The Online Encyclopedia and Dictionary






Ventricular fibrillation

Ventricular fibrillation (V-fib) is a cardiac condition which consists of a lack of coordination of the contraction of the muscle tissue of the large chambers of the heart. The ventricular muscle twitches randomly, rather than contracting in unison, and so the ventricles fail to pump blood into the arteries.

Ventricular fibrillation is a medical emergency: if the arrhythmia continues for more than a few seconds, the blood pressure will collapse, blood circulation will cease, and death will ensue. Ventricular fibrillation, along with an arrhythmia called pulseless ventricular tachycardia are the two most common causes of cardiac arrest.

This condition can sometimes be reversed with the use of a defibrillator and the usage of antiarrhythmic agents like amiodarone or lidocaine.



Sudden cardiac arrest is the leading cause of death in the industrialised world. It exacts a significant mortality with approximately 70-90,000 sudden cardiac deaths each year in the United Kingdom and survival rates are only 2% (National Institute for Clinical Excellence Guidelines 2000). The majority of these deaths are due to ventricular fibrillation secondary to ischaemic heart disease [Myerburg RJ et al. 1995]. During ventricular fibrillation cardiac output drops to nil and unless terminated promptly death usually ensues within minutes. Ventricular fibrillation has been described as "chaotic asynchronous fractionated activity of the heart" [Moe et al. 1964]. Another more complete definition is that ventricular fibrillation is a "turbulent, disorganised electrical activity of the heart in such a way that the recorded electrocardiographic deflections continuously change in shape, magnitude and direction" [Robles de Medina 1978]. Ventricular fibrillation most commonly occurs within diseased hearts, and in the vast majority, it is a manifestation of underlying ischaemic heart disease. Ventricular fibrillation is also seen in those with cardiomyopathy, myocarditis and other heart pathologies. It is also notable that ventricular fibrillation occurs where there is no discernible heart pathology, the so-called idiopathic ventricular fibrillation. Idiopathic ventricular fibrillation occurs with a reputed incidence of approximately 1% of all cases of out-of-hospital arrest, as well as 3-9% of the cases of ventricular fibrillation unrelated to myocardial infarction and 14% of all ventricular fibrillation resuscitations in patients under the age of 40 [Viskin S et al 1990]. It follows then that on the basis of the fact that ventricular fibrillation itself is common, then idiopathic ventricular fibrillation accounts for an appreciable mortality. Recently described syndromes such as the Brugada Syndrome may give clues to the underlying mechanism of ventricular arrhythmias. In the Brugada syndrome, changes may be found in the resting ECG with evidence of RBBB and ST elevation in the V1-V3 chest leads with an underlying propensity to sudden cardiac death [Brugada P et al. 1992]. The relevance of this is that theories of the underlying pathophysiology and electrophysiology must account for the occurrence of fibrillation in the apparently "healthy" heart. There are obviously mechanisms at work, which we do not fully appreciate and understand. Investigators are exploring new techniques of detecting and understanding the underlying mechanisms of sudden cardiac death in these patients without pathological evidence of underlying heart disease [Saumarez RC et al 1995].

Historical Aspects

Lyman Brewer suggests that the first recorded account of ventricular fibrillation dates as far back as 1500 BC and can be found in the Ebers Papyrus of Ancient Egypt. The extract recorded 3500 years ago may even date from as far back as 3500 BC. It states - "When the heart is diseased, its work is imperfectly performed: the vessels proceeding from the heart become inactive, so that you cannot feel them … if the heart trembles, has little power and sinks, the disease is advanced and death is near". Whether this is a description of ventricular fibrillation is debatable [Brewer LA 1983]. The next recorded description occurs 3000 years later and is recorded by Vesalius who described the appearance of "worm like" movements of the heart in animals prior to death. The significance and clinical importance of these observations and descriptions possibly of ventricular fibrillation were not recognised until John Erichsen in 1842 described ventricular fibrillation following the ligation of a coronary artery [Erichsen JE 1842]. Subsequent to this in 1850, fibrillation was described by Ludwig and Hoffa when they demonstrated the provocation of Ventricular fibrillation in an animal by applying a "faradic" current to the heart [Hoffa M et al. 1850]. In 1874, Edmé Félix Alfred Vulpian coined the term "mouvement fibrillaire", a term, which he seems to have used to describe both atrial and ventricular fibrillation [Vulpian A 1874]. John A. MacWilliam, a physiologist who had trained under Ludwig and who subsequently became Professor of Physiology at Aberdeen University gave an accurate description of the arrhythmia in 1887. This definition still holds today and is interesting in the fact that his studies and description predate the use of electrocardiography. His description is as follows - "The ventricular muscle is thrown into a state of irregular arrhythmic (sic) contraction, whilst there is a great fall in the arterial blood pressure, the ventricles become dilated with blood as the rapid quivering movement of their walls is insufficient to expel their contents; the muscular action partakes of the nature of a rapid incoordinate twitching of the muscular tissue…The cardiac pump is thrown out of gear, and the last of its vital energy is dissipated in the violent and the prolonged turmoil of fruitless activity in the ventricular walls." MacWilliam spent many years working on ventricular fibrillation and was one of the first to show that ventricular fibrillation could be terminated by a series of induction shocks through the heart [MacWilliam JA 1887]. The first Electrocardiogram recording of Ventricular Fibrillation was by August Hoffman in a paper published in 1912 [Hoffman A 1912]. At this time two other researchers, Mines and Garrey working separately produced work demonstrating the phenomenon of circus movement and re-entry as possible substrates for the generation of arrhythmias. This work was also accompanied by Lewis who performed further outstanding work into the concept of "circus movement". Later milestones include the work by Kerr and Bender in 1922 who produced an electrocardiogram showing ventricular tachycardia evolving into ventricular fibrillation [Kerr WJ et al. 1922]. The re-entry mechanism was also advocated by DeBoer who showed that ventricular fibrillation could be induced in late systole with a single shock to a frog heart [De Boer S 1923]. The concept of "R on T ectopics " was further brought out by Katz in 1928 [Katz LN 1928]. This was called the “vulnerable period” by Wiggers and Wegria in 1940 who brought to attention the concept of the danger of premature ventricular beats occurring on a T wave . Another definition of VF was produced by Wiggers in 1940. He described ventricular fibrillation as - "an incoordinate type of contraction which, despite a high metabolic rate of the myocardium, produces no useful beats. As a result, the arterial pressure falls abruptly to very low levels, and death results within six to eight minutes from anaemia of the brain and spinal cord." [Wiggers CJ et al. 1940]. Spontaneous conversion of ventricular fibrillation to a more benign rhythm is rare in all but small animals. Defibrillation is the process that converts ventricular fibrillation to a more benign rhythm. This is usually by application of an electric shock to the myocardium and will be discussed later.

