Fundamentals of Intracardiac Mapping

Underlying Basis for the Extracellular Electrogram

Cardiac electrical activity originates from activation of ion channels across cell membranes. The cardiac action potential is generated within individual cells and reflects cardiac electrical activation. Although some net charge flow occurs in the extracellular space, most cardiac electrical activity is generated within individual myocardial cells. Extracellular electrodes record potentials generated in the extracellular space and therefore differ markedly from action potentials recorded intracellularly. The differences are caused not only by differences in recording location (extracellular vs. intracellular) but also by the inherent summation of electrical activity from multiple cells that occurs when an extracellular potential is generated.

The field of view of extracellularly recorded electrograms reflects the relative contribution of individual cells both near to and far from recording electrodes that generate extracellular potentials. Computer modeling studies have created simulated extracellular potentials using various assumptions regarding intracellular action potentials. Factors that affect the field of view of recording electrodes are whether the recordings are unipolar or bipolar, interelectrode distance (for bipolar recordings), electrode size and composition, and inherent myocardial properties such as tissue resistivity and space constant. In view of the summation of intracellular potentials that occurs to generate extracellular potentials, some fundamental questions regarding cardiac mapping require further investigation. For example, when one asks the activation time determined from extracellular potentials, there may not be a single answer, because different cells within the field of view of an extracellular recording electrode may be activated at different times. Thus uniform rules regarding interpretation of extracellular electrograms need to account for differences in underlying physiology, and different rules may be appropriate in different circumstances. For example, it has generally been accepted that the HV interval should be measured from the onset of the His bundle electrogram. The basis for this approach is that regardless of the location of the recording electrode, one wishes to determine the onset of activation within the His bundle to best evaluate conduction time within the His–Purkinje system. By contrast, mapping of tachycardia origin uses techniques such as the baseline crossing in a bipolar electrogram that seeks to determine not the onset of activation but the occurrence of activation at a specific location, and in most cases to predict the effects of ablation at that site. Thus understanding the physiology that generates extracellularly recorded potentials and the purpose of a particular mapping technique is required to determine the best theoretical as well as practical techniques to use for intracardiac mapping.

Electrogram Recording: Amplification, Filtering, and Digitization of Signals

Physiologic signals acquired through intracardiac electrodes are typically less than 10 mV in amplitude and therefore require considerable amplification before they can be digitized, displayed, and stored. In most modern electrophysiology laboratories, the signal processor (filters and amplifiers), visualization screen, and recording apparatus are often incorporated as a computerized laboratory recording system. The amplifiers used for recording intracardiac electrograms must have the ability to gain modification as well as to alter both high-band–pass and low-band–pass filters to permit appropriate attenuation of the incoming signals.

After amplification, signals are digitized and filtered by a computerized data-acquisition system and are written to a hard disk or optical drive, while displaying signals in real time on a monitor. Digitization is a form of data reduction, whereby an analog waveform is sampled at a constant rate (sampling rate) by an analog-to-digital (A/D) converter. The amplitude of the analog waveform is translated into a binary number (e.g., 8-, 10-, or 12-bit conversion resolution of the A/D converter), which represents the full dynamic input range of the A/D converter (typically ±2.5 or ±10 V). Sampling rates of 600 Hz or more allow for the recording of most of the data contained in the intracardiac electrogram waveforms, with most modern mapping systems sampling at or about 1000 Hz. The ideal system should be able to provide a variety of display configurations with a wide range of sweep speeds (most modern systems can display up to 400 mm/second) and should allow adjustment of the size and gain and other characteristics of the amplified electrogram.

