| The Self-Regulating Double-Pump: A Working Model of the Human Heart |
 |
|
|
| | Clear Image Window | |
The findings in chapter 6 initiated the design of a heart model consisting of two pumps in series working with a common displaceable wall (cf. the ventricular septum of the human heart). To demonstrate that a double-pump can work according to the basic balancing principle established for the human heart in situ (Concept 2, see Chapter 2), a self-regulating prototype was built [108]. It had the inlet and outlet of the respective pump halves connected in the same way, that the human heart fits into the circulation system.
The pump is shown in Fig. 7-1 and Fig. 7-2, and its design principles in Fig. 7-3. Its working mode will be explained with reference to Fig. 7-3. Numbers in the text (identifying individual pump parts) refer to Fig. 7-2 and/orFig. 7-3.
The pump is mounted in a rigid outer casing 1. A partition-wall S (corresponding to the ventricular septum) is arranged in the casing. S has semi-flexible, membrane-like sections 2 and 3. The spacing (septal volume, SV) can be increased with the aid of some external means e.g., by compressed air or in this case by injection of water.Injection of water is effected by a piston driven by Pressurised Air Cylinder, PAC (71). The piston, consisting of a membrane with a rigid centre part, has a constant stroke length. It displace a certain amount of water back and forth between the activation pressure transducer (73) and SV. The total system displace water back and forth to SV (septal volume), RV (right volume) and LV (left volume), is called PAC-SV, PAC-RV and PAC-LV.
The wall structure S divides the space within the casing into two chambers, L and R (corresponding to the left and right ventricles). each chamber is provided with an inflow flap valve, MV (corresponding to the mitral valve) and TV (corresponding to the tricuspid valve), and an outflow flap valve, AV(corresponding to the aortic valve) and PV (corresponding to the pulmonary valve).
When the pressurised air cylinder PAC-SV connected to the spacing between membranes 2 and 3 is activated, they move apart laterally by the force of the injected water. The rising of pressure in chambers L and R will cause the inflow valves MV and TV to close and the outflow valves AV and PV to open.
If the prevailing external pressures of the outflow valves are different, one outflow valve will open before the other, and the corresponding chamber will begin to empty before the other.
The other chamber will not start emptying until either the outlet pressures become equal, or the respective flexible membrane (6) has been extended to the limit of its movement. (That means e.g., by mechanical restriction between membranes 2 and 3 or externally thereof).
When water in SV is removed, valves AV and PV will close, provided there is a generally higher pressure maintained at the outflow side than at the inflow side; valves MV and TV will open, whereupon chambers L and R will take in more fluid.
If the prevailing pressures at the inflow side of the pump are different, membranes 2 and 3 will move towards the chamber with the lower inflow pressure. It will thus decrease in volume.
Provided that the pre-set ranges of movement are the same for both membranes, the membrane at the chamber with the smaller volume will thus be closer to the limit of its movement in the direction of that chamber. At the next activation of PAC-SV, that membrane will be displaced until its expansion limit, provided enough water is injected. Further expansion of volume SV can only be achieved by displacement of the other membrane. In such a situation, more fluid will be expelled from the chamber with the initially larger volume, no matter what the prevailing pressures are at the outlets of chambers R and L.
When the double-pump is connected to a circulatory system in which the same fluid flows in two circuits i.e., a system corresponding to circulation of blood in humans, the filling mechanism will automatically achieve a balance of the volume displacements effected.
For example, should one chamber pump out "too much" fluid, this "surplus" will be returned to the fluid flowing to the other chamber, so in order to compensate. This means that a balance in volume is achieved without any complicated regulating mechanisms.
The pump effect can also be supplemented by providing each of the ventricle-simulating chambers with additional laterally movable wall portions (membranes). These are named 12 and 13, and are arranged by highly flexible membranes 7 attached to the casing 1. Volumes LV and RV between membranes 12 and 13 and the casing are connected to pressurised air cylinders PAC-LV and PAC-RV. They can be activated individually, to inject a specific volume of water as described for PAC-SV.
The expansion of the volumes delimited by membranes 12 and 13 towards L and R, respectively, will not have any effect on the regulating function of the pump, but will only contribute to its total stroke volume.
In the experimental model, chamber L corresponds to the left ventricle in man, and has thus to work against a higher pressure than chamber R. A copper gauze 61 bulging towards R restricts the movement of membrane 3 towards chamber R.
Membranes 2, 3, 12 and 13 have been made by pouring fast-setting silicone resin into appropriate moulds, in order to obtain the flexible quality of the membrane 6. Each pressure-transmitting chamber, SV, LV and RV, has been arranged to displace 50 cm3 fluid maximally. This means that in a situation where left and right inflow pressures are identical, efflux from each of the chambers L and R will be 75 cm3.
