Biomedical subjects

J C Jarvis

Publications and source records attributed to J C Jarvis.

At least 37 records · Page 2Linked to original sources

PubMed1998

Nitroglycerine reduces neutrophil activation and acute damage in latissimus dorsi muscle grafts.

BACKGROUND: Damage to the latissimus dorsi muscle (LDM) may jeopardize a successful outcome to dynamic cardiomyoplasty. We and others have demonstrated muscle damage in LDM in various species including humans. Ischemia is now recognized to be an important contributory factor. We postulated that glyceryl trinitrate, a nitric oxide donor, might protect against ischemic endothelial dysfunction and so reduce resultant muscle damage. METHODS: In 20 adult rats the left LDM was mobilized on its thoracodorsal neurovascular pedicle and maintained as an orthotopic graft. Half of the animals received glycerol trinitrate intraoperatively and postoperatively for 24 hours. The other half served as untreated controls. Each group was further subdivided into two groups (n = 5 in each): animals in which the LDM was excised after 4 hours for myeloperoxidase studies, and animals in which the LDM was excised at 24 hours for analysis of muscle damage by histology and enzyme macrohistochemistry. Blood samples were taken at 24 hours for assay of plasma nitrite and nitrate as nitric oxide metabolites. RESULTS: Glycerol trinitrate-treated animals had higher plasma nitric oxide metabolite levels after 24 hours (after nitrate reductase treatment, total nitrite, 78.3+/-11.8 nmol/mL, mean +/- SEM) than controls (42.1+/-3.7 nmol/ mL, p = 0.008). The proportion of viable LDM in glycerol trinitrate-treated animals was greater than in untreated animals, mainly in the middle and distal regions of the graft (middle region, 96.3%+/-0.5% versus 75.7%+/-4.1%, p<0.001; distal region, 94.4%+/-0.8% versus 40.9%+/-3.1%, p<0.001). Macrohistochemical findings correlated well with the histologic findings. Myeloperoxidase activity (U/g) was markedly lower in glycerol trinitrate-treated LDMs, mainly in the distal part of the graft (glycerol trinitrate versus control, 20.5+/-2.1 versus 40.9+/-3.1 U/g, p<0.001). CONCLUSIONS: Glycerol trinitrate significantly reduced acute damage to the distal two-thirds of the mobilized LDM, possibly by modifying leukocyte activation and endothelial dysfunction associated with ischemic injury.

Animals↗

PubMed1998

Observation and basis of improved blood flow to the distal latissimus dorsi muscle: a case for electrical stimulation prior to grafting.

OBJECTIVE: Dynamic cardiomyoplasty, using a functional graft of the latissimus dorsi muscle, has shown promise as a treatment for selected patients with advanced heart failure. The success of this approach depends on maintaining the viability of the muscle, whose distal portion is susceptible to ischaemic damage. We investigated the effects of surgical mobilization on regional muscle blood flow and the influence of electrical stimulation of the muscle. METHODS: Ten sheep were randomly assigned to two equal groups. In one group, the latissimus dorsi muscle was stimulated continuously in situ at 2 Hz for two weeks; in the other group, the muscle was not stimulated. Regional blood flows in the muscles were determined by a fluorescent microsphere technique. Serial measurements were made (a) under baseline conditions before intervention, (b) with the thoracodorsal artery occluded and (c) after interruption of the perforating collateral arteries. RESULTS: Surgical mobilization of the unstimulated latissimus dorsi muscles had little effect on blood flow in the proximal region, which remained at 93.1 +/- 16.9% of baseline (mean +/- SEM). The distal region was rendered significantly more ischaemic (55.8 +/- 13.5% of baseline, p < 0.002 compared to the proximal region). Electrical prestimulation abolished any significant proximodistal gradient in blood flow and improved distal muscle perfusion following mobilization (proximal vs. distal: 75.0 +/- 8.8 vs. 63.0 +/- 10.9%; p > 0.4). CONCLUSIONS: Distal muscle ischaemia occurred when the entire latissimus dorsi muscle was acutely elevated on the thoracodorsal pedicle alone. Electrical prestimulation of the muscle in situ improved the thoracodorsal perfusion of the distal muscle by abolishing the proximal-to-distal gradient in flow, with a substantial benefit to distal flow after mobilization. Although electrical stimulation is known to induce vascular proliferation, we argue that this effect of stimulation is brought about mainly by enhancement of the flow through anastomotic connections between proximal and distal arterial territories.

