Ventricular reentry around a fixed barrier. Resetting with advancement in an in vitro model

MEMORY AND BISTABILITY IN A ONE-DIMENSIONAL LOOP OF MODEL CARDIAC CELLS

Unidirectional propagation has been studied in a one-dimensional loop of model cardiac cells represented as a homogeneous and isotropic cable. Membrane ionic currents were represented by a modified Beeler-Reuter model. The time constants of the gate variables of the slow inward current acting during the plateau of the action potential were divided by a parameter K ≥1. In the space-clamped model, increasing K shortens the action potential duration, changes the shape of the restitution curve and adds a slow memory component to the dynamics. In a paced regime, it promotes bistability in which period-1 and period-2 patterns coexist over an interval of pacing frequencies. In the loop, bistability is created between periodic and aperiodic modes of sustained reentry for an interval of loop length. The bistability of the space-clamped and loop model are both related to the form of the restitution curve.

The effect of cryoinjury on ventricular tachycardia in the swine right ventricle

This study was performed to assess the influence of the cryoinjury on the dynamics of wavefronts and to determine whether they can convert ventricular fibrillation (VF) to ventricular tachycardia (VT) in fibrillating right ventricular (RV) of swines using an optical mapping system. A cryoinjury with a diameter of 12 mm was created on the epicardium of perfused RV of swines (n = 6) and optical mapping were taken from baseline until 10 minutes after the cryoinjury. Out of 35 cryoinjuries, the images were possible to be interpreted in 32. The optical action potential could not be observed in either the cryoinjury or peri-injury sites at 1 and 3 minutes, was observed in only the cryoinjury site at 5 minutes, and recovered in both sites at 10 minutes. The cycle length of the tachycardia was 135.9 +/- 23.6 msec at baseline, 176.2 +/- 79.3 msec at 1 minute, 187.6 +/- 97.9 msec at 3 minutes, 185.5 +/- 19.2 msec at 5 minutes, and 152.1 +/- 64.1 msec at 10 minutes. The cycle lengths at 1, 3, and 5 minutes after the cryoinjury were significantly more prolonged than that at baseline (p = 0.001, p = 0.006, p = 0.016). After the cryoinjury, the VF changed to VT in 9 (28.0%), and terminated in 2 (6.3%). These changes were observed mainly within 5 minutes after cryoinjury. The cryoinjury had anti-fibrillatory effects on the tissue with VF. This phenomenon was related to a decreasing mass and stabilizing wavefronts.

Basic mechanisms of cardiac impulse propagation and associated arrhythmias

Propagation of excitation in the heart involves action potential (AP) generation by cardiac cells and its propagation in the multicellular tissue. AP conduction is the outcome of complex interactions between cellular electrical activity, electrical cell-to-cell communication, and the cardiac tissue structure. As shown in this review, strong interactions occur among these determinants of electrical impulse propagation. A special form of conduction that underlies many cardiac arrhythmias involves circulating excitation. In this situation, the curvature of the propagating excitation wavefront and the interaction of the wavefront with the repolarization tail of the preceding wave are additional important determinants of impulse propagation. This review attempts to synthesize results from computer simulations and experimental preparations to define mechanisms and biophysical principles that govern normal and abnormal conduction in the heart.

Reentrant waves in a ring of embryonic chick ventricular cells imaged with a Ca2+ sensitive dye

According to the classic model initially formulated by Mines, reentrant cardiac arrhythmias may be associated with waves circulating in a ring geometry. This study was designed to study the dynamics of reentry in a ring geometry of cardiac tissue culture. Reentrant calcium waves in rings of cultured embryonic chick cardiac myocytes were imaged using a macroscope to monitor the fluorescence of intracellular Calcium Green-1 dye. The rings displayed a variety of stable rhythms including pacemaker activity and spontaneous reentry. Waves originating from a localized pacemaker could lead to reentry as a consequence of unidirectional block. In addition, more complex patterns were observed due to the interactions between reentrant and pacemaker rhythms. These rhythms included instances in which pacemakers accelerated the reentrant rhythm, and instances in which the excitation was blocked in the vicinity of pacemakers. During reentrant activity an appropriately timed electrical stimulus could induce resetting of activity or cause complete annihilation of the propagating waves. This experimental preparation reveals many spontaneously occuring complex rhythms. These complex rhythms are hypothesized to reflect interactions between spontaneous pacemakers, wave propagation, refractory period, and overdrive suppression. This preparation may serve as a useful model system to further investigate complex dynamics arising during reentrant rhythms in cardiac tissue.