The experimental model has also been provided with a means of preventing the complete closure of each of the valves, for simulation of various types of valve insufficiency. It have also been provided with a means for restriction of the opening of these valves for the purpose of simulating various types of stenosis.
Water was used as circulating fluid in all experiments.
Casing 1 is made from polymethyl methacrylate, which is transparent. It allows the observation of membranes and valves in operation.
In the first experiments with the double-pump, it has been seen that pulsating inflow into chambers L and R creates two problems. Added to this, is the working mode of the pressurised air cylinders activating the pump. The problems and their solutions are explained below.
Pulsating inflow demands that, each time the chambers are to be filled, the incoming fluid has to be accelerated. In the event of the circulation not being in a state of equilibrium, the fluid that is supposed to fill the larger chamber will need excess force or time for acceleration. The only force effecting the filling of chambers L and R is the static inflow (filling) pressure. Therefore filling time has to be of a maximum length, to secure the filling of the larger volume (in case the pump is not in equilibrium). That means that the pumping rate must not be set too high.
It is important to reduce the mass of fluid which had to be accelerated at the inflow side. Therefore the flexible afferent tubes to the pump chambers were made of equal length, and as short as possible.
But this was not sufficient. Rather high filling pressures were still required. Moreover, the closing of valves MV and TV gave rise to sharp pressure pulses, followed by resonating phenomena. This situation did not allow determination of the regulating function of the pump to be carried out with sufficient precision.
The problem was eliminated by the addition of first expansion vessels 21 and 22 in connection with the inflow tract. The goal was to reduce the need for high filling pressures and, to a large extent, eliminate pressure transients. In the human body, this is naturally provided by the atria and the venous system.
Volume activation by pressurised air from cylinders (PAC) created another problem.
Increased outflow from one of the volumes L or R, can be caused of the regulating function of the intermediate wall. That results in a higher outlet pressure at a given peripheral resistance, which necessitates a higher air pressure, and thus a greater amount of compressed air.
In diastole this compressed air volume must leave the system, to allow fluid to enter LV and RV simultaneously.
If means for controlled recession of the walls are not provided, the regulating function of S will be severely impaired. The incoming fluid will preferably rush into the chamber that dilates most readily. A little later a somewhat higher filling pressure on the opposite side should push the intermediate wall towards the chamber that had accepted the in-flowing fluid most readily; its inflow valve will close, preventing the regulating function of the pump from becoming operative.
This problem was overcome by the addition of restricting means (not shown) to the inlets and outlets of the electromagnetic valves providing PAC-RV, PAC-SV and PAC-LV with pressurised air.
The pump was connected to circulatory loops simulating that of man (Fig. 7-1 and Fig. 7-4).
Air pressure was set at a level sufficient for rapid injection of water, via PAC-SV, PAC-LV and PAC-RV into chambers SV, LV and RV respectively.
The pumping rate was set at about 50 strokes per minute. Activating periods of the air cylinders were adapted to pressures and flows expected in the system. Additional pressure monitors (not shown) were arranged, at the bottom of the second open expansion vessels 23 and 24 for inflow pressure registration.
A Third open expansion vessel 25, was connected to vessel 23 and another third open expansion vessel 26, to vessel 24. These open third expansion vessels 25 and 26 were arranged so as to be easily raised or lowered.
Outflow tubes from chambers R and L were connected to closed fourth expansion vessels 27 and 28. Air enclosed in the fourth expansion vessels 27 and 28 imitated the Windkessel effect of the aorta and the pulmonary artery.
The outlets of the fourth expansion vessels 27 and 28 were provided with two restriction valves (not shown). The action of the restriction valves corresponds to that of the peripheral resistance and of the resistance in the pulmonary circulation respectively.
Pressure monitors 55 and 56 were arranged on top of 27 and 28, for the registration of the pressures at the outflow side.
The outlets of 27 and 28 were provided with turbo flow-rate meters 51 and 52. They had fixed tubes 81 and 82 ending above the respective aforementioned open and movable expansion vessels 25 and 26. This arrangement transforms pulsating into non-pulsating flow i.e., it assumes part of the function of the arteriolar-alveolar system (flow rate independent of pressure drop, the "waterfall" concept [130]).
Inlet-outlet connections of the two pump halves were arranged in a configuration of an eight.
Turbo flow meters 51 and 52 had considerable self-impedance, especially at higher flow rates. There was a sharp increase in flow rate in certain heart defects simulated with this pump arrangement. Giving chamber R a higher outflow than L, flow meter 51 on the right outflow side was shunted (via tube 83) to prevent damage to membranes 2 and 3. The turbo flow meter 51 thus showed reduced flow rates compared with those measured on the left side, 52. This had to be taken into account when analysing the results given in the diagrams.
A corresponding shunt 84 (to be opened in certain experiments) was provided in the left circulation loop.
Results are presented in the form of eight electrostatic-recorder traces.