Animals↗

PubMed1998

Muscle transformation in cardiomyoplasty: the effect of conditioning and mobilisation on perfusion, oxygenation and fatigue resistance in the latissimus dorsi muscle.

BACKGROUND: In the clinical application of transformed skeletal muscle to cardiac assistance there is evidence that the latissimus dorsi muscle (LDM) wrap can undergo atrophy, which would prevent it from providing a sustained functional improvement. Possible causes are ischaemia and degeneration related to the conditioning process. We studied the nutritional and structural changes occurring under different stimulation regimes with the aim of improving the conditioning protocol. METHODS: Microelectrodes were used to measure regional perfusion and oxygenation in the rabbit LDM during mobilisation and subsequent repeated contraction. Group A muscles (n = 10) were conditioned for 6 weeks at 10 Hz, Group B muscles (n = 10) for 2 weeks at 2.5 Hz. Each muscle was then mobilised and tested in a hydraulic apparatus which recorded the pressure generated in a closed circuit. RESULTS: Muscles of Group A and Group B demonstrated transformation of fibre type, with a predominance of type I (62 +/- 4%) fibres in Group A and type IIa (68 +/- 9%) fibres in Group B. There was no evidence of muscle degeneration. After 10 min of fatigue testing the pressure produced was 53 +/- 5% of initial values in Group A and 51 +/- 8% in Group B, compared to 8 +/- 1% in the control group (P < 0.001). Maximum rate of relaxation was faster in Group B than in Group A (46 +/- 3% vs. 36 +/- 3% of control muscle, P < 0.05). Mobilisation resulted in a decrease in the distal perfusion of the control muscles (P < 0.05) and PO2 decreased by 8.7 +/- 1.7 mmHg during a fatigue test, which resulted in rapid loss of contractile function to 46 +/- 1% of the initial value within 1 min. In both Groups A and B the perfusion of all regions of the muscles both before and after mobilisation was greater than that of controls. During the same fatigue test, the PO2 of the distal regions was maintained and the contractile function fell more slowly to between 70 and 80% of initial values within 1 min. CONCLUSION: We showed that ischaemia in the distal region of the control LDM could result from mobilisation and repeated contraction. Muscle transformation improved perfusion and prevented a fall in tissue PO2 during a sustained series of contractions. Muscles that were conditioned at 2.5 Hz shared the improved perfusion of the fully transformed muscle, but had faster relaxation characteristics. Short periods of in situ conditioning prior to mobilisation may help to avoid ischaemic changes in distal parts of the LDM while achieving fatigue resistance in the grafted muscle at an earlier postoperative stage.

Animals↗

PubMed1998

Morphological and functional evidence, and clinical importance, of vascular anastomoses in the latissimus dorsi muscle of the sheep.

Mobilisation of the latissimus dorsi muscle as a functional graft necessarily involves division of perforating arteries that enter the distal portion of the muscle, rendering it vulnerable to ischaemic damage when the muscle is stimulated electrically. Using a fluorescent microsphere technique we showed that the blood flow contributed by the thoracodorsal artery decreases in a proximal-to-distal direction, and that of the perforating arteries in a distal-to-proximal direction, but for neither does the flow decline to zero. This is consistent with earlier reports of anastomotic connections between the 2 arterial territories. We went on to use fluorescence microscopy to demonstrate the existence of these vascular anastomoses, the first such evidence obtained under physiological conditions of pressure and flow. In clinical applications, the existence of anastomotic connections offers the prospect of maintaining flow to the distal part of the grafted muscle without the delays inherent in neovascularisation procedures.

Animals↗

PubMed1997

Relation between muscle contraction speed and hydraulic performance in skeletal muscle ventricles.