Filtering is an important aspect of electrogram processing ( Table 6.1 ). High-pass filters eliminate components below a given frequency (allowing signal content of higher frequency to pass without attenuation). In the surface electrocardiogram (ECG), components such as the T wave are of relatively low frequency, and high-pass filtering of 0.05 to 0.1 Hz is used to preserve these components while eliminating baseline drift. When examining bipolar intracardiac electrograms, high-frequency components are of the most interest, and high-pass filtering of 30 to 50 Hz is used to eliminate the low-frequency components. Unipolar electrograms usually go through a high-pass filter of 0.05 Hz, however, because the polarity of the signal (which reflects the direction of myocardial activation) and the signal morphology (and therefore low-frequency components) must be preserved. To eliminate noise at higher frequencies, low-pass filters are generally set to about 500 Hz for intracardiac signals (because there are essentially no intracardiac signals of interest much above 300 Hz). Notch filters remove specific frequencies such as 60-Hz noise from power supplies. Fig. 6.1 shows the effect of different filter settings on intracardiac atrial, ventricular, and His–Purkinje signals.

Unipolar and Bipolar Signals

The morphology and amplitude of the recorded electrograms depend on (1) the type of normal or abnormal depolarization responsible for the electrical potential and on local myocardial characteristics such as ischemia or infarction; (2) the orientation of the activation wave front in relation to myocardial fiber orientation ; (3) the distance between the source of the potential and the recording electrode; (4) the size, configuration, and interpolar distance of the recording electrode ; (5) the orientation of the wave front in relation to the poles of a bipolar electrode; (6) the conducting medium in which electrograms are recorded ; and (7) other factors.

The unipolar electrogram is recorded as the potential difference between a single electrode in direct contact with the heart (exploring electrode) and an indifferent electrode, which is placed at a distance from the heart (such as in the inferior vena cava) or at the Wilson central terminal. The recording is therefore not truly unipolar because all recordings depend on voltage differences between two poles; the unipolar designation signifies that one of the poles is distant from the heart. During cardiac activation, the approach of the dipole of an activation wave front toward an exploring electrode gives a small positive deflection, and its passage gives a rapid deflection in the negative direction, with a final return to baseline. The amplitude of the unipolar electrogram is proportional to the area of the dipole layer and the reciprocal value of the square of the distance between the dipole layer and the recording site. Thus the unipolar electrogram records a combination of local and distant electrical events, with the contribution of distant electrical events decreasing in proportion to the square of the distance from the exploring electrode. As mentioned earlier, extracellular recordings are not synonymous with intracellular microelectrode recordings. Nonetheless, in normal myocardium with relatively homogeneous conduction and repolarization, several studies (as well as theoretical models) have shown conformity of activation times between intracellular microelectrode and extracellular recordings ( Fig. 6.2 ), with the maximal downslope of the unipolar electrogram coinciding with the upstroke of the transmembrane potential. Therefore at least in normal hearts, there is agreement on using the maximal downslope of the unipolar electrogram for activation detection; there is significant controversy about the optimal value of the slope threshold, however, with recommended thresholds from different studies ranging from − 0.2 to − 2.5 mV per ms; the large range can be at least partially attributed to the fact that these studies were performed under a variety of different baseline conditions in normal, acutely ischemic as well as chronically infarcted hearts (animal and human).

The bipolar electrogram is recorded as the potential difference between two closely spaced electrodes in direct contact with the heart; it can be calculated as the difference between two unipolar electrograms at each of the two electrode sites ( Fig. 6.3 ). In the electrophysiology laboratory, this is typically done with analog amplifiers rather than with digital subtraction between unipolar signals. The amplitude of the bipolar electrogram is inversely proportional to the third power of the distance between recording site and dipole.