Electrophysiology of Cardiac cells

If a microelectrode is placed on the surface of a resting myocardial cell and a second (indifferent) microelectrode is placed in a remote location such as the extracellular space then no electrode potential (zero potential) is recorded due to the high impedance of the cell membrane. If however, a glass microelectrode with a tip diameter of less than 0.5 µm penetrates the cell membrane then a potential gradient is detected with a resting membrane potential of -80 to -90 mV in most cardiac cells. This however can range between -50 to -95 mV recorded during diastole depending on the cardiac cell type. The transmembrane resting potential is generated by the difference in ionic concentrations on either side of the cell membrane. The contributing ions include Sodium, Potassium, Chloride and Calcium [Guyton AC 1991]. Potassium is the major cation responsible for the production of a transmembrane potential due to an intracellular concentration of 150 mmol/l and extracellular of about 5 mmol/l with a 30:1 ratio [Carmeliet E 1992]. An opposite gradient exists for sodium ions (Na+) resulting in a higher extracellular concentration of Na+ ions compared with the intracellular concentration. The membrane however is much less permeable to Na+ than K+ ions and so the Na+ gradient does not change the resting gradient appreciably. The Na+/K+ pump actively removes Na+ ions from the cell against its electrochemical gradient at the same time as it pumps K+ ions into the cell against its chemical gradient. This pump is fueled by a Na+/K+-ATPase enzyme that hydrolyses ATP for energy and is bound to the membrane. It requires both Na+ and K+ to function and for every three Na+ ions pumped out from the cell two K+ ions are pumped inwards so an electrical gradient is established. There is a net pumping out of positive charge. The activity of the pump must maintain the same ionic balance even as heart rate increases since the cell loses a slight amount of K+ and gains a slight amount of Na+ with each depolarisation. Cardiac glycosides such as digoxin act directly to block this pump [Zipes P 1995].

Membrane Channels

The cell membrane consists of a bilayer of phospholipid molecules. The tail end is nonpolar and hydrophobic and points inwards towards the centre of the membrane and the polar end is hydrophilic and points towards the outer and inner surfaces of the cell membrane. The Cell membrane or sarcolemma (particularly that area that is hydrophobic) provides a high resistance insulation wrapped around the cell contents allowing only selective permeability to ions. These channels are proteins or phospholipoproteins and may be selective in favouring the passage of a particular ion. It is this lack of free movement of ions across the sarcolemma that results in a potential gradient. The cell has channels that allow flow and span the cell membrane allowing the passage of ions. Other more selective channels exist. These are protein or phospholipoprotein channels, which act, in a more selective fashion. Ion channels can cycle through one of three states – Open (activated), Closed (inactive) and Closed (resting), which can return to Open (activated) with the correct stimulus. This stimulus may be a neurotransmitter binding to an extracellular site – a receptor operated channel or a voltage change – a voltage operated channel. These channels are foci for therapeutic drugs.

Intercalated Discs

The cell membranes of some cell types of adjacent channels form close margins called intercalated discs. The cells are bound by areas of strong adhesion – macula adherens, desmosome and fascia adherens. This aids the transfer of mechanical energy. In the intercalated disc is an area called the nexus or gap junction, which allows close contact across a gap of only 10 to 20 angstroms (1 nm to 2 nm) between cells. These gap junctions provide an area of low resistance ideal for coupling and the passage of ions and small molecules from neighbouring cell cytoplasms. They also confer the myocardium with the property of being anisotropic – that is that the anatomical and biophysical properties vary according to the direction in which they are measured. Resistivity is lower longitudinally and conduction velocity is greater longitudinally.

Cellular Depolarisation and Repolarisation

When the appropriate stimulus is delivered to excitable tissue an action potential is evoked. This is characterised by a sudden voltage change due to depolarisation followed by repolarisation. This is seen in both myocardial cells and in the cells of the nervous system. Different cell types have different and relatively fixed time and voltage relationships. Action potentials, which last only milliseconds in nerve tissue last several hundred milliseconds in cardiac cells. The onset of depolarisation is accompanied by an abrupt increase in permeability to the Na+ ion. Na+ and Ca2+ enter the cell through their respective channels. The Na+ current is regenerative in that the entrance of a little Na+ depolarises the membrane more and leads to increased conductance to Na+ so that even more enters. This raises the intracellular potential to a positivity of about +20 mV. This phase of depolarisation is termed Phase 0 and reflects the fast inward Na+ current typical of working myocardial cells and purkinje fibres. The maximum rate of depolarisation of ventricular cells is 200 volts per second (V/s) and of atrial cells is 100-200 V/s. For purkinje fibres the rate is much faster at 500 V/s. Depolarisation of the cell is followed by a gradual return to the resting membrane potential. It encompasses three phases. Phase 1, were there is an initial rapid return of intracellular potential to 0 mV largely due to the closure of Na+ channels. This is followed by Phase 2, which is seen as a plateau phase resulting from the slow entry of Ca2+ into the cell. Phase 3 is next where there is a return of the intracellular potential to resting potential due to extrusion of K+ ions. At the end of Phase 3, the normal resting membrane potential is re-established and the excess of Na+ remaining within the cells is removed actively via the Na+/K+ pump.

Excitation and Threshold Potential

Excitation of a cardiac cell occurs when a stimulus reduces the transmembrane potential to a critical level known as the threshold potential which is about -60 mV in atrial and ventricular muscle cells and only -40 mV in SA and AV nodal cells. Any stimulus that raises the resting membrane potential towards the threshold potential will promote excitation and vice-versa.


The period of refractoriness is divided into a period of absolute refractoriness during which no stimulus can evoke a response, followed by a period of relative refractoriness when only a stimulus stronger than usual can evoke a response. During this time, the Na+ channels become inactivated. The relative refractory period includes the time when Na+ channels are becoming available. This period is followed by the period of supernormal excitability during which a relatively weak stimulus can provoke a response.

Cardiac Neural Innervation and Vagal Stimulation

The atrioventricular (AV) node and the Bundle of His together share a rich and densely innervated supply of cholinergic and adrenergic nerve fibres. The density of innervation exceeds that seen in the ventricular myocardium. Ganglia, nerve fibres and nerve nets lie close to the AV node. In the canine heart, parasympathetic nerves to the AV node region enter at the junction of the Inferior Vena Cava on the inferior border of the left atrium adjacent to the coronary sinus. Nerves in contact with the AV nodal fibres have been noted along with agranular and granular vesicular processes, presumably representing cholinergic and adrenergic processes. Acetylcholine release may be concentrated around the N region of the AV node [Zipes DP 1994] It is found that the autonomic neural input demonstrates some sidedness. The right sympathetic and vagal nerves affect the sinus node more than the AV node and vice-versa. Despite much overlap in the innervation, specific branches of the sympathetic and Vagal nerves can be seen to innervate specific parts of the conducting system [Braunwald E 1998]. The vagus modulates the cardiac sympathetic activity at prejunctional and postjunctional sites by two methods. The first involves regulating the amount of noradrenaline released and the second by inhibiting cyclic AMP-induced phosphorylation of cardiac proteins such as phospholamban.