The variables recorded are:
| SV Activated; LV and RV idle |
| SV, LV and RV activated |
The regulating action of the pump solely by partition-wall S was tested, with lateral pressurising volumes LV and RV in an idle state. By raising or lowering the open expansion vessels 25 and 26, filling pressures on the right and left side were set.
An increase in filling pressure on the right side, carries with it an incremental increase in volume at the same side. By the balancing action of the wall S, the stroke volume and outflow pressure on the right side was increased (Fig. 7-5).
The stroke volume and outflow pressure on the left side decreased accordingly.
A few strokes later, balance had been attained again and flow rates and flow pressures were the same as before the change in inflow pressure at the right side. Filling pressure coincided, but at a slightly increased level, due to the fluid volume from the expansion vessel 25, which had been added to the system.
Decrease of filling pressure on the right side resulted in an opposite reaction (this is not clearly shown, since flow meter 51 (Fig. 7-4) was by-passed).
Corresponding reactions were elicited by raising and lowering filling pressure on the left side.
Experimental procedure was as above, but with lateral pressurising volumes LV and RV in working order.
The results (Fig. 7-6) with respect to balancing are the same, but recorded pressures and flow are consequently higher.
Keeping LV idle compares with the situation in a lateral left ventricular infarction.
When starting from the condition in the previous experiment (and by disconnecting LV), the volume was reduced immediately by 50 cm3 (Fig 7-7); it had been 75 cm3 for each of the pump halves in a position of balance. This is seen as a decrease in pressure and flow from chamber L, and an increase of its filling pressure. Partition-wall S deviated towards chamber R already in systole due to this increased filling pressure and thereby lent its total stroke volume, 50 cm3 to the left chamber. Correspondingly, right chamber stroke decreased by 25 cm3. Because of the equilibrating output function of partition-wall S, the system could thus still be kept in balance with an increase in filling pressure on the left side.
Chamber R had its filling pressure somewhat reduced, by transfer of fluid to expansion vessels 22, 24 and 26.
A loss of two thirds of pumping capacity at chamber L, thus resulted in an outflow decrease from that chamber by only one third. The partition-wall was still able to compensate for an increased filling pressure on the right side, but not for a corresponding increase on the left side. The reason for this is, that the maximal regulation capacity of the pump had been fully utilised. The maximal regulation capacity is given by the volume of expansion of the wall S, and the position of copper gauze 61.
The human heart stabilises the ventricular septum in systole by its activated network of muscle fibres, which corresponds to the copper gauze in the pump. This means that ventricular septum, in the case of a lateral left ventricular infarction, can bulge over towards the right ventricle in accordance with left-side filling pressure increasing. That activity can thereby raise the left ventricle systolic stroke volume.
Right ventricle stroke volume will be reduced correspondingly.
The net result is a decrease in total pumping capacity for the damaged heart. Balancing capacity will be preserved through a raised left ventricular filling pressure.
Keeping the right lateral volume idle, compares with the situation of an anterior right ventricular infarction (Fig. 7-8).
The experiment started under "normal" conditions. That is, with the partition-wall and both lateral walls RV and LV activated.
RV was then disconnected. After deactivating RV, flow and pressure at the outflow side corresponded to that in the preceding experiment.
Some consideration had to be given to the less than ideal configuration of silicone membranes 2 and 3, which showed some resistance in flap-over. Increase in filling pressure on the right side caused the intermediate wall S to bulge into chamber L. This caused the membrane pump-like action of S to compensate for the deficit on the right side, to which even the lateral wall LV contributed for a short period. A prerequisite for this compensation of the pumping to be effective, is that the pressure at the outlet of L is higher than at the outlet of R.
On increasing filling pressure by raising 25, the regulation by S, brings the pump into balance after a few beats. This can be seen by the fact that flow and outgoing pressure are the same as they were before the raising of 25.
The filling pressure on both sides is high because, of the transfer of fluid to the system.
A relative increase in filling pressure on the left side, achieved by lowering vessel 25, implies that the partition-wall in the passive pumping phase deviates towards R. The stroke volume in R is reduced, until balance is obtained.
The experiment shows what happens in a right ventricular infarction without the ventricular septum being damaged; the action of the septum together with the power resources of the rest of the left ventricle wall, can compensate for the loss in pumping capacity of the right ventricle [59, 90, 168].
This is again effected at the cost of total stroke volume.
It is worthy of note that traumatising RV infarction seems to be rare (cf. [84]).
Keeping the partitioning wall S volume idle, corresponds to a septal infarction.
In this arrangement, membranes 2 and 3 had only the role of dividing the interior of the pump into chambers L and R. To avoid removing copper gauze 61 (which simulates systolic stabilising properties of VS), flow meter 52 was provided with a shunt 84, identical to that at flow meter 51. That is permitting rapid change in pressure on the flux side.
The pump was started, filling pressures for both pumping chambers were equally set and recorder traces for the starting filