Computerized mapping of fibrillation in normal ventricular myocardium

It is well known that the ability to fibrillate is intrinsic to a normal ventricle that exceeds a critical mass. The questions we address are how is ventricular fibrillation (VF) initiated and perpetuated in normal myocardium, and why is VF not seen more often in the general population if all ventricles have the ability to fibrillate. To study the mechanisms of VF, we used computerized mapping techniques with up to 512 channels of simultaneous multisite recordings for data acquisition. The data were then processed for dynamic display of the activation patterns and for mathematical analyses of the activation intervals. The results show that in normal ventricles, VF can be initiated by a single strong premature stimulus given during the vulnerable period of the cardiac cycle. The initial activations form a figure-eight pattern. Afterward, VF will perpetuate itself without any outside help. The self-perpetuation itself is due to at least two factors. One is that single wave fronts spontaneously break up into two or more wavelets. The second is that when two wavelets intersect perpendicular to each other, the second wavelet is broken by the residual refractoriness left over from the first wavelet. Mathematical analyses of the patterns of activation during VF revealed that VF is a form of chaos, and that transition from ventricular tachycardia (VT) to VF occurs via the quasiperiodic route. In separate experiments, we found that we can convert VF to VT by tissue size reduction. The physiological mechanism associated with the latter transition appears to be the reduction of the number of reentrant wave fronts and wandering wavelets. Based on these findings, we propose that the reentrant wave fronts and the wandering wavelets serve as the physiological equivalent of coupled oscillators. A minimal number of oscillators is needed for VF to perpetuate itself, and to generate chaotic dynamics; hence a critical mass is required to perpetuate VF. We conclude that VF in normal myocardium is a form of reentrant cardiac arrhythmia. A strong electrical stimulus initiates single or dual reentrant wave fronts that break up into multiple wavelets. Sometimes short-lived reentry is also generated during the course of VF. These organized reentrant and broken wavelets serve as coupled oscillators that perpetuate VF and maintain chaos. Although the ability to support these oscillators exists in a normal ventricle, the triggers required to generate them are nonexistent in the normal heart. Therefore, VF and sudden death do not happen to most people with normal ventricular myocardium. (c) 1998 American Institute of Physics.     

Dispersion of refractoriness and induction of reentry due to chaos synchronization in a model of cardiac tissue

Ventricular fibrillation is a lethal condition caused by multiple chaotically wandering electrical wavelets in the heart, reentering their own and each other's territories. The development of effective therapies requires a detailed understanding of how these reentrant waves are initiated. In this Letter, we demonstrate a novel mechanism for inducing reentry, in which chaos synchronization causes large-scale heterogeneities of refractoriness transverse to the direction of propagation. These regions of increased refractoriness create localized conduction block, which induces spiral wave reentry.     

Two forms of spiral-wave reentry in an ionic model of ischemic ventricular myocardium