The Electrophysiology of Ischaemic Myocardium

It is impossible to discuss the underlying mechanisms of the initiation and maintenance of lethal cardiac arrhythmias without some recourse to the electrophysiological environment in the Peri-arrest State. For this review I have used the most salient points from the excellent reviews by Cascio and Janse [Cascio WE et al. 1995, Janse MJ et al. 1989, Janse MJ et al. 1992] Myocardial ischaemia has been defined by Hearse as a condition in which arterial perfusion is inadequate to meet the energy of the cells. This leads to adaptive biochemical mechanisms that alter the ionic homeostasis [Hearse DJ 1994]. In clinical practice, myocardial ischaemia can be encountered in a variable spectrum of scenarios. These range from the abrupt and complete cessation of coronary blood flow and myocardial perfusion that can accompany acute plaque rupture and vessel occlusion which typically leads to myocardial infarction. A different scenario involves those patients with recurrent episodes of ischaemia with significant flow impairment with or without increased demands. These differences can be separated into the no-flow and low-flow conditions or alternatively the anoxic vs. hypoxic phenomena [Cascio WE et al. 1995]. One important issue of the occluded no-flow situation is the lack of washout with the accumulation of energy substrates and metabolic by products. No-flow is associated with an earlier inhibition of anaerobic glycolysis compared to the low-flow situation due to the accumulated metabolic products of the enzymes of the glycolytic pathway. Both scenarios can obviously co-exist with the low-flow picture being present at the ischaemic penumbra around an area of no-flow induced infarcted tissue. The contribution of each depends on the underlying cause and the existence of collaterals [Cascio WE et al. 1995]. The loss of tissue perfusion results in a cascade of reactions. Myocardial ischaemia results in metabolic adaptations that alter both impulse formation and propagation. The main changes seen are within the intracellular and extracellular compartments with alterations in the concentration of high-energy phosphate compounds and the accumulation of metabolic by-products. These can be measured in terms of intra and extracellular pH and the change of concentration of K+ ions and the following ions and molecules such as Ca2+, Na+, Mg2+, ATP, ADP and inorganic phosphate. Alterations to the transmembrane currents occur as a consequence of these ionic changes as well as changes to adrenergic receptor stimulation and the accumulation of lactate, amphipathic compounds and adenosine. These local changes of the underlying electrophysiological properties of the heart lead to a disorder of active and passive membrane properties that underlie alterations in excitability, abnormal automaticity, refractoriness and conduction [Janse MJ et al. 1989]. In other words, it forms an environment suitable for the potential generation of arrhythmias. As we will discuss in more depth later on, impulse propagation in cardiac tissue requires the rapid depolarisation of an excitable myocyte and the efficient transfer of the resultant action potential from cell to cell. This requires local circuits of excitation involving the sarcolemmal membrane and the intra and extracellular spaces that spread the depolarisation along the individual fibre, nexal membrane and gap junctions. Following depolarisation, the myocyte remains refractory for a period of time until Na+ and Ca2+ currents allow the reconfiguration of the appropriate state to again allow trans-sarcolemmal current flow. The length of the refractory period is dependent on voltage and time dependent components. It is sensitive to local ionic and metabolic conditions as well as the presence of secondary messengers brought about due to ischaemia [Janse MJ et al. 1992].

High Energy Phosphate Compounds

High-energy phosphate compounds provide the energy for the maintenance of ionic homeostasis. They include but are not limit to Adenosine Triphosphate (ATP), Adenosine Diphosphate (ADP) and Creatine Phosphate (CP). They are most efficiently produced by glycolysis with subsequent oxidative phosphorylation in the cellular mitochondria. Following the onset of ischaemia, they are rapidly consumed. Anaerobic metabolism in comparison to aerobic metabolism is a very poor producer of ATP. During the early (15 seconds - 10 minutes) and late (10 - 25 minutes) phases of ischaemia, the ATP levels are sufficient to maintain cellular viability. During later stages, the channels, exchangers, pumps and enzyme systems dependent on ATP are notably affected. The intracellular levels of Ca2+ and Na+ and extracellular K+ then begin to rise.


Ventricular arrhythmias including ventricular ectopics and ventricular fibrillation occur during ischaemia with a bimodal distribution. The first phase occurs within 5 and 7 minutes of the arrest of perfusion. The second phase peaks between 12 and 30 minutes [Kaplinsky E et al. 1979, Meesman W et al. 1978].

Extracellular Potassium (K+e)

During the early phase, potassium (K+e) begins to accumulate in the extracellular space in both animal and human hearts. Changes in K+e occur in three phases. The initial phase begins within 15 seconds of the arrest of tissue perfusion and can last from 5 to 10 minutes in duration. During this period K+e rises rapidly and achieves levels as high as 15 mmol/l. the rise is related to the fall in the availability of high-energy phosphates. The potassium effluxes across the cell membrane. Hypoxia is known to activate K+ channels, which may then act as conduits for cellular K+ loss and cause shortening of the action potential. K+ATP channels which are activated when ATP falls during hypoxia, metabolic inhibition or ischaemia may be responsible for some of the K+ efflux [Noma A et al. 1983]. Pharmacological blockade of the K+ATP channel during ischaemia with sulphonylurea reduces but does not prevent cellular K+ loss [Yan G-X et al. 1993]. Some work suggests that K+ATP channel blockage does seem to prevent ventricular fibrillation in ischaemic rats [Wollebethe CD et al. 1989]. During the next 10 to 15 minutes a plateau phase develops and K+e does not rise any further and may in fact fall. There is a net absence of efflux at this time. During this phase there may well be some Na+-K+ATPase pump activity as administration of cardiac glycosides leads to the attenuation of the plateau that is characteristically seen at this stage. The third and final phase begins as a secondary rise in K+ occurs. This begins approximately 20 to 30 minutes after the onset of no-flow anoxic ischaemia. K+ rises to levels in excess of 30 mmol/l. At this stage, cell to cell uncoupling occurs, the onset of ischaemic contracture begins and an increase in long-chain acylcarnitines occurs in addition. There are secondary falls in the free energy of hydrolysis of ATP [Owens L et al. 1993, Yamada KA et al. 1994]. At the same time Ca2+ increases dramatically. Whether this is the cause or the harbinger of cell death remains debatable. Here I have discussed myocardial ischaemia with its local increase in extracellular K+ [Weiss JN et al. 1995] and shortening of the action potential duration in ischaemic regions of the heart [Wilde AAM et al. 1990]. Myocardial ischaemia leads to regional heterogeneity between ischaemic and non-ischaemic regions. This as we will later see is believed to be the substrate required for the genesis of potentially lethal arrhythmias.