Resetting and annihilation of reentrant activity in a model of a one-dimensional loop of ventricular tissue

Resetting and annihilation of reentrant activity by a single stimulus pulse (S1) or a pair (S1-S2) of coupled pulses are studied in a model of one-dimensional loop of cardiac tissue using a Beeler-Reuter-type ionic model. Different modes of reentry termination are described. The classical mode of termination by unidirectional block, in which a stimulus produces only a retrograde front that collides with the activation front of the reentry, can be obtained for both S1 and S1-S2 applied over a small vulnerable window. We demonstrate that another scenario of termination-that we term collision block-can also be induced by the S1-S2 protocol. This scenario is obtained over a much wider range of S1-S2 coupling intervals than the one leading to a unidirectional block. In the collision block, S1 produces a retrograde front, colliding with the activation front of the pre-existing reentry, and an antegrade front propagating in the same direction as the initial reentry. Then, S2 also produces an antegrade and a retrograde front. However, the propagation of these fronts in the spatial profile of repolarization left by S1 leads to a termination of the reentrant activity. More complex behaviors also occur in which the antegrade fronts produced by S1 and S2 both persist for several turns, displaying a growing alternation in action potential duration ("alternans amplification") that may lead to the termination of the reentrant activity. The hypothesis that both collision block and alternans amplification depend on the interaction between the action potential duration restitution curve and the recovery curve of conduction velocity is supported by the fact that the dynamical behaviors were reproduced using an integro-delay equation based on these two properties. We thus describe two new mechanisms (collision block and alternans amplification) whereby electrical stimulation can terminate reentrant activity. (c) 2002 American Institute of Physics.

Predicting the entrainment of reentrant cardiac waves using phase resetting curves

Excitable media, such as the Belousov-Zhabotinsky medium or the heart, are capable of supporting excitation waves that circulate in a closed repetitive path--a phenomenon known as reentrant excitation. A single stimulus, depending on its magnitude, timing, and location, can cause a time shift of the reentrant excitation called resetting. The present study examines the ability of resetting data to predict the effects of periodic stimuli on reentrant excitation circulating on an annular domain. We compare the results of the theoretical models with experiments carried out in an animal model of a dangerous reentrant cardiac rhythm. The current work may lead to improved approaches to therapy through a better understanding of how typical clinical stimuli interact with abnormal reentrant cardiac rhythms.

Obstacle-induced transition from ventricular fibrillation to tachycardia in isolated swine right ventricles: insights into the transition dynamics and implications for the critical mass

OBJECTIVES

The study was done to test the hypothesis that an artificial anatomical obstacle prevents the maintenance of ventricular fibrillation (VF) by stabilizing reentrant wavefronts (RWF) and increases the critical mass (CM) of myocardium required to sustain VF.

BACKGROUND

Artificial obstacles can anchor RWF in simulated models of VF. Whether an artificial obstacle affects multiple-wavelet VF in real tissue is unclear.

METHODS

The endocardial surfaces of seven isolated, perfused swine right ventricles were mapped using a plaque of 477 bipolar electrodes with 1.6-mm resolution. An 8-mm hole was punched in the tissue. The CM was reached by tissue mass reductions, at which VF converted to periodic activity (ventricular tachycardia, VT).

RESULTS

After the creation of the obstacle, the VF cycle length increased from 71.6+/-18.4 ms to 87.5+/-13.0 ms (p<0.05). The obstacle, together with the papillary muscle, facilitated the transition from VF to VT by serving as attachment sites for the RWF. When one RWF attaches to the obstacle and another attaches to the papillary muscle, it may result in stable VT with figure-eight patterns. The CM for VF in the presence of an 8-mm hole (28.7+/-3.8 g) was higher than in the control group (swine right ventricles without holes, 24.0+/-3.4 g, p<0.05).

CONCLUSIONS

An artificial anatomical obstacle induces slowing and regularization of VF, impairs the persistence of VF as judged by an increase of the CM, and can convert VF to VT by serving as an attachment site to reentrant excitation.