It is well known that there is considerable spatial inhomogeneity in the electrical properties of heart muscle, and that the many interventions that increase this initial degree of inhomogeneity all make it easier to induce certain cardiac arrhythmias. We consider here the specific example of myocardial ischemia, which greatly increases the electrical heterogeneity of ventricular tissue, and often triggers life-threatening cardiac arrhythmias such as ventricular tachycardia and ventricular fibrillation. There is growing evidence that spiral-wave activity underlies these reentrant arrhythmias. We thus investigate whether spiral waves might be induced in a realistic model of inhomogeneous ventricular myocardium. We first modify the Luo and Rudy [Circ. Res. 68, 1501-1526 (1991)] ionic model of cardiac ventricular muscle so as to obtain maintained spiral-wave activity in a two-dimensional homogeneous sheet of ventricular muscle. Regional ischemia is simulated by raising the external potassium concentration ([K(+)](o)) from its nominal value of 5.4 mM in a subsection of the sheet, thus creating a localized inhomogeneity. Spiral-wave activity is induced using a pacing protocol in which the pacing frequency is gradually increased. When [K(+)](o) is sufficiently high in the abnormal area (e.g., 20 mM), there is complete block of propagation of the action potential into that area, resulting in a free end or wave break as the activation wave front encounters the abnormal area. As pacing continues, the free end of the activation wave front traveling in the normal area increasingly separates or detaches from the border between normal and abnormal tissue, eventually resulting in the formation of a maintained spiral wave, whose core lies entirely within an area of normal tissue lying outside of the abnormal area ("type I" spiral wave). At lower [K(+)](o) (e.g., 10.5 mM) in the abnormal area, there is no longer complete block of propagation into the abnormal area; instead, there is partial entrance block into the abnormal area, as well as exit block out of that area. In this case, a different kind of spiral wave (transient "type II" spiral wave) can be evoked, whose induction involves retrograde propagation of the action potential through the abnormal area. The number of turns made by the type II spiral wave depends on several factors, including the level of [K(+)](o) within the abnormal area and its physical size. If the pacing protocol is changed by adding two additional stimuli, a type I spiral wave is instead produced at [K(+)](o)=10.5 mM. When pacing is continued beyond this point, apparently aperiodic multiple spiral-wave activity is seen during pacing. We discuss the relevance of our results for arrythmogenesis in both the ischemic and nonischemic heart. (c) 1998 American Institute of Physics.     

Defibrillation via the elimination of spiral turbulence in a model for ventricular fibrillation

Ventricular fibrillation, the major reason behind sudden cardiac death, is turbulent cardiac electrical activity in which rapid, irregular disturbances in the spatiotemporal electrical activation of the heart make it incapable of any concerted pumping action. Methods of controlling ventricular fibrillation include electrical defibrillation as well as injected medication. Electrical defibrillation, though widely used, involves subjecting the whole heart to massive, and often counterproductive, electrical shocks. We propose a defibrillation method that uses a very low-amplitude shock (of order mV) applied for a brief duration (of order 100 ms) and over a coarse mesh of lines on our model ventricle.     

Wave front fragmentation due to ventricular geometry in a model of the rabbit heart

The role of the heart's complex shape in causing the fragmentation of activation wave fronts characteristic of ventricular fibrillation (VF) has not been well studied. We used a finite element model of cardiac propagation capable of simulating functional reentry on curved two-dimensional surfaces to test the hypothesis that uneven surface curvature can cause local propagation block leading to proliferation of reentrant wave fronts. We found that when reentry was induced on a flat sheet, it rotated in a repeatable meander pattern without breaking up. However, when a model of the rabbit ventricles was formed from the same medium, reentrant wave fronts followed complex, nonrepeating trajectories. Local propagation block often occurred when wave fronts propagated across regions where the Gaussian curvature of the surface changed rapidly. This type of block did not occur every time wave fronts crossed such a region; rather, it only occurred when the wave front was very close behind the previous wave in the cycle and was therefore propagating into relatively inexcitable tissue. Close wave front spacing resulted from nonstationary reentrant propagation. Thus, uneven surface curvature and nonstationary reentrant propagation worked in concert to produce wave front fragmentation and complex activation patterns. None of the factors previously thought to be necessary for local propagation block (e.g., heterogeneous refractory period, steep action potential duration restitution) were present. We conclude that the complex geometry of the heart may be an important determinant of VF activation patterns. (c) 2002 American Institute of Physics.     

Epicardial wavefronts arise from widely distributed transient sources during ventricular fibrillation in the isolated swine heart

It has been proposed that VF waves emanate from stable localized sources, often called "mother rotors." However, evidence for the existence of these rotors is conflicting. Using a new panoramic optical mapping system that can image nearly the entire ventricular epicardium, we recently excluded epicardial mother rotors as the drivers of Wiggers' stage II VF in the isolated swine heart. Furthermore, we were unable to find evidence that VF requires sustained intramural sources. The present study was designed to test the following hypotheses: 1. VF is driven by a specific region, and 2. Rotors that are long-lived, though not necessarily permanent, are the primary generators of VF wavefronts. Using panoramic optical mapping, we mapped VF wavefronts from 6 isolated swine hearts. Wavefronts were tracked to characterize their activation pathways and to locate their originating sources. We found that the wavefronts that participate in epicardial reentry were not confined to a compact region; rather they activated the entire epicardial surface. New wavefronts feeding into the epicardial activation pattern were generated over the majority of the epicardium and almost all of them were associated with rotors or repetitive breakthrough patterns that lasted for less than 2 s. These findings indicate that epicardial wavefronts in this model are generated by many transitory epicardial sources distributed over the entire surface of the heart.     