BACKGROUND: The fatigue resistance and power-to-weight ratio of skeletal muscle that has been conditioned by electrical stimulation makes cardiac assistance from a graft of such muscle a realistic prospect. A skeletal muscle must be surgically reconfigured to act on the circulating blood, but little is known about the power losses that accompany such interventions. We investigated in acute experiments the hydraulic performance of approximately cylindrical pumps made from sheep latissimus dorsi (LD) muscles, having first characterized the performance of each muscle in situ. METHODS AND RESULTS: Force-length and force-velocity relations were measured in situ for LD that had received either 8 weeks of stimulation at 2 Hz or no chronic stimulation. Two sizes of skeletal muscle ventricle (SMV) were formed from the same muscles, and their hydraulic performance was measured. The hydraulic performance was also calculated from the linear data, models of the force-length and force-velocity curves, and a description of the stress distribution within the SMV wall. The model predicted well the isovolumetric function of the ventricles and the optimum afterload but overestimated the flow and therefore the power. In conditioned ventricles the performance was particularly poor because of the slow contractile properties of the muscles. CONCLUSIONS: If SMVs are to pump effectively against the arterial impedance, the pressure drop caused by flow (or the internal resistance) should be lower than that of the ventricles we constructed. Progress can be made through refinement of surgical technique and stimulation protocols that generate faster fatigue-resistant muscles.

Animals↗

PubMed1997

Intramural blood flow of skeletal muscle ventricles functioning as aortic counterpulsators.

BACKGROUND: Skeletal muscle ventricles (SMVs) working as aortic counterpulsators have provided long-term left ventricular assistance under experimental conditions. However, gradual deterioration of SMV pump function and rupture have been observed, and this may be related to compromised intramural blood flow during synchronized counterpulsation under systemic working conditions. METHODS: Transformed, double-layered SMVs in 6 sheep were stimulated for 3-minute periods (5 V, 30 Hz, burst duration and delay from QRS both 40% of the cardiac cycle) to work as diastolic counterpulsators in the systemic circulation at a 1:2 (SMV:heart) and 1:1 ratio, and on a mock circulation with low-pressure loading conditions at a 1:2 ratio. Thoracodorsal artery blood flow was monitored by ultrasonic flow probe, and intramural blood flow distribution was investigated by fluorescent microspheres. Thoracodorsal venous lactate concentrations were measured before and after each period of stimulation. RESULTS: Thoracodorsal artery blood flow increased significantly (p < 0.001) after stimulation. The magnitude of augmentation (89%; 95% confidence interval, 36% to 163%) was similar for all working conditions studied. Reactive hyperemia was observed after most 1:1 regimens but was rare after 1:2 regimens. A significant (p < 0.05) 15% increase in serum lactate levels was present after 1:1 regimens only. All regimens of stimulation resulted in a significant increase (p < 0.01) in blood flow to sections in the outer wall of the SMV, but a significant increase (p < 0.05) in blood flow to sections in the inner wall was observed only under low loading conditions. CONCLUSIONS: Skeletal muscle ventricles subjected to 1:1 systemic counterpulsation regimens work under partly anaerobic conditions. High loading conditions may compromise SMV inner wall blood flow.

Animals↗

PubMed1997

Flow in a simple model skeletal muscle ventricle: comparison between numerical and physical simulations.

Flow patterns generated during ventricular filling have been investigated for three different combinations of flow rate and injection volume. The numerical solutions from a commercially available computational fluid dynamics package were compared with observations made under identical flow conditions in a physical model for the purpose of code validation. Particle pathlines were generated from the numerical velocity data and compared with corresponding flow-visualization pictures. A vortex formed at the inlet to the ventricle in both cases: During the filling phase, the vortex expanded and traveled toward the apex of the ventricle until, at the end of filling, the vortex occupied the full radial extent of the ventricle; the vortex continued to travel once the filling process had ended. The vortices in vitro were more circular in shape and occupied a smaller volume than those generated by the numerical model. Nevertheless, comparison of the trajectories of the vortex centres showed that there was good agreement for the three conditions studied. Postprocessing of velocity data from the numerical solution yielded wall shear-stress measurements and particle pathlines that clearly illustrate the mass-transport qualities of the traveling vortex structure. For the cases considered here, the vortex transit produced a time-dependent shear stress distribution that had a peak value of 20 dynes cm-2, with substantially lower levels of shear stress in those regions not reached by the traveling vortex. We suggest that vortex formation and travel could reduce the residence time of fluid within a skeletal muscle ventricle, provided that the vortex travels the complete length of the ventricle before fluid is ejected at the start of the next cycle.

Blood Flow Velocity↗

PubMed1997

High-energy phosphates and tension production in rabbit tibialis anterior/extensor digitorum longus muscles.