Mechanism for site-dependent differences in the shape of the resetting response curve in fixed barrier reentry

INTRODUCTION

We investigated whether the site of stimulation within a reentrant circuit affects the resetting response curve (RRC). RRCs are used to characterize the excitable gap of reentrant circuits, including the duration of the fully excitable gap and the presence of partially excitable tissue.

METHODS AND RESULTS

We reset proximal and distal to a site of interval-dependent conduction (IDC) in canine in vitro atrial tricuspid rings. Adjustable reentry allowed changes in the cycle length and direction of reentry. In nine preparations we reset 26 tachycardias. In the 16 tachycardias with one site of IDC, RRCs were significantly different when stimulating distal and proximal to a site of interval-dependent conduction. For the distal curves, the duration of the flat portion was 42 +/- 26 msec greater (P < 0.001), the slope of the increasing portion was 0.20 +/- 0.17 less (P < 0.02), and the increase in the return cycle was 14 +/- 9 msec less (P < 0.001). These differences resulted from early activation of the site of IDC by the antidromic premature impulse when stimulating from distal sites. As a result, the coupling interval of the orthodromic impulse at the site of IDC was the same or greater than at the stimulation site. In 10 tachycardias with multiple sites of IDC, significant differences in the resetting responses did not occur even when the antidromic impulse penetrated one site of IDC.

CONCLUSION

In a fixed anatomic barrier reentrant circuit with one site of IDC, resetting distal to this site misrepresents the properties of the excitable gap of the entire circuit.

Influence of propafenone on resetting and termination of canine atrial flutter

Previous studies on atrial flutter (AF) presumed that resetting was due to the prematurity effect (PE) in which the stimulated antegrade wavefront travels in the tail of the AF preexisting wavefront. We studied the collision effect (CE) between the AF and the stimulated retrograde wavefronts, its contribution to resetting, and its relationship to AF termination and how they are affected by the Class IC agent propafenone (PPF). A canine model of AF was created using a Y-shaped lesion in the right atrium in 14 dogs (33 +/- 3 kg). Five atrial bipolar electrodes were positioned around the tricuspid valve. In a subsequent set of 11 dogs, we used 16 bipolar electrodes for recording. AF was induced by burst pacing. Single and multiple stimuli were applied to measure conduction time and reset-response curves (RRCs). This was repeated after the administration of PPF (1 mg/kg loading dose for 10 minutes, followed by 1.8 mg/kg/per hour infusion). Three distinct mechanisms were found to contribute to the RRC: the PE, the CE, and heterogeneity. PPF stabilized the RRC, increased significantly the cycle length (CL), the duration of the effective refractory period, as well as the duration of the excitable gap. However, PPF did not alter the duration of the fully excitable portion. We studied 36 annihilations without and 48 with PPF. Transient fibrillation was found in 75% of the episodes without, compared to 22% with PPF. Other types of termination such as conduction block, CL oscillations, and reversal of activation were found for 25% of the episodes without and 78% with PPF. In many cases, conduction block and CL oscillations were associated with a failure of propagation of the stimulated antegrade wavefront in the region of collision. Termination by reversal of activation suggests that propagation was two dimensional and could not be represented by a one dimensional movement. The average coupling interval (in percent of CL), that induced fibrillation was not significantly different from that at which conduction block occurred. This suggests that transient fibrillation is associated with a weak CE rather than with rapid pacing. The CE is amplified by multiple stimuli and PPF. The incidence of transient fibrillation in AF annihilation diminishes with PPF as the CE becomes more important. This suggests that the evaluation of PE and CE in AF may be an indication of the risk of atrial fibrillation.

Catheter ablation of mitral isthmus ventricular tachycardia using electroanatomically guided linear lesions

Mitral isthmus ventricular tachycardia uses a reentrant circuit with a critical isthmus of conduction bounded by the mitral valve proximally and a remote inferior infarction scar distally. Successful catheter ablation requires placement of a lesion to transect the isthmus so as to prevent wavefront propagation. We report a case with previously unsuccessful ablation in which focal isthmus ablation failed to eliminate arrhythmia. Electroanatomic mapping demonstrated a wide tachycardia isthmus, and a linear lesion placed from the edge of the inferior infarct (as demonstrated on the three-dimensional voltage electroanatomic map) to the base of the mitral valve successfully eliminated tachycardia. In some patients with mitral isthmus VT, a wide isthmus requires linear lesion placement to fully transect the isthmus and eliminate tachycardia. Electroanatomic mapping can be used to define isthmus boundaries and thus guide successful ablation.

