexes, where the median normal filter output template has a median alue for each value in the normal filter output

3 The system of claim 2, here the template generating circuit determines the normal template from the normal complexes

The system of claim 3, where the template generating circuit determines a median value for each of the plurality of feature points

The system of claim 2, where the input circuitry receives a complex, the signal moφhology analyzing circuit locates the plurality of feature points on the complex based on moφhological features of the complex, the filter output response circuit performs a numencal convolution of the normal template ith the plurality of feature points on the complex to giv e a complex output and sums a numerical difference between the v alues of the complex output and the median normal filter output template

The system of claim 1 , including a microprocessor which receives the sum of the numerical difference from the filter output response circuit and classifies the complex as a rapid complex when the sum of the difference is greater than or equal to a sum of residual threshold value

The system of claim 6, where the microprocessor declares a rapid rhythm w hen the number of rapid complexes exceeded a predetermined threshold value

The system of claim 7, where the microprocessor provides a signal to an output circuit to generate therapy when the rapid rhythm is declared

The system of claim 6, where the microprocessor classifies the complex as a supra-rapid complex w hen the sum of the difference is less than the sum of residual threshold v alue

The system of claim 9, where the microprocessor declares a supra-rapid rhythm when the number of supra-rapid complexes exceeded a predetermined threshold value

The system of claim 10, where the microprocessor provides a signal to an output circuit to generate supra-rapid therapy hen a supra- rapid rhythm is declared

A method, compnsing sampling a complex,

-n locating a plurality of feature points on the complex; performing a numerical convolution of a normal template with the plurality of feature points on the complex to give a complex output; and summing a numerical difference between the values of the complex output and a median normal filter output template.

13. The method of claim 12, including: locating a plurality of feature points on normal complexes; determining the normal template from the normal complexes; performing a numerical convolution of the normal template and the plurality of feature points for each of the normal complexes to give a normal filter output; and determining the median normal filter output template from the normal complexes, where the median normal filter output template has a median value for each value in the normal filter output.

14. The method of claim 13, where determining the normal template includes determining a median value for each of the plurality of feature points for the normal template.

15. The method of claim 12, including classifying the complex as a rapid complex when the sum of the difference between the values of the complex output and the median normal filter output template is greater than or equal to a sum of residual threshold value.

16. The method of claim 15. including classifying the complex as a supra-rapid complex when the absolute value of the difference between the values of the complex output and the median normal filter output template is less than the sum of residual threshold value.

17. The method of claim 16. including sampling a plurality of complexes, classifying each of the plurality of complexes as either a rapid complex or a supra-rapid complex, and determining whether the number of rapid complexes exceeded a predetermined threshold value.

18. The method of claim 17. including declaring rapid event when the number of rapid complexes exceeded the predetermined threshold value.

19. The method of claim 18, including providing a signal to an output circuit to generate therapy when a rapid event is declared.

20. The method of claim 17, including declaring supra-rapid event when the number of supra-rapid complexes exceeded the predetermined threshold value.

21. The method of claim 20, including providing a signal to an output circuit to generate therapy when a supra-rapid event is declared.

22. The method of claim 12, where summing a numerical difference includes summing an absolute value of the difference between the values of the complex output and the median normal filter output template.