The effects of repetitive muscle contraction on energy state and tension production were studied in rabbit tibialis anterior/extensor digitorum longus muscles that had been subjected to 90 days of continuous indirect electrical stimulation at 10 Hz. Anesthetized chronically stimulated and control rabbits were challenged with 15 min of stimulation at 4 and 15 tetani/min. Pi-to-phosphocreatine (PCr) ratio (Pi/PCr) was measured in vivo before, during, and after acute stimulation by 31P-magnetic resonance spectroscopy, and tension was recorded at the same time. Although Pi/PCr was low at rest, it was significantly higher in chronically stimulated muscle than in control muscle (0.20 +/- 0.02 vs. 0.05 +/- 0.01, P < 0.05). Stimulation of control muscle for 15 min at both 4 and 15 tetani/min induced a significant rise in Pi/PCr, whereas the same conditions in chronically stimulated muscle did not produce any significant departure from initial levels. The tension produced by control muscle fell to 93 +/- 3% of its initial value during stimulation at 4 tetani/min and to 61 +/- 7% at 15 tetani/min, respectively. In chronically stimulated muscle, on the other hand, tension was potentiated above its initial level at both stimulation rates (135 +/- 15 and 138 +/- 11%, respectively) and remained significantly elevated throughout each trial. The ability of chronically stimulated muscle to sustain high levels of activity with minimal perturbations in Pi/PCr or decrement in tension is attributable to cellular adaptations that include a well-documented increase in oxidative capacity.

Animals↗

PubMed1996

Fast-to-slow transformation in stimulated rat muscle.

Several previous studies have failed to demonstrate changes due to chronic stimulation in contractile speed of innervated fast rat muscles, and it has been suggested that the adaptive capacity of skeletal muscle in this species is limited. We have reassessed this contention. Fast muscles of the rat hind limb were stimulated continuously at 10 or 20 Hz for 55-61 days. The maximum shortening velocity of the extensor digitorum longus muscles was reduced to 50% of the control value. The proportion of type 1 fibers increased from 4% in control muscle to 34% in stimulated muscles and there was a corresponding reduction in type 2B/D fibers. The proportion of type 2A fibers after stimulation was similar to that in control muscles. These results, taken together with our published analyses of myosin isoform composition of these muscles, show that the mechanisms that control gene expression in response to activity are not exclusive to larger mammals.

Animals↗

PubMed1996

A novel video technique for visualizing flow structures in cardiovascular models.

We describe a video system that produces good quality images of particle trajectories in seeded fluid flows. The operation of a liquid crystal optical shutter and a modified charge-coupled device (CCD) camera were synchronized to generate images of particle trajectories which were stored in a framegrabber before being transferred to S-VHS tape. The camera system is particularly appropriate for visualizing transient or unsteady flows in models of the cardiovascular system as the integration time may be varied to produce particle trajectories of variable length.

Cardiac Output↗

PubMed1996

Formation and travel of vortices in model ventricles: application to the design of skeletal muscle ventricles.

Vortex-ring production was studied in axisymmetric elastomeric ventricles designed to stimulate flow in a cardiovascular assist device. A flow visualization technique was used to investigate the effects of reducing the inlet diameter and predilating the ventricle on vortex travel in two ventricles of different shape and size. In most cases, vortex rings formed during the filling phase. They were bounded by the incoming jet of fluid and the ventricular wall. The velocity of their centres during the filling period was proportional to the inflow velocity. During filling, vortex velocity was substantially independent of the shape and diameter of the two ventricles studied. It was dependent mainly on orifice diameter: a narrower inlet led to greater inflow velocities and proportionately greater vortex velocities. At the end of the filling phase, each vortex increased in size to occupy the full radial extent of the ventricle. This process was associated with a decrease in the axial velocity and strength of the vortex. At low flow rates, these losses resulted in the arrest of the vortex at end filling. Vortex motion in ventricles is particularly important in the design of a cardiovascular device such as the skeletal muscle ventricle (SMV), where small ejection fractions may leave blood at the apex of the ventricle relatively undisturbed. It is suggested that inlet diameter could be selected to favour the formation and travel of vortices, with a resultant reduction in apical residence time and hence a reduced risk of thrombus formation.

Coronary Circulation↗

PubMed1996

Factors influencing vortex development in a model of a skeletal muscle ventricle.