Dispersion-based reentry: mechanism of initiation of ventricular tachycardia in isolated rabbit hearts

The aim of the study was to determine whether facilitation of reentry by potassium-channel openers is related to dispersion of refractoriness and/or modification of anisotropic properties of ventricular myocardium. The dispersion of ventricular effective refractory period (VERP), longitudinal and transverse ventricular conduction velocities (thetaL and thetaT, respectively), and wavelength [lambda = VERP x theta(L or T)] were studied in Langendorff-perfused left ventricular epicardium in 20 rabbits during infusion of incremental doses of levcromakalim or nicorandil. Dispersion of refractoriness was assessed using standard deviation of VERP mean (SD-VERP), dispersion index (DI; SD-VERP/mean VERP), and maximum dispersion (Dmax = VERPmax - VERPmin). Ventricular conduction velocities and anisotropic ratio were not modified, whatever the dose used. VERP and lambda were significantly shortened at high concentrations of levcromakalim and nicorandil. At these doses, SD-VERP, DI, and Dmax were increased significantly. Analysis of ventricular tachycardia induction, performed using a high-resolution ventricular mapping system, confirmed that heterogeneity and shortening of VERP were factors inducing functional conduction block. Our data suggest that, in rabbit left ventricular epicardium, functional conduction block facilitating the occurrence of reentry could be initiated by shortening and, especially, by dispersion of refractoriness during infusion of potassium-channel openers.

Reversal of reentry and acceleration due to double-wave reentry: two mechanisms for failure to terminate tachycardias by rapid pacing

OBJECTIVES

We sought to demonstrate mechanisms by which rapid pacing can cause conduction block without terminating reentry.

BACKGROUND

Rapid pacing can fail to terminate or can accelerate tachycardias in patients. Mechanisms for these responses are poorly understood.

METHODS

We studied reentry in the canine atrial tricuspid ring and a left ventricular ring in vitro in 12 preparations. Activations were recorded from 10 sites around the ring, and monophasic action potentials were recorded from critical sites of block. Rapid pacing at cycle lengths that intermittently caused conduction block was performed at multiple sites.

RESULTS

Action potential alternans contributed to block of an orthodromic impulse during rapid pacing. When pacing continued for two stimuli after orthodromic block, a second episode of block could reverse the direction of tachycardia. Continued pacing at this site was likely to produce block of an antidromic impulse, which may initiate double-wave reentry. Double-wave reentry could be sustained or nonsustained. Its cycle length was 56% to 77% of the single-wave cycle length. The ratio of double-wave cycle length to single-wave cycle length was inversely correlated with the relative excitable gap (p < 0.01). Double-wave reentry can be a mechanism for persistent cycle length alternation during tachycardia.

CONCLUSIONS

Successful termination of reentry by rapid pacing required block of an othrodromic impulse and stopping pacing within one stimulus after orthodromic block. Reversal of reentry makes the circuit resistant to termination from this site of pacing. Antidromic block can cause acceleration due to double-wave reentry when there is a substantial excitable gap.

The variable contribution of functional and anatomic barriers in human ventricular tachycardia: an analysis with resetting from two sites

OBJECTIVES

This study sought to investigate the influence of stimulation site on the properties of the circuit in ventricular tachycardia.

BACKGROUND

A fully excitable gap can be demonstrated in most human ventricular tachycardias. This requires the presence of an arc of block so that the entire circuit can recover from refractoriness within the period of the cycle length. Resetting characterizes the conduction properties of the tissue within the ventricular tachycardia circuit. Previous studies have not investigated the possibility of site-dependent differences in the resetting response.

METHODS

Resetting was performed from the right ventricular apex and outflow tract in 23 patients. Two characteristics of the resetting response were analyzed: 1) the total duration of the flat portion, and 2) the slope of the increasing portion.

RESULTS

A flat portion of the resetting response was observed in 18 tachycardias; in 8 of the 18, there was a significant site-dependent difference (> or = 40 ms) in the duration of the flat portion. A significant site-dependent difference in the slope of the increasing portion of the resetting curve was seen in 6 of 22 tachycardias. In all, a stimulation site-dependent change in at least one characteristic of the resetting response was seen in 12 (52%) of the 23 tachycardias.