Mechanisms of Ventricular Fibrillation

Zipes divides the mechanisms of arrhythmia genesis into disorders of impulse formation and disorders of impulse conduction or both [Zipes DP 1994]. Zipes reminds us of the caveat that the present diagnostic tools do not permit unequivocal determination of the electrophysiological mechanisms responsible for most clinically occurring arrhythmias or their ionic basis. This he states is especially true for ventricular arrhythmias. In general terms it is almost impossible to separate re-entry and automaticity. In most circumstances we are only able to suggest that such an arrhythmia is consistent with a particular underlying mechanism. For many years due to the practical problems involved with mapping large areas of the heart simultaneously ventricular fibrillation has been hard to study. Most observers have confined their interest and work to the induction and termination of ventricular fibrillation. Much of the current data on the dynamic electrophysiological changes during cardiac arrhythmias comes either from computer modeling, electrode studies or the use of high resolution optical mapping. In optical mapping, the emitted fluorescence caused by transmembrane potential changes of cardiac cells is recorded by a video camera. The video images of the epicardial surface can be analysed pixel by pixel by image processing techniques. A typical high-resolution image may consist of 200 by 200 pixels giving data on 40,000 epicardial sites [Gray RA et al. 1999]. Mathematical models such as the Panfilov and Keener model also exist which facilitate the understanding of the relationships between the electrocardiogram and the underlying spread of patterns of depolarisation throughout a three dimensional myocardium. The Panfilov and Keener model specifically incorporates a realistic anatomical representation of the right and left ventricular structure. There is also an accurate representation of myocardial fibre orientation and distribution of the dispersion of action potential dynamics [Panfilov AV et al. 1993, Nielson PMF et al. 1991].

Re-entry -"disorders of impulse conduction"

The role of re-entry or circus motion was demonstrated separately by Mines and Garrey [Mines GR 1913, Garrey WE 1914]. Mines created a ring of excitable tissue by cutting the atria out of the ray fish. Garrey cut out a similar ring from the turtle ventricle. They were both able to show that if a ring of excitable tissue was stimulated at a single point that the subsequent waves of depolarisation would pass around the ring. The waves eventually meet and cancel each other out, but if an area of transient block occurred with a refractory period that blocked one wavefront and subsequently allowed the other to proceed retrogradely over the other path then a self-sustaining circus movement phenomena would result. For this to happen however there needs to be some form of non-uniformity. In practice, this may be an area of ischaemic or infarcted myocardium or underlying scar tissue. It is possible to think of the advancing wave of depolarisation as a dipole with a head and a tail. Evidently therefore the length of the refractory period and the time taken for the dipole to travel a certain distance - the propagation velocity will determine whether such a circumstance will arise for re-entry to occur. Factors that promote re-entry would include a slow propagation velocity, a short refractory period with a sufficient size of ring of conduction tissue. These would enable a dipole to reach an area that had been refractory and was now able to be depolarised with continuation of the wavefront. In clinical practice therefore, factors that would lead to the right conditions to favour such re-entry mechanisms include - increased heart size through hypertrophy or dilatation, drugs which alter the length of the refractory period and areas of cardiac disease. Therefore, the substrate of ventricular fibrillation is transient or permanent conduction block. Block due either to areas of damaged or refractory tissue leads to areas of myocardium for initiation and perpetuation of fibrillation through the phenomenon of re-entry.

Abnormal Automaticity - "disorders of impulse formation"

Automaticity is a measure of the propensity of a fibre to initiate an impulse spontaneously. The product of a hypoxic myocardium can be hyperirritable myocardial cells. These may then act as pacemakers. The ventricles are then being stimulated by more than one pacemaker. This may well lead to the generation of a circus entry arrhythmia. Scar and dying tissue is inexcitable but around these areas usually lies a penumbra of hypoxic tissue which is excitable. Ventricular excitability may be the trigger to generate re-entry arrhythmias. It is interesting that most cardiac pathologies with an associated increased propensity to arrhythmia development have an associated loss of membrane potential. That is, the maximum diastolic potential is less negative and therefore exists closer to the threshold potential. Cellular depolarisation can be due to a raised external concentration of K+, a decreased intracellular concentration of K+, increased permeability to Na+ or a decreased permeability to K+. The ionic basis of automaticity is the net gain of an intracellular positive charge during diastole in the presence of a voltage dependent channel activated by potentials negative to –50 to –60 mV. Myocardial cells are exposed to different environments. Normal cells may be exposed to hyperkalaemia, abnormal cells may be perfused by normal environment. For example with a healed myocardial infarction, abnormal cells can be exposed to an abnormal environment such as with a myocardial infarction with myocardial ischaemia. In conditions such as myocardial ischaemia possible mechanism of arrhythmia generation include the resulting decreased internal K+ concentration, the increased external K+ concentration, noradrenaline release and acidosis [Ho K 1993].

Triggered Activity

Triggered activity can occur due to the presence of afterdepolarisations. These are depolarising oscillations in the membrane voltage induced by preceding action potentials. These can occur before or after full repolarisation of the fibre and as such are termed either early (EADs) or delayed afterdepolarisations (DADs). All afterdepolarisations may not reach threshold potential but if they do they can trigger another afterdepolarisation and thus self perpetuate.

Theories of Ventricular Fibrillation

Most theories of arrhythmias are focused around the basic premise that the heart is an example of a generic excitable medium. The propagation of cardiac action potentials occurs from cell to cell through low-resistance intercellular gap junctions, with the transmembrane current (Im) being a major index of the cellular response. There are subsequently many theories about fibrillation in excitable media but it is only now with the increasing sophistication of experimental techniques that have become available that we can now try to study the complex spatial patterns that underlie sustained fibrillation. Over the past 80 years, with the work of Mines, Moe, Garrey and Lewis, numerous theories have been put forwards to attempt to explain the disordered electrical activity of fibrillation. Theories include a single ectopic focus, multiple automatic foci, a single circus pathway or multiple re-entrant circuits. Others have suggested that irrespective of the underlying mechanism that fibrillation should be explained in terms of many electrical waves spreading through the 3D myocardium, and what we see is really complex three dimensional patterns of electrical excitation of the myocardium [Jalife et al. 1996]. Others have looked at ventricular fibrillation in terms of chaos theory [Janse MJ 1995].