Localization of ischemia in canine hearts using tagged rotated long axis MR images,endocardial surface stretch and wall thickening

Tagged magnetic resonance imaging allows the noninvasive measurement of regional systolic myocardial deformations and helps localize ischemic regions in the left ventricle (LV). The objective of this study was to evaluate the potential accuracy of localizing ischemic regions in the LV using endocardial and epicardial data obtained from tagged rotated long axis images. Nine canine hearts with acute ischemia induced by coronary artery ligation were imaged along four long axis planes rotated around the LV long axis, at end diastole and end systole. Each plane was tagged by four parallel lines perpendicular to the LV long axis. Tracing the endocardial and epicardial intersection points of the tag lines, 24 myocardial cuboids were reconstructed for each LV at end diastole and end systole. Endocardial surface stretch and transmural systolic thickening were calculated for each cuboid. The functional data were compared to perfusion data obtained from postmortem monastral blue staining of the heart. The ability of each functional index to discriminate between ischemic and non-ischemic regions was assessed using the “t”-statistic. The potential accuracy in localizing ischemia was evaluated by studying the corresponding sensitivity-specificity curves. The results demonstrate that adequate discrimination and localization can be obtained with both functional indices. However, endocardial surface stretch is advantageous as it uses only endocardial data and can save 50% of the post-processing time.     

基于心脏腔式结构的心电图元胞自动机建模

建立了包含心房肌、心室肌、房室腔、室间隔并考虑心室肌分层结构的心电图元胞自动机模型.利用所建立的模型,仿真了电信号在心脏中的传导,计算了正常和缺血情况下的场点电势走势.数值结果表明:正常情况下,模拟所得的场点电势呈现与标准心电图一致的P波、QRS波群、T波和J波;在心内膜下肌细胞缺血情况下,出现T波倒置的现象;在心外膜下肌细胞缺血情况下,T波变得高耸;在透壁缺血情况下,T波提前形成,QT间期缩短.将正常和异常情况下的场点电势走势与临床结果进行了对比,并分析了其形成与持续机制.研究结果可为准确阐明心电图与心肌细胞电活动之间的关系、探讨心电图的产生与持续机制提供参考.     

Alternans and higher-order rhythms in an ionic model of a sheet of ischemic ventricular muscle