CONCLUSIONS

A stimulation site-dependent change in the flat portion of the resetting response is compatible with an arc of block that is at least partially functional in nature. A change in the slope of the increasing portion is compatible with either partially functional circuit barriers or variation in properties of conduction and refractoriness at different locations within the circuit, or both. These observations suggest that a spectrum of circuit properties may exist in humans, with a variable contribution of anatomic and functional characteristics.

The dynamics of sustained reentry in a ring model of cardiac tissue

This paper describes the dynamics of circus movement around a fixed obstacle, using a one-dimensional continuous and uniform ring model of cardiac tissue to simulate sustained reentry. The membrane ionic current is simulated by a modified Beeler-Reuter formulation in which the kinetics of the fast sodium current were updated using more recent voltage-clamp data. Changes in the ring length are used to modify the dynamics of reentry. Reentry is stable if the ring length (X) exceeds a critical value (Xcrit) and complete block occurs if X is below a minimum (Xmin). Irregular sustained reentry is observed at intermediate ring lengths, as a narrow range of aperiodic reentry near Xcrit, and a larger range of quasi-periodic reentry at shorter ring lengths. The basic pattern of irregular reentry is an alternation between long and short cycle length, action potential duration (APD), diastolic interval (DIA), wavelength, and excitable gap. In aperiodic reentry cycle length variations are small, APD and DIA fluctuations are of medium amplitude, and conduction velocity over the whole pathway is essentially constant during successive turns. Much larger fluctuations in these various quantities occur during quasi-periodic reentry, and they increase in size as X approaches Xmin. The complexity of quasi-periodic reentry patterns is related to three factors: the slope of the APD versus DIA relation, which is greater than 1, the existence of a zone of slow conduction on the ring when the excitable gap becomes quite short, and the occurrence of triggered waves of secondary repolarization and excitability recovery. In the present model, quasi-periodic reentry with triggered secondary recovery covers most of the range of ring lengths, giving rise to sustained irregular reentry. There is very close agreement between our simulation results and experimental data obtained on rings of cardiac tissue.

Pleomorphic ventricular tachycardia with antegrade His-bundle activation: elucidation by multiple His-bundle recordings

The use of multiple His bundle-right bundle branch recordings in electrophysiologic studies has facilitated definition of the mechanism and elucidation of the direction of impulse propagation in bundle branch reentrant tachycardia, "Mahaim" fiber reciprocating tachycardia, and retrograde His depolarization in fascicular or ventricular tachycardias. This report details the electrophysiologic evaluation of pleomorphic ventricular tachycardia in a patient with advanced coronary heart disease. The ventricular tachycardia at baseline revealed variation in the QRS duration without alteration of the electrocardiographic (ECG) morphology. Following flecainide administration, a ventricular tachycardia with close resemblance of the ECG morphology to sinus rhythm was induced. Proximal and distal His-bundle recordings revealed early antegrade His-bundle activation during this tachycardia. Programmed stimulation converted this tachycardia back to the clinical ventricular tachycardia with intermittent narrowing of the QRS complexes. Early His activation was evident only during the narrower complexes but not in the tachycardia beats with wide complex. Penetration of the His bundle by ventricular tachycardia, with resultant fusion from intramyocardial ventricular activation and His-Purkinje activation, could have accounted for the near normalization of the QRS morphology during the ventricular tachycardia.

Abnormal automaticity as mechanism of atrial tachycardia in the human heart--electrophysiologic and histologic correlation: a case report

INTRODUCTION

A 32-year-old woman was operated upon because of drug refractory atrial tachycardia.

METHODS AND RESULTS

Electrophysiologic study was performed prior to operation. During surgery, epicardial mapping of the electrical activity of the left atrium was performed. The left atrial appendage was resected and studied in a tissue bath. Thereafter, histologic examination was performed. Polarity of the P wave in the surface ECG suggested that the tachycardia originated high in the left atrium. Epicardial mapping disclosed earliest activation in the apex of the left atrial appendage. Intracellular recordings from surgical specimen made at the site of origin, which was marked during surgery, revealed cells with phase 4 depolarization at cycle lengths ranging from 360 to 540 msec. Exit block prevented spread of activation from the spontaneously firing cells to surrounding tissue. Histology showed that spontaneous activity arose in an area with abnormal cells--characterized by an amorphous, pale eosinophilic staining cytoplasm and absence of nuclei--surrounded by normal myocytes.