The Multiple Wavelet hypothesis

The multiple wavelet hypothesis of ventricular fibrillation was formulated by Moe and Abildskov more than 40 years ago. It suggests that fibrillation is due to the presence of a number of independent wavelets that propagate randomly through the myocardium around multiple areas of refractory tissue. Maintenance of fibrillation is thought to depend on the number of wavelets. With a paucity of wavelets they either die out or fuse into a single activation front leading to a resumption of a more organised regular tachycardia or flutter. Support for this theory was found in the mapping studies performed by Allessie and others studying acetylcholine induced atrial fibrillation in normal canine hearts [Moe GK et al. 1959]. The work of Mines and Garrey demonstrated that a critical mass of myocardium was a prerequisite for ventricular fibrillation to persist [Mines GR. 1913, Garrey WE. 1914]. In dogs, the critical mass is 25% of the total ventricular mass. Ventricular fibrillation can only be sustained in large hearts eg. Man, dogs and porcine. Inducing ventricular fibrillation is difficult in smaller hearts e.g. frogs. When ventricular fibrillation starts it can be assumed that the number of wavelets is few and that this number increases. In 1930 Wiggers studied fibrillation using cinematographic techniques was able to break it temporally into four distinct stages. The first stage, he named the undulatory or tachysystolic phase generally lasted less than 1 second in the dog, during which 2 to 8 peristaltic waves frequently appeared to arise from one source and swept across the ventricles. This was then followed by the conclusive incoordination phase lasting 15 to 40 seconds with waves of contraction with distinctly different rhythms and sequences that passed over the ventricles. The third stage was called the stage of tremulous incoordination where coarse convulsive movements were visible and this stage lasted 2-3 minutes. Contraction waves appeared to spread rapidly over short distances. Finally, he described a fourth stage characterised by progressive atonic incoordination - waves become coarser and slower and no longer maintained the even levels of elevated intraventricular pressure that characterised stages one to three. Movement decreased until eventually only slight movements persisted in a few regions Moe using the work of Garrey and Mines and other observations revived the theory that re-entry provides the basic mechanism for ventricular fibrillation to develop the well known multiple wandering wavelet hypothesis. He constructed a computer model of the heart and modeled the atria as a hexagonal sheet one unit thick with each unit naturally having six neighbours. A unit when fired would then automatically transmit excitation to the 6 adjoining units and whether they do likewise will depend on their degree of refractoriness. He was therefore able to produce activity that resembled fibrillation and consisted of multiple wandering wavelets, which exhibited self-sustaining turbulence. The model of ventricular fibrillation was therefore characterised by multiple wavelets and re-entrant circuits. Fibrillation was inducible in this computer model if a calculated value determined by the product (fibrillation number) of the following parameters: - characteristic length (L), inverse of conduction velocity (T), and the inverse of a constant that determines the refractory period (K-2) reached a critical value [Moe GK et al. 1964]. This “fibrillation number” hypothesis suggests that the physiological complexity of cardiac fibrillation is determined in part by the characteristic length or tissue mass and that a tissue mass reduction may result in progressively less electrophysiological behaviour. Interestingly work on atrial fibrillation by Allessie and other investigators has confirmed Moe’s multiple wavelet theory as the likely basis for atrial fibrillation [Moe GK et al. 1964]. It was Mines who initially demonstrated the importance of ventricular mass in the development and maintenance of ventricular fibrillation [Mines GR 1914]. More recently, similar work has been performed to look at the spatio-temporal complexity of ventricular fibrillation and its relationship to tissue mass. These have shown that as tissue mass decreased the number of wave fronts decreased, the life-span of re-entrant wave fronts increased and the cycle length, the diastolic interval and the duration of the action potential actually lengthened. The researchers looked at the relationship of ventricular fibrillation to chaos theory and found that a decrease in the number of wavefronts in ventricular fibrillation by tissue mass reduction caused a transition from chaotic to periodic dynamics via the quasiperiodic route. They regarded ventricular fibrillation as chaos and a regular rhythm as periodicity [Kim et al. 1997]. Chen and co-workers showed that during ventricular fibrillation in situ in normal ventricles the re-entrant wave fronts had a limited life span lasting only 1.36 seconds or 9.6 cycles before termination [Chen PS et al. 1988]. In another study, Cha looked at the termination of re-entrant activity and found that re-entrant wave-front activity terminated when wavefronts that had arisen initially from outside the mapped tissue interfered with the re-entrant pathways. It was noted that the refractory period of the fibrillating myocardium was between 48 and 77 milliseconds. Because the refractory period was much shorter than the cycle length, a large excitable gap was present in the re-entrant circuit. It was also noted that myocardial fibre orientation was an important determinant of the site of conduction block and although subendocardial ablation slowed the wavefront propagation, it did not prevent the generation and the maintenance of re-entry and fibrillation [Cha YM et al. 1994]. Mapping studies have looked at ventricular fibrillation that has developed spontaneously during the early phases of myocardial ischaemia and reperfusion in experimental animals. Other studies have also looked at ventricular fibrillation that was induced electrically in normal hearts. These studies have confirmed the multiple wavelet hypothesis. There is also evidence that a further role is played by more focal mechanisms [Pogwizd SM 1987, Pogwizd SM 1990, Coronel R 1992, Chen PS 1990].

Spiral Waves

Another hypothesis supports the idea that ventricular fibrillation may be due to spiral wave activity. This mechanism does not depend on any peculiarities of the cardiac muscle and may be demonstrated in any excitable medium. For example, spiral waves may be observed in the so-called Belousov-Zhabotinsky reaction as well as in the formation of multicellular aggregates in the social amoeba, the spreading depression in the retina of birds, and the propagation of calcium waves in the cytosol of Xenopus oocytes. In the case of the heart, the spiral wave may be complementary to already traditional ideas, which are based on the classical idea of re-entry [Pertsov AM et al. 1993]. However, as we have discussed, theories of ventricular fibrillation generation and maintenance are based around the idea that abnormal myocardial tissue sets up changes in refractoriness that allow an environment for re-entry to exist. This does not explain however the cases of ventricular fibrillation that occur in those hearts without any obvious structural pathology. To address this problem other workers have tried to formulate theories to study fibrillation in normal myocardium. A novel theory and description of fibrillation has been suggested by Winfree. It involves the existence of three-dimensional rotors of electrical activity that become unstable when the heart thickness and exceeds some critical value. A rotor is defined as a wave of excitation propagating around a topological defect, which is known as a phase singularity. A spatial phase singularity is a site in an excitable medium at which the phase of the site is arbitrary. It is seen that the neighbouring and surrounding elements exhibit a continuous progression of phase that is equal to ± 2 around this site [Gray et al. 1998]. Winfree states "several pinned rotors would collectively resemble fibrillation in the electrical summation of a body surface electrocardiogram, and epicardial electrodes or catheter electrodes would still reveal their individual local periodicities" [Winfree AT 1994]. Gray has not refuted this theory but has suggested that it is the speed of the rotors that contributes to the characteristic appearance of ventricular fibrillation [Gray et al. 1995]. Interestingly it was demonstrated that when the normal myocardium is activated by a single pinned rotor that it presents as a "monomorphic tachycardia". Winfree has suggested that if this rotor shows drift or simple meander then polymorphic tachycardia results, which resembles torsades de pointes. If it exhibits hypermeander or something faster and less regular than seen elsewhere then it presents as fibrillation. Gray and co-workers were able in a simulation to place seven stationary rotors in the heart, which resulted in a regular periodic ECG quite unlike fibrillation. They presented evidence to suggest that a single rapidly moving rotor gives rise to ECG patterns of activity identical to ventricular fibrillation. Although not all episodes of ventricular fibrillation were due to just one rotor they suggested that there must be a critical thickness of many rotors for fibrillation to occur. They also suggest that the Doppler phenomenon provides a robust explanation for the narrow banded frequency spectra characteristic of fibrillation that has perplexed investigators for years. Gray has also been able to produce a new algorithm that markedly reduces the amount of data required to depict the complex spatiotemporal patterns of fibrillation [Gray RA et al. 1998]. A phase variable q was calculated at each site and this was found to be a more useful than simply using the fluorescence signal F that simplified the analysis of fibrillation. With this new method he was able to study fibrillation by phase q(x,y,t) and to study the detailed dynamics of spatial phase singularities and rotors. It was possible then to look at the initiation, maintenance, and termination of arrhythmias. Panfilov has suggested a refinement to Winfree’s proposition by modelling turbulence in 3D. Their hypothesis is that the scroll wave in 3D has more complicated dynamics because of the meandering of the curved filament, which, together with 2D pulse instabilities cause local distortions [Panfilov AV 1995]. Jalife has recently reviewed the theoretical concepts of ventricular fibrillation and he has proposed a new algorithm, which demonstrates that ventricular fibrillation is not random and may be analysed quantitatively. The approach is based on video imaging of voltage sensitive dyes from 20,000 sites recorded simultaneously on the epicardium of sheep and rabbits' ventricles. Ventricular fibrillation activity shows a strong periodic component of 500 beats per minute. Phase maps showed that ventricular fibrillation depends on organisation of wavefronts around a small number of field irregularities with short lifespans and form as a result of interactions of wavefronts with obstacles in their paths. It seems there is a high degree of temporal and spatial organisation [Jalife J 1999]. ==Ischaemic==, Electrically-induced and Idiopathic Ventricular Fibrillation It can be useful to study ventricular fibrillation grouped by cause to determine differences and clues to the underlying mechanisms. Potential initiators include ventricular fibrillation induced by ischaemia in abnormal hearts, electrical stimuli in normal or diseased hearts and those in whom ventricular fibrillation occurs for no clear underlying reason.