Life-threatening arrhythmias such as ventricular tachycardia and fibrillation often occur during acute myocardial ischemia. During the first few minutes following coronary occlusion, there is a gradual rise in the extracellular concentration of potassium ions ([K(+)](0)) within ischemic tissue. This elevation of [K(+)](0) is one of the main causes of the electrophysiological changes produced by ischemia, and has been implicated in inducing arrhythmias. We investigate an ionic model of a 3 cmx3 cm sheet of normal ventricular myocardium containing an ischemic zone, simulated by elevating [K(+)](0) within a centrally-placed 1 cmx1 cm area of the sheet. As [K(+)](0) is gradually raised within the ischemic zone from the normal value of 5.4 mM, conduction first slows within the ischemic zone and then, at higher [K(+)](0), an arc of block develops within that area. The area distal to the arc of block is activated in a delayed fashion by a retrogradely moving wavefront originating from the distal edge of the ischemic zone. With a further increase in [K(+)](0), the point eventually comes where a very small increase in [K(+)](0) (0.01 mM) results in the abrupt transition from a global period-1 rhythm to a global period-2 rhythm in the sheet. In the peripheral part of the ischemic zone and in the normal area surrounding it, there is an alternation of action potential duration, producing a 2:2 response. Within the core of the ischemic zone, there is an alternation between an action potential and a maintained small-amplitude response ( approximately 30 mV in height). With a further increase of [K(+)](0), the maintained small-amplitude response turns into a decrementing subthreshold response, so that there is 2:1 block in the central part of the ischemic zone. A still further increase of [K(+)](0) leads to a transition in the sheet from a global period-2 to a period-4 rhythm, and then to period-6 and period-8 rhythms, and finally to a complete block of propagation within the ischemic core. When the size of the sheet is increased to 4 cmx4 cm (with a 2 cmx2 cm ischemic area), one observes essentially the same sequence of rhythms, except that the period-6 rhythm is not seen. Very similar sequences of rhythms are seen as [K(+)](0) is increased in the central region (1 or 2 cm long) of a thin strand of tissue (3 or 4 cm long) in which propagation is essentially one-dimensional and in which retrograde propagation does not occur. While reentrant rhythms resembling tachycardia and fibrillation were not encountered in the above simulations, well-known precursors to such rhythms (e.g., delayed activation, arcs of block, two-component upstrokes, retrograde activation, nascent spiral tips, alternans) were seen. We outline how additional modifications to the ischemic model might result in the emergence of reentrant rhythms following alternans. (c) 2000 American Institute of Physics.     

Some observations on the question: Is ventricular fibrillation “chaos”?

Ventricular fibrillation has been modeled as cardiac chaos occuring after a series of subharmonic bifurcations. However, previous experimental studies have suggested that fibrillatory oscillations have a relatively narrow-band frequency spectrum inconsistent with a turbulent process. Similarly, during the first minute of canine fibrillation we observed only a few localized frequency peaks from the epicardial and body surface electrocardiogram rather than a broadband type of spectrum as would be predicted for chaotic dynamics. Further narrowing of the frequency spectrum occured during the second minute of fibrillation. The frequency spectrum of ventricular fibrillation contrasts with scaled, broadband spectra observed in normal cardiac function. We suggest that ventricular fibrillation may serve as a general model for transitions from broadband stability to certain types of pathological periodicities in other physiological perturbations.     

Hybrid segmentation of left ventricle in cardiac MRI using gaussian-mixture model and region restricted dynamic programming

Segmentation of the left ventricle from cardiac magnetic resonance images (MRI) is very important to quantitatively analyze global and regional cardiac function. The aim of this study is to develop a novel and robust algorithm which can improve the accuracy of automatic left ventricle segmentation on short-axis cardiac MRI. The database used in this study consists of three data sets obtained from the Sunnybrook Health Sciences Centre. Each data set contains 15 cases (4 ischemic heart failures, 4 non-ischemic heart failures, 4 left ventricle (LV) hypertrophies and 3 normal cases). Three key techniques are developed in this segmentation algorithm: (1) ray scanning approach is designed for segmentation of images with left ventricular outflow tract (LVOT), (2) a region restricted technique is employed for epicardial contour extraction, and (3) an edge map with non-maxima gradient suppression approach is put forward to improve the dynamic programming to derive the epicardial boundary. The validation experiments were performed on a pool of data sets of 45 cases. For both endo- and epi-cardial contours of our results, percentage of good contours is about 91%, the average perpendicular distance is about 2 mm. The overlapping dice metric is about 0.92. The regression and determination coefficient between the experts and our proposed method on the ejection fraction (EF) is 1.01 and 0.9375, respectively; they are 0.9 and 0.8245 for LV mass. The proposed segmentation method shows the better performance and is very promising in improving the accuracy of computer-aided diagnosis systems in cardiovascular diseases.     