A video technique was used to visualize flow patterns in a pulsatile model of a skeletal muscle ventricle (SMV). The shape and duration of SMV filling/emptying curves were determined by the stroke action of a computer-controlled piston pump. In this way we examined the effect of filling duration and injection volume on the flow structures. To simulate the interaction between the left ventricle and the SMV in the proposed clinical application, two filling/emptying regimes were employed with the period of each SMV cycle being either equal to or twice that of a typical cardiac cycle. Vortex formation at the inlet to the ventricle was a feature of both flow cycles during filling for all the combinations of injection volume and filling duration considered. When a pacing ratio of 1:1 was simulated, SMV filling was immediately followed by ejection, and the formed vortex was expelled. Under this flow regime, fluid in the apical end of the ventricle remained undisturbed. When the pacing ratio was 2:1, however, the quiescent period that followed filling allowed time for the formed vortex to travel a significant distance toward the apical end of the ventricle. Under certain conditions of injection volume and flow rate, the vortex reached the apex by the end of the quiescent period. We suggest that a ventricle connected to the aorta by a single, valveless conduit would require a flow regime that included a significant period of quiescence after filling. This would allow the vortex to reach and displace blood that would otherwise stagnate in the apex, thereby reducing the risk of thrombus formation.

Biomechanical Phenomena↗

PubMed1996

The use of coloured dye-extraction microspheres to measure blood flow in rabbit skeletal muscle: a validation study with special emphasis on repeated measurements.

We evaluated the use of coloured dye-extraction microspheres for measuring blood flow in the skeletal muscle of anaesthetized rabbits. Spheres were injected into the left atrium after sternotomy or lateral thoracotomy, or into the left ventricle via the left carotid artery. Both routes of administration produced adequate mixing of microspheres with the blood. In general, the values of flow obtained from spheres of different colour were well correlated. Resting flows were difficult to measure, owing to the low number of spheres and the correspondingly low absorption of the samples. Occlusion of the capillary bed by trapped spheres was less than 10% in both tibialis anterior and extensor digitorum longus muscles after injection of a total of 15 x 10(6) spheres. Injection via the left ventricle, a closed-chest procedure, gave a more stable preparation than injections via the left atrium. However, in both cases there was a progressive decline in hyperaemic muscle blood flow during a series of injections. These results indicate that it is possible to use this technique to measure skeletal muscle blood flow in the rabbit, but that it is difficult to exploit the advantages of sequential microsphere injection in this species.

Animals↗

PubMed1996

Induction of a fast-oxidative phenotype by chronic muscle stimulation: mechanical and biochemical studies.

We studied changes in the mechanical properties and myosin isoform composition of rabbit tibialis anterior muscles that were subjected to continuous stimulation at 2.5 Hz for up to 12 wk. The effects of stimulation at 2.5 Hz were less profound than those observed for the same duration of stimulation at 10 Hz (12). Stimulation at 10 Hz for 12 wk induced complete transformation to a slow-contracting muscle homogeneous in slow myosin isoforms; stimulation for the same period at 2.5 Hz resulted in moderate changes in contractile speed and a very small increase in the synthesis of slow myosin isoforms. On the other hand, the fatigue resistance of muscles stimulated at 2.5 Hz was as great, in both isometric and dynamic fatigue tests, as that of the muscles stimulated at 10 Hz. Thus entire fast skeletal muscles can be transformed to a state in which fast myosin isoforms continue to be synthesized, but the oxidative capacity is sufficient to support sustained working at a higher power output than that associated with slow muscle.

Animals↗

PubMed1996

Induction of a fast-oxidative phenotype by chronic muscle stimulation: histochemical and metabolic studies.

Chronic electrical stimulation of skeletal muscle at 10 Hz induces fast-to-slow fiber type transformation. Does a lower aggregate amount of activity lead to a less complete transformation, or does it produce the same transformation over a longer time course? We examined this question by subjecting adult rabbit tibialis anterior and extensor digitorum longus muscles to continuous stimulation at 2.5 Hz for 2-12 wk. Most of the fibers acquired the histochemical and immunocytochemical characteristics of type 2A, not type 1, fibers. There was a corresponding rise in oxidative activity, but this was accompanied by a marked decline in anaerobic glycolysis. The activities of hexokinase and 3-oxoacid CoA-transferase stopped increasing after 2 wk, glutamate oxaloacetate transaminase after 4 wk, and beta-hydroxyacyl-CoA dehydrogenase after 6 wk of stimulation. Succinate dehydrogenase, citrate synthase, lactate dehydrogenase, and creatine phosphokinase continued to change up to 12 wk of stimulation. Changes in enzyme activity were not as rapid or as marked as those observed for stimulation at 10 Hz, and none showed the typical two-phase response of oxidative enzyme activities to stimulation at 10 Hz. The latter may therefore be dependent on induction of type 1 myosin isoforms.