Ischaemia related ventricular fibrillation

Elharrar with Janse and co-workers have looked at ventricular fibrillation secondary to ischaemia and reperfusion and have described several possible mechanisms. For example, ventricular fibrillation may be induced by injury currents crossing the border between normal tissue and an ischaemic region. This then sets in motion a re-entrant mechanism involving re-entry into the ischaemic myocardium. The precise mechanisms of ventricular fibrillation development in myocardial ischaemia remain to be elucidated. As we have discussed, myocardial ischaemia is characterised by a variety of electrophysiological alterations such as slow conduction and variable degrees of conduction block that provide the necessary conditions for re-entry to occur [Elharrar V et al. 1977]. Much work has involved the use of epicardial and myocardial electrocardiographic mapping of multiple sites following coronary artery occlusion and reperfusion to look at injury currents and excitation patterns. Janse used 60 electrodes to study extracellular epicardial and intramural sites in the left ventricle of isolated porcine and canine hearts during the first 15 minutes following vessel occlusion and reperfusion of the left anterior descending artery. The arrhythmias that developed showed fragmented wavefronts and multiple wandering wavelets. These followed tortuous paths on the epicardial surface. When circus type movements occurred they were seldom complete and had small diameters of only about 0.5 cm. The small area of ventricle studied is a common major limitation of mapping studies [Janse MJ et al. 1980]. Ideker’s teams were able to induce ventricular fibrillation in 8 out of 10 canine hearts in open chested dogs following the occlusion of the proximal circumflex artery. They made simultaneous recordings from 27 electrodes spaced over both ventricles and showed that a period of organised epicardial activation occurred in the transition to ventricular fibrillation. Analysis of the first 1.5-2.5 seconds in the period of transition from sinus rhythm or ventricular tachycardia to ventricular fibrillation showed that a single, organised, orderly and rapid wavefront broke across the epicardium near the border of the ischaemic-reperfused region. It then passed across the non-ischaemic portion of the ventricles to the opposite side of the heart. Each activation front arose from the border of ischaemic and non-ischaemic tissue. The time for each activation cycle to traverse the ventricles increased and conduction velocity slowed so that cycles of activation came closer together. This resulted in activation fronts arriving at the epicardium of the ischaemic-reperfused zone before they had terminated over the right ventricle. This overlap resulted in as many as three co-existing activation fronts co-existing on the epicardium. Analysis of the mechanisms of the development of ventricular fibrillation during early myocardial ischaemia was performed using a three-dimensional mapping system and multiple bipolar electrodes (232) at up to eight levels in the feline heart [Pogwizd SM et al. 1987]. Ventricular fibrillation was induced by occlusion of the proximal LAD in open-chested cats with the subsequent development of ventricular tachycardia and then ventricular fibrillation. A study of malignant reperfusion arrhythmias during ischaemia showed that ventricular activation time was significantly delayed and ventricular tachycardia occurred within 15 seconds post reperfusion in all size animals and progressed to ventricular fibrillation in 50%. Non-sustained VT was seen to occur due to both re-entrant and non-re-entrant means. 75% of non-sustained VT was usually initiated by non-re-entrant means at the border of the reperfused zone. 25% of non-sustained VT was due to intramural re-entry. Finally, VT always preceded the development of ventricular fibrillation, and was usually initiated by non-re-entrant mechanisms in the subendocardium at the border of the reperfused zone and was maintained by re-entrant and non-re-entrant means. More recently, Pogwizd’s team, in a similar study, have looked at ventricular fibrillation during myocardial ischaemia without perfusion, again using 232 bipolar electrodes in an open chested feline model. They found that initiation of VT, which led to VF, occurred by intramural re-entry in 3 out of 4 cases. The coupling intervals of the initiating beats of VT that lead ultimately to fibrillation were no different from those responsible for the non-sustained VT. Maintenance of the VT that led to VF was due primarily to intramural re-entry (84% of cases) involving multiple activation sites in and around the border of the ischaemic zone. Non-re-entrant mechanisms arising in the subendocardium and subepicardium also contributed to the maintenance of VT before development of ventricular fibrillation. Transition from ventricular tachycardia to ventricular fibrillation was due exclusively to intramural re-entry with initiation of the re-entrant beats in the subendocardium and occasionally subepicardium. Acceleration of the tachycardia by re-entry along with a rapid and inhomogeneous recovery of excitability led to increased functional block and conduction delay. It appears that the initiation and maintenance of VT leading to ventricular fibrillation during early ischaemia is due to intramural re-entry with some contribution from non-re-entrant mechanisms. The development of ventricular fibrillation is due to continued intramural re-entry and rapid recovery of excitability [Pogwizd SM et al. 1990]. Ideker’s team more recently have documented that ventricular fibrillation which is perpetuated by functional re-entry occurs when an activation front blocks and rotates around tissue that is excitable. It has been noted that electrocardiograms recorded near these regions typically contain two sequential deflections representing activation on either side of the block. From this, they have been able to develop an algorithm to detect functional block during ventricular fibrillation from a single electrode recording [Evans FG et al. 1999].