Spatial organization,predictability, and determinism in ventricular fibrillation

The degree of spatial organization of ventricular fibrillation (VF) is a fundamental dynamical property of the arrhythmia and may determine the success of proposed therapeutic approaches. Spatial organization is closely related to the dimension of VF, and hence to its predictability and controllability. We have explored several techniques to quantify spatial organization during VF, to predict patterns of activity, and to see how spatial organization and predictability change as the arrhythmia progresses. Epicardial electrograms recorded from pig hearts using rectangular arrays of unipolar extracellular electrodes (1 mm spacing) were analyzed. The correlation length of VF, the number of Karhunen-Loeve modes required to approximate data during VF, the number, size and recurrence of wavefronts, and the mean square error of epicardial potential fields predicted 0.256 seconds into the future were all estimated. The ability of regularly-timed pacing stimuli to capture areas of fibrillating myocardium during VF was confirmed by a significant increase in local spatial organization. Results indicate that VF is neither "low-dimensional chaos" (dimension <5) nor "random" behavior (dimension= infinity ), but is a high-dimensional response with a degree of spatial coherence that changes as the arrhythmia progresses. (c) 1998 American Institute of Physics.     

Patterns of spiral tip motion in cardiac tissues

In support of the spiral wave theory of reentry, simulation studies and animal models have been utilized to show various patterns of spiral wave tip motion such as meandering and drifting. However, the demonstration of these or any other patterns in cardiac tissues have been limited. Whether such patterns of spiral tip motion are commonly observed in fibrillating cardiac tissues is unknown, and whether such patterns form the basis of ventricular tachycardia or fibrillation remain debatable. Using a computerized dynamic activation display, 108 episodes of atrial and ventricular tachycardia and fibrillation in isolated and intact canine cardiac tissues, as well as in vitro swine and myopathic human cardiac tissues, were analyzed for patterns of nonstationary, spiral wave tip motion. Among them, 46 episodes were from normal animal myocardium without pharmacological perturbations, 50 samples were from normal animal myocardium, either treated with drugs or had chemical ablation of the subendocardium, and 12 samples were from diseased human hearts. Among the total episodes, 11 of them had obvious nonstationary spiral tip motion with a life span of >2 cycles and with consecutive reentrant paths distinct from each other. Four patterns were observed: (1) meandering with an inward petal flower in 2; (2) meandering with outward petals in 5; (3) irregularly concentric in 3 (core moving about a common center); and (4) drift in 1 (linear core movement). The life span of a single nonstationary spiral wave lasted no more than 7 complete cycles with a mean of 4.6+/-4.3, and a median of 4.5 cycles in our samples. Conclusion: (1) Patently evident nonstationary spiral waves with long life spans were uncommon in our sample of mostly normal cardiac tissues, thus making a single meandering spiral wave an unlikely major mechanism of fibrillation in normal ventricular myocardium. (2) A tendency toward four patterns of nonstationary spiral tip motion was observed. (c) 1998 American Institute of Physics.     

Influence of cardiac tissue anisotropy on re-entrant activation in computational models of ventricular fibrillation

The aim of this study was to establish the role played by anisotropic diffusion in (i) the number of filaments and epicardial phase singularities that sustain ventricular fibrillation in the heart, (ii) the lifetimes of filaments and phase singularities, and (iii) the creation and annihilation dynamics of filaments and phase singularities. A simplified monodomain model of cardiac tissue was used, with membrane excitation described by a simplified 3-variable model. The model was configured so that a single re-entrant wave was unstable, and fragmented into multiple re-entrant waves. Re-entry was then initiated in tissue slabs with varying anisotropy ratio. The main findings of this computational study are: (i) anisotropy ratio influenced the number of filaments sustaining simulated ventricular fibrillation, with more filaments present in simulations with smaller values of transverse diffusion coefficient, (ii) each re-entrant filament was associated with around 0.9 phase singularities on the surface of the slab geometry, (iii) phase singularities were longer lived than filaments, and (iv) the creation and annihilation of filaments and phase singularities were linear functions of the number of filaments and phase singularities, and these relationships were independent of the anisotropy ratio. This study underscores the important role played by tissue anisotropy in cardiac ventricular fibrillation.     