Anaerobiosis↗

PubMed1996

Changes in ATP, phosphocreatine, and 16 metabolites in muscle stimulated for up to 96 hours.

Rabbit tibialis anterior muscles were stimulated continuously at 10 Hz for periods ranging from 2 min to 96 h and were analyzed for energy reserves and metabolic intermediates. Glycogen, ATP and phosphocreatine fell rapidly during the first 5 min of stimulation. Glycogen continued to fall to very low levels, whereas ATP and phosphocreatine rose, reaching 70% of control by 1 h, despite ongoing stimulation. After 2 h, glycogen also increased, regaining control levels in 4 days. Glucose rose to 4.5 times control in 30 min and still exceeded 2.5 times control at 24 h. In the first 2 min, glycolytic intermediates, glucose 6-phosphate (G-6-P), fructose 1,6-bisphosphate, lactate, and pyruvate more than doubled and then returned to control levels or below. Malate and 3-glycerophosphate rose 600 and 200%, respectively. Both of these compounds participate in shuttling reducing equivalents from cytoplasm into mitochondria. Citrate and alpha-ketoglutarate underwent much more modest changes. Glucose 1,6-bisphosphate (G-1,6-P2) fell to one-third of control by 2 h and then rose dramatically at 4 h. At 4 days it was still twice control. The 6-phosphogluconate (6PG) doubled at 2 min, then rose to 12 times control at 2 h, fell somewhat, and peaked at 16 times control at 24 h. Aspartate and alanine both exhibited a biphasic rise in concentration, whereas glutamate fell to 30% in 15 min and rose slowly after 4 h. The rise in glucose was interpreted to be the consequence of rapid glycogenolysis together with inhibition of hexokinase by G-1,6-P2 and elevated G-6-P. Paradoxically, glycogen resynthesis apparently occurred when the glycogen synthase stimulator, G-6-P, was very low, and the glycolysis stimulator, G-1,6-P2, was high. Although G-1,6-P2 is an inhibitor of 6PG dehydrogenase, the timing of the changes in G-1,6-P2 and 6PG levels suggests that the accumulation of 6PG was initiated by some other influence.

Adenosine Triphosphate↗

PubMed1995

The input-output relations of skeletal muscle.

We used three approaches to determine the stimulation patterns that maximize the isometric force-time integral per impulse (FTIpP) available from tibialis anterior muscles of the rabbit. Initially the interval between two pulses was fixed at the value that gave the maximum force-time integral, and successive pulses were added at intervals that maximized the FTIpP. We checked this iterative approach by a second method, in which a computer-generated protocol was used to deliver randomized bursts to the muscles. These experiments confirmed that optimal stimulation patterns for fast muscles consisted of an initial high-frequency portion followed by a train of impulses at a lower frequency. However, for muscles that had been stimulated chronically at a constant low frequency, an initial high-frequency portion conferred no advantage. In a third set of experiments we used constant-frequency bursts to generate contour surfaces that represented the dependence of FTIpP on the frequency and number of impulses. The results agreed with those from the earlier methods. We conclude that optimized patterns have potential for clinical use, but their value will depend strongly on the activation characteristics of the stimulated muscle.

Animals↗

PubMed1995

A radial basis function model of muscle stimulated with irregular inter-pulse intervals.

Paralysed muscle, or skeletal muscle which is to be used for cardiac assistance, may be given an artificial function if it is electrically stimulated to contract and the response can be adequately controlled. To design a controller, a model of the muscle or system is usually required. The most commonly used models are analogues, originating from A.V. Hill's model. However muscles exhibit many nonlinear and time-varying phenomena which, if they are to be modelled, make the analogue complex and cumbrous to work with. The system may further be complicated by pathological changes and secondary effects of stimulation. We propose that such a system can be modelled by nonlinear networks ('neural networks'). The radial basis function network (RBF) has two advantages over the better-known multi-layer perceptron (MLP). We describe the use of an RBF network to model rabbit muscle that is supramaximally stimulated at irregular inter-pulse intervals.

Animals↗