Electrically-induced ventricular fibrillation

Ventricular fibrillation can also be induced by delivery of a large electrical stimulus during the "vulnerable period" of the cardiac cycle. It was thought that this would result in an activation front that propagated away from the site of stimulation and blocked uni-directionally when it reached areas that had not yet recovered excitability. A non-uniform dispersion of refractoriness was thought to result in re-entry as adjacent areas of tissue remain in different states of recovery. This would provide the substrate for propagation fronts to circle back into areas which were formerly refractory areas and now excitable. This re-entry would then lead ultimately to ventricular fibrillation. Other researchers have shown that ventricular fibrillation can occur where there is a uniform dispersity of refractoriness [Frazier et al. 1989, Chen P-S 1988, Shibata N 1988]. It had been felt that non-uniform dispersity of refractoriness was a sine qua non of ventricular fibrillation. Frazier and co-workers have shown that ventricular fibrillation can be induced when there is uniform refractoriness – it changes by the same amount over a distance throughout the region of interest. Frazier and colleagues looked at the hypothesis that the field of a premature stimulus interacting with relatively refractory tissue could create unidirectional block and re-entry in the absence of non-uniform dispersity of recovery. Recordings of up to 120 points on a small region of the right ventricle were made. Pacing was performed from electrodes placed along the side of the area being mapped. Shocks were delivered via a second electrode and these were able to induce circus re-entry. It was found that when the field created by the second electrode was studied that field strengths and tissue refractoriness were uniformly dispersed at an angle to each other. Circus re-entry occurred around a critical point where the second field of ~5 V/cm intersected tissue approximately at the end of its refractory period [Frazier DW et al. 1989]. Jalife’s team has used both computer simulations and video imaging of the rabbit heart using fluorescence to demonstrate the transmembrane potential with high-resolution optical mapping. Their work has suggested that irrespective of the underlying mechanism that fibrillation should be explained and considered in terms of many electrical waves spreading through a three dimensional myocardium. What we see is complex three-dimensional spatio-temporal patterns of electrical excitation of the myocardium with "drifting vortices of electrical activity" [Jalife J et al. 1996].

Idiopathic ventricular fibrillation

Ventricular fibrillation occurring in the structurally normal heart is classified as "primary electrical disease" or “idiopathic ventricular fibrillation”. This suggests that the cause or trigger lies within some property of the conducting system of the heart. The model of "primary electrical disease" is the long QT syndrome (LQTS) in which the altered membrane ionic channel function underlies QT prolongation. The altered channel activity is due to mutations in genes encoding ion channels [Kass RS et al. 1996]. In 1992, Brugada and Brugada described a small group of 8 patients with a history of aborted sudden death and a distinctive ECG consisting of right bundle branch block with ST elevation in the right precordial leads and a normal QT interval. Structural heart was absent as determined clinically, biochemically and by echocardiographic and angiographic examinations [Brugada P et al. 1992]. Interestingly, the ECG pattern was dynamic and modifiable by changes in autonomic tone and antiarrhythmic drugs. It has been suggested that "Marked dispersion of refractoriness of cardiac tissue or extreme anisotropic conduction properties of the conduction system and the ventricular muscle" represent the underlying pathophysiological abnormality [Brugada P et al. 1992]. Similar findings have also been described in a subgroup of males in southern Asia. In northeastern Thailand, sudden unexpected death typically occurring during sleep has been described in young men, many of whom have a family history [Tatsanavivat P et al. 1992]. These syndromes provide an interesting field of studying the mechanisms of the initiation and maintenance of ventricular fibrillation without the need to consider extrinsic initiators. Study of these patients may help further determine the mechanism of ventricular fibrillation.

Characteristics of the Ventricular Fibrillation Waveform

Ventricular fibrillation can be described in terms of its electrocardiographic waveform appearance. All waveforms can be described in terms of certain features such as amplitude and frequency. Researchers have looked at the frequency of the ventricular fibrillation waveform to see if it helps to elucidate the underlying mechanism of the arrhythmia or holds any clinically useful information. More recently, Gray has suggested an underlying mechanism for the frequency of the waveform that has puzzled investigators as possibly being a manifestation of the Doppler effect of rotors of fibrillation [Gray RA et al. 1998]. Analysis of the fibrillation waveform is performed using a mathematical technique known as Fourier analysis.

The Fourier Transform

The Fourier Transform is based on the discovery that it is possible to take any periodic function of time x(t) and resolve it into an equivalent infinite summation of sine waves and cosine waves with frequencies that start at 0 and increase in integer multiples of a base frequency f0 = 1/T, where T is the period of x(t). Here is what the expansion looks like

An expression of the form of the right hand side of this equation is called a Fourier Series. The job of a Fourier Transform is to figure out all the ak and bk values to produce a Fourier Series given the base frequency and the function x(t). You can think of the a0 term outside the summation as the cosine coefficient for k=0. There is no corresponding zero-frequency sine coefficient b0 because the sine of zero is zero, and therefore such a coefficient would have no effect. Of course, we cannot do an infinite summation of any kind on a real computer, so we have to settle for a finite set of sines and cosines. It turns out that this is easy to do for a digitally sampled input, when we stipulate that there will be the same number of frequency output samples as there are time input samples. In addition, we are fortunate that all digital recordings have a finite length. We can pretend that the function x(t) is periodic, and that the period is the same as the length of the recording. In other words, imagine the recording repeating forever, and call this repeating function x(t). The duration of the repeated section defines the base frequency f0 in the equations above. To verify that the transform is functioning correctly, you could then generate all the sines and cosines at these frequencies, multiply them by their respective ak and bk coefficients, add these all together, and you will get your original recording back.

Discrete Fast Fourier Transform algorithm

The Discrete Fast Fourier Transform is an algorithm formulated by Cooley and Tukey [Cooley JW and Tukey JW 1965] and converts a sampled complex-valued function of time into a sampled complex-valued function of frequency Fast Fourier analysis is a mathematical method to separate and express a waveform as its component sine and cosine waves. The improvements in computer theory and algorithm design that enabled Cooley and Tukey in the 1960s and others beforehand provided a method of doing this that reduced the computational time and thus allowed FFT analysis to be performed by microprocessors. This equation shows the exact relationship between the inputs and outputs. In this equation, xk is the kth complex-valued input (time-domain) sample, yp is the pth complex-valued output (frequency-domain) sample, and n = 2N is the total number of samples. Note that k and p are in the range 0.. n-1.