Contraction and relaxation-induced oscillations of the left ventricle of the heart during the isovolumic phases

A theoretical analysis is presented for the transient dynamical response of the left ventricle of the heart during the isovolumic contraction and relaxation phases of the cardiac cycle. Small oscillations of the left ventricular cavity pressure and wall motion are excited by the initial rates of filling and emptying of the ventricle as well as the rate of change in muscle fiber activation. The analysis applies to the genesis of the first and second heart sounds. The ventricle is modeled as a finite, thick-walled incompressible cylinder having a continuum of imbedded axial and circumferential active muscle fibers, which interacts with a fixed volume of an incompressible, ideal fluid. The solution is obtained using a two-timing asymptotic expansion procedure. The theoretical calculations of left ventricular pressure waveforms compare favorably with published recorded pressure waveforms. The amplitude spectra of computed waveforms contain information concerning the active elastic modulus of the fibers, which is a measure of cardiac contractility.     

Ventricular fibrillation and atrial fibrillation are two different beasts

Although the mechanisms of fibrillation are no doubt multi-faceted, the geometry of the heart may play a major role in the dynamics of wave propagation during fibrillation [A. T. Winfree, Science 266, 1003-1006 (1994)]. The ventricles are thick chambers made up of sheets of parallel muscle fibers with the direction of fibers rotating across the ventricular walls (rotational anisotropy). The thick walls of the ventricles allow reentry to develop transmurally, provided the wavelength is sufficiently small. Depending on the kinetics of heart cells, the dynamics of rotating waves in three dimensions may be fundamentally different than in two dimensions, leading to destabilization of reentry and ventricular fibrillation (VF) in thick ventricles. The atria have an intricate geometry comprised of a thin sheet of cardiac tissue attached to a very complex network of pectinate muscles. The branching geometry of the pectinate muscles may lead to destabilization of two-dimensional reentry via "long-distance" electrical connections giving rise to atrial fibrillation (AF). Therefore, although fibrillation occurs via complex three-dimensional wave propagation in the ventricles and the atria, the underlying mechanisms and factors that sustain VF and AF are probably different.(c) 1998 American Institute of Physics.     

Self-organization and the dynamical nature of ventricular fibrillation

This article reviews recent data supporting the conjecture that, in the structurally and electrophysiologically normal heart, cardiac fibrillation is not a totally random phenomenon. Experimental and numerical studies based on the theory of excitable media suggest that fibrillation in the mammalian ventricles is the result of self-organized three-dimensional (3-D) electrical rotors giving rise to scroll waves that move continuously (i.e., drift) throughout the heart at varying speeds. A brief review of studies on the dynamics of rotors in two-dimensional (2-D) and 3-D excitable media is presented with emphasis on the experimental demonstration of such dynamics in cardiac muscle of various species. The discussion is centered on rotor dynamics in the presence and the absence of structural heterogeneities, and in the phenomena of drifting and anchoring, which in the electrocardiogram (ECG) may manifest as life-threatening cardiac rhythm disturbances. For instance, in the rabbit heart, a single electrical rotor that drifts rapidly throughout the ventricles gives rise to complex patterns of excitation. In the ECG such patterns are indistinguishable from ventricular fibrillation. On the other hand, a rotor that anchors to a discontinuity or defect in the muscle (e.g., a scar, a large artery or a bundle of connective tissue) may result in stationary rotating activity, which in the ECG is manifested as a form of so-called "monomorphic" ventricular tachycardia. More recent data show that ventricular fibrillation occurs in mammals irrespective of size or species. While in small hearts, such as those of mice and rabbits, a single drifting or meandering rotor can result in fibrillation, in larger hearts, such as the sheep and possibly the human, fibrillation occurs in the form of a relatively small number of coexisting but short-lived rotors. Overall, the work discussed here has paved the way for a better understanding of the mechanisms of fibrillation in the normal, as well as diseased human heart. (c) 1998 American Institute of Physics.     