CONCLUSION

The observations indicate that the mechanism of the atrial tachycardia was based on abnormal automaticity in an area consisting of a conglomeration of normal and abnormal myocytes.

Criteria for local myocardial electrical activation: effects of electrogram characteristics

Detection of local electrical myocardial activation by means of extracellular recordings is often difficult in the presence of polyphasic electrograms. The purpose of this investigation was to compare the ability of several variables to distinguish unipolar deflections due to local activation from those due to nonlocal activity. A model of polyphasic deflections based on atrial recordings during reentrant tachycardia was used to facilitate distinction of local and distant activity by methods independent of the test variables. The performance of variables were assessed by comparing areas under receiver operating characteristic curves. Optimal thresholds of test variables were identified by maximizing statistics which corrected for the pretest probability of local activation. We found that the greatest negative first derivative of the unipolar potential discriminated between local and distant ventricular signals, but performed less well than the ratio of the first derivative to the potential for distinguishing between local atrial signals and distant ventricular signals. A linear combination of the potential and the ratio of the first derivative and the potential performed well for all groups of signals studied. We conclude that optimal criteria for detecting local activation depends on the characteristics of the population of signals and that a statistical approach can be used to identify optimal criteria for a given population.

Frequency and implications of resetting and entrainment with right atrial stimulation in atrial flutter

Thirty-three patients (24 with typical and 9 with atypical flutter-wave morphology) were studied to evaluate the incidence and implications of resetting and entrainment of atrial flutter with right atrial stimulation. Resetting with single extrastimulus was present in 23 cases (group A) and absent in 10 (group B). Most cases of reset flutter were typical (20 of 23). Fixed fusion indicative of entrainment was observed in all 29 cases with pacing trains. Groups A and B did not differ significantly in flutter cycle length (230 +/- 20 vs 223 +/- 19 ms), atrial functional refractory period (165 +/- 18 vs 167 +/- 22 ms) or longest paced cycle length producing entrainment (213 +/- 19 vs 210 +/- 19 ms). In contrast, the return cycle after the longest paced cycle length producing entrainment was significantly shorter in group A (228 +/- 27 vs 284 +/- 56 ms; p = 0.001). The return cycle in group A was virtually identical to the flutter cycle length, whereas in group B it was greater (p = 0.002 compared with group A). Resetting was more frequent in typical than atypical flutter (20 of 24 vs 3 of 9; p = 0.01). Both typical and atypical flutter can be transiently entrained by right atrial pacing. Lack of resetting and longer return cycle, suggesting a longer conduction time between the reentrant circuit and the stimulation site, were mostly observed in atypical flutter. The data suggest a different location for both types of flutter, and may have implications for ablation techniques. A more cautious approach, with more extensive mapping, appears appropriate for ablation attempts of atypical flutter.

Unidirectional block between isolated rabbit ventricular cells coupled by a variable resistance

We have used pairs of electrically coupled cardiac cells to investigate the dependence of successful conduction of an action potential on three components of the conduction process: (a) the amount of depolarization required to be produced in the nonstimulated cell (the "sink" for current flow) to initiate an action potential in the nonstimulated cell, (b) the intercellular resistance as the path for intercellular current flow, and (c) the ability of the stimulated cell to maintain a high membrane potential to serve as the "source" of current during the conduction process. We present data from eight pairs of simultaneously recorded rabbit ventricular cells, with the two cells of each pair physically separated from each other. We used an electronic circuit to pass currents into and out of each cell such that these currents produced the effects of any desired level of intercellular resistance. The cells of equal size (as assessed by their current threshold and their input resistance for small depolarizations) show bidirectional failure of conduction at very high values of intercellular resistance which then converts to successful bidirectional conduction at lower values of intercellular resistance. For cell pairs with asymmetrical cell sizes, there is a large range of values of intercellular resistance over which unidirectional block occurs with conduction successful from the larger cell to the smaller cell but with conduction block from the smaller cell to the larger cell. We then further show that one important component which limits the conduction process is the large early repolarization which occurs in the stimulated cell during the process of conduction, a process that we term "source loading."