The major advantage of bipolar recordings lies in the distinction between local and distant activity. A limitation of bipolar electrograms is their directional sensitivity; as a result, if the activation wave front is parallel in relation to the electrode pair, the bipolar spike will be of maximal amplitude, whereas if it is perpendicular, both electrodes will record the same waveform at the same time, and no spike will result. In addition, activation at the two poles of the bipolar electrogram is not simultaneous, making activation detection more difficult in bipolar electrograms. The following criteria have been suggested for activation detection in bipolar electrograms: (1) the maximal absolute value of the bipolar electrogram; (2) the first elevation of the electrogram of more than 45 degrees from the baseline (obviously subject to display gain); (3) the baseline crossing with the steepest slope; and (4) morphologic algorithms that search for symmetry in the bipolar waveform. Of these, the maximal amplitude of the bipolar electrogram is the most easily measured and has been shown to coincide closely with the maximal downslope of the unipolar electrogram (and the maximal upstroke of the monophasic action potential).

Characteristics of Intracardiac Signals: Normal Heart

Unipolar endocardial electrograms obtained from experiments in normal canine hearts and from the human atrium and ventricle are characterized by a QS morphology with a rapid downstroke in the first part of the QRS complex (intrinsic deflection). Bipolar endocardial electrograms in the normal human left ventricle from catheters with 10-mm interelectrode distance have amplitudes greater than 3 mV and durations less than 70 ms, and no split, fractionated electrograms are found. Unipolar and bipolar electrograms in diseased myocardium are typically characterized by slower upstrokes and fragmentation (see later discussion).

Although slow conduction and fractionated, low amplitude signals are not as widespread in the normal heart compared with the diseased atrium or ventricle, low-amplitude, high-frequency electrograms have been well described at the thoracic vein–atrium junction, within both the pulmonary veins and the vena cava, as well as at other sites in the normal heart, such as the crista terminalis and coronary sinus.

Electrical Abnormalities in the Absence of Structural Heart Disease

Electrograms need not be entirely normal even in the absence of overt fibrosis or fatty replacement. For example, low-amplitude, fractionated bipolar electrograms may be seen at the earliest activation sites of focal tachycardias, even in the presence of a normal looking unipolar electrogram. Fig. 6.4 shows the bipolar electrogram at the successful ablation site of two different focal tachycardias in the absence of any overt structural heart disease: a focal tachycardia originating from the coronary sinus ostium and a focal ventricular tachycardia (VT) arising from the inferolateral left ventricle. Late potentials during sinus rhythm are also noted in hearts that are otherwise morphologically normal; Fig. 6.5 shows late potentials in the right ventricle of a patient who had exercise-induced left bundle branch block (LBBB) superior axis VT, but with no evidence of scar or any other abnormality on ECG or cardiac magnetic resonance imaging (MRI).

The basis for these potentials is not entirely clear. In the absence of overt scar on sensitive imaging modalities such as MRI, it is tempting to speculate that at least some of the delay seen in these electrograms may be functional in nature. In fact, functional anisotropic conduction is well described in the normal heart (e.g., at the crista terminalis) and may be responsible for the creation of macroreentrant circuits, such as isthmus-dependent atrial flutter. Anisotropic conduction has also been demonstrated within the pulmonary veins and the superior vena cava. Fig. 6.6 shows that the conduction times into and out of the pulmonary vein can be very different, which is an example of anisotropic conduction.

Conceivably, therefore at least some of the focal tachycardias described in the normal heart may be microreentrant in origin. Other factors that have been postulated are poor intercellular coupling, such as that caused by decreased gap junctions at sites of origin of focal tachycardias, leading to decreased electrotonic inhibition of the focus by surrounding tissue. In general, electrograms at successful ablation sites for focal tachycardias are less than 50 ms presystolic. However, pulmonary vein tachycardias are usually more than 50 seconds presystolic, at least partially on account of the slow conduction noted in this area.

Fractionated, bipolar electrograms (both during tachycardia and during sinus rhythm) have also been well described in idiopathic VT arising from the left ventricular septum. In fact, it is the presence of these Purkinje-like potentials some distance away from the septum that has led some to postulate a reentrant circuit of considerable size as the basis for this VT. Split electrograms may sometimes be recorded in the normal heart, such as in the posteroseptal right atrium from conduction block across the Eustachian ridge.