Death,dynamics and disorder: Terminating reentry in excitable media by dynamically-induced inhomogeneities

Formation of feedback loops of excitation waves (reentrant circuit) around non-conducting ventricular scar tissue is a common cause of cardiac arrhythmias, such as ventricular tachycardia, often leading to death. This is typically treated by rapid stimulation from an implantable device (ICD). However, the mechanisms of reentry termination success and, more importantly, failure, are poorly understood. To study such mechanisms, we simulated pacing termination of reentry in a model of cardiac tissue having significant restitution and dispersion properties. Our results show that rapid pacing dynamically generates conduction inhomogeneities in the reentrant circuit, leading to successful pacing termination of tachycardia. The study suggests that more effective pacing algorithms can be designed by taking into account the role of such dynamical inhomogeneities.     

Vortex dynamics in three-dimensional continuous myocardium with fiber rotation: Filament instability and fibrillation

Wave propagation in ventricular muscle is rendered highly anisotropic by the intramural rotation of the fiber. This rotational anisotropy is especially important because it can produce a twist of electrical vortices, which measures the rate of rotation (in degree/mm) of activation wavefronts in successive planes perpendicular to a line of phase singularity, or filament. This twist can then significantly alter the dynamics of the filament. This paper explores this dynamics via numerical simulation. After a review of the literature, we present modeling tools that include: (i) a simplified ionic model with three membrane currents that approximates well the restitution properties and spiral wave behavior of more complex ionic models of cardiac action potential (Beeler-Reuter and others), and (ii) a semi-implicit algorithm for the fast solution of monodomain cable equations with rotational anisotropy. We then discuss selected results of a simulation study of vortex dynamics in a parallelepipedal slab of ventricular muscle of varying wall thickness (S) and fiber rotation rate (theta(z)). The main finding is that rotational anisotropy generates a sufficiently large twist to destabilize a single transmural filament and cause a transition to a wave turbulent state characterized by a high density of chaotically moving filaments. This instability is manifested by the propagation of localized disturbances along the filament and has no previously known analog in isotropic excitable media. These disturbances correspond to highly twisted and distorted regions of filament, or "twistons," that create vortex rings when colliding with the natural boundaries of the ventricle. Moreover, when sufficiently twisted, these rings expand and create additional filaments by further colliding with boundaries. This instability mechanism is distinct from the commonly invoked patchy failure or wave breakup that is not observed here during the initial instability. For modified Beeler-Reuter-like kinetics with stable reentry in two dimensions, decay into turbulence occurs in the left ventricle in about one second above a critical wall thickness in the range of 4-6 mm that matches experiment. However this decay is suppressed by uniformly decreasing excitability. Specific experiments to test these results, and a method to characterize the filament density during fibrillation are discussed. Results are contrasted with other mechanisms of fibrillation and future prospects are summarized. (c)1998 American Institute of Physics.     

Progress toward controlling in vivo fibrillating sheep atria using a nonlinear-dynamics-based closed-loop feedback method

We describe preliminary experiments on controlling in vivo atrial fibrillation using a closed-loop feedback protocol that measures the dynamics of the right atrium at a single spatial location and applies control perturbations at a single spatial location. This study allows investigation of control of cardiac dynamics in a preparation that is physiologically close to an in vivo human heart. The spatial-temporal response of the fibrillating sheep atrium is measured using a multi-channel electronic recording system to assess the control effectiveness. In an attempt to suppress fibrillation, we implement a scheme that paces occasionally the cardiac muscle with small shocks. When successful, the inter-activation time interval is the same and electrical stimuli are only applied when the controller senses that the dynamics are beginning to depart from the desired periodic rhythm. The shock timing is adjusted in real time using a control algorithm that attempts to synchronize the most recently measured inter-activation interval with the previous interval by inducing an activation at a time projected by the algorithm. The scheme is "single-sided" in that it can only shorten the inter-activation time but not lengthen it. Using probability distributions of the inter-activation time intervals, we find that the feedback protocol is not effective in regularizing the dynamics. One possible reason for the less-than-successful results is that the controller often attempts to stimulate the tissue while it is still in the refractory state and hence it does not induce an activation. (c) 2002 American Institute of Physics.