Although this formula tells you what the FFT is equivalent to, this formula is not how the FFT algorithm is implemented. This formula requires O(n2) operations, whereas the FFT itself is O(n*log2(n)). In other words, if you were to use the formula above, it would be much slower than using the FFT algorithm.

Power Spectrum

The distribution of frequency and power of a waveform can be expressed as a power spectrum in which the contribution of different waveform frequencies to the waveform under analysis is measured. This can be expressed as either the dominant or peak frequency i.e. the frequency with the greatest power or the median frequency which divides the spectrum in two halves. Frequency analysis has many other uses in medicine and in cardiology including analysis of heart rate variability and assessment of cardiac function as well as in imaging and acoustics [Shusterman V et al. 1999, Kaplan SR et al. 2000].

Frequency Characteristics of Ventricular Fibrillation

Ventricular fibrillation despite its appearance of a random waveform has a clear dominant frequency with a narrow bandwidth and a peak in the power spectrum around 9 to 12 Hz. Carlisle induced ventricular fibrillation in dogs using ischaemia, reperfusion and electrical stimulation in the presence of ischaemia and then examined the frequency response. Fast Fourier transform analysis was used to determine whether the underlying initiator of fibrillation would alter the frequency content of the ventricular fibrillation electrocardiographic waveform [Carlisle EFG et al. 1990]. It was found that the frequency in those hearts with ventricular fibrillation induced by an electrical stimulus was 9.9 Hz and remained above 9 Hz for 70 seconds and then fell to 5 Hz. A similar pattern was seen in hearts with ventricular fibrillation induced by acute ischaemia and one undergoing reperfusion. In both these cases the dominant frequency was initially 12.2 Hz and 12.3 Hz respectively and they both followed a similar pattern as the frequency fell. However, fibrillation due to the administration of potassium and ouabain had a dominant frequency that was much slower at 4.8 Hz and 7.1 Hz respectively. Wiggers in his description of fibrillation 70 years earlier had also noted that after 1 or 2 minutes the ventricular fibrillation frequency fell and the rate slowed down and the organisation of physical activity appeared to increase as ventricular fibrillation progressed [Wiggers CJ 1930]. Other operators have also found that the frequency characteristics of repeated episodes of ventricular fibrillation induced in the same subjects show fair-to-good but not excellent reproducibility. Bipolar recordings were far more reproducible than unipolar recordings, but both bipolar configurations had similar reproducibility. These findings have implications for both the pathophysiology of induced ventricular fibrillation and the design of ventricular fibrillation detection algorithms [Taneja T et al. 1997]. Fibrillation frequency falls with time and Dzwoncyzk was able to produce an algorithm based on frequency analysis, which could predict the total elapsed time since the onset of ventricular fibrillation with an average error of  0.86 min. It is possible that analysis of the waveform frequency would enable different defibrillation and resuscitation strategies to be implemented [Dzwonczyk R et al. 1990]. Strohmenger, using frequency analysis to analyse fibrillation was able to show that median frequency, dominant frequency, and amplitude were predictive of countershock success in humans [Strohmenger HU et al. 1997]. As for the underlying mechanism, Mandapati and colleagues examined the effects of global ischaemia on ventricular fibrillation in an isolated rabbit heart model. They suggest that rotating spiral waves are the most likely underlying mechanism of ventricular fibrillation and that they contribute to its frequency content. It was postulated that the ischaemia-induced decrease in the frequency of ventricular fibrillation was due to an increase in the periodicity of the spiral waves that occurred secondary to an increase in the core area [Mandapati R et al. 1998]. Work performed by Gray and others in an attempt to look at the spatial and temporal organisation of fibrillation has involved the use of potentiometric dyes and video imaging to record the dynamics of transmembrane potentials at multiple sites during fibrillation. They have shown that the transmembrane signals at each site exhibit a strong periodic component centred near 8 ± 3 Hz [Gray RA 1998]. Analysis of the frequency spectrum of the fibrillation waveform during ischaemia shows that the action potential duration of myocytes is reduced and the frequency of depolarisations increased [Akiyama T et al. 1981]. The time interval between the activation fronts correlates with the frequency with activation fronts having been measured of 120-130 ms in dogs and pigs. This correlates with a ventricular fibrillation frequency of 7-9 Hz. The precise origin of the power spectrum however remains controversial. As mentioned, Gray has suggested that the signal is due to the Doppler phenomenon induced by moving rotors of fibrillation and suggests that this may provide a robust explanation for the narrow banded frequency spectra [Gray RA et al. 1998]. The frequency signal is relatively constant and comprised of a narrow frequency band which has lead others to suggest that this signifies some form of organisational activity or even chaos theory [Goldberger AL 1996].

Fibrillation frequency and drugs

Investigators have looked at the effects of various anti-arrhythmic agents on ventricular fibrillation frequency. Lignocaine in dogs has been shown to reduce the high frequencies seen [Carlisle EJ et al. 1990]. Stewart showed that Class I antiarrhythmic agents, namely lignocaine, mexiliteine and disopyramide significantly reduced the dominant frequency of the power spectrum in a pig model [Stewart AJ et al. 1996]. The Class I agents are involved with the voltage dependent fast sodium channels and it would seem that these channels contribute to the higher frequency component of the power spectrum. Calcium Blockers such as Verapamil, which act upon the slow-inward calcium channel were also studied. Verapamil was seen to prevent the usual fall and decline in the ventricular fibrillation frequency after 1-2 minutes of ventricular fibrillation in dogs [Carlisle EJ et al. 1990]. Verapamil was also noted to actually increase the ventricular fibrillation dominant frequencies in a pig model and it has been suggested that the initial fall in frequency and leftward shift of the power spectrum may be due to the blockade of inward L-type calcium channels [Stewart AJ et al. 1996]. However, recent work has found that no significant change in peak ventricular fibrillation frequency was seen with Verapamil throughout the arrhythmia in isolated feline hearts [Amitzur G et al. 2000].


In summary, ventricular fibrillation is a sudden lethal arrhythmia responsible for many deaths in the western world mostly in the setting of ischaemic heart disease. Despite much work, the underlying nature of fibrillation is incompletely understood. Most episodes of fibrillation occur in diseased hearts, however others occur in so-called normal hearts. Theories of ventricular fibrillation generation and maintenance must account for both. The multiple wandering wavelet theory and the further modifications to these discussed above are not mutually exclusive. Most studies showing wandering wavelets are focused on a small area of myocardium and are restricted to surface measurements and more or less show the micro-electrophysiological conditions. The theories proposed by Winfree and Panfilov are based on experimental observations and more so on computer modelling and describe ventricular fibrillation more on a macro-electrophysiological level. Evidently, a better understanding of the causes of ventricular fibrillation initiation and maintenance can only help in the struggle to find better therapeutic strategies to both prevent the onset of ventricular fibrillation and to end the ongoing process by immediate defibrillation.

See also:

Last updated: 02-07-2005 00:03:12
Last updated: 05-03-2005 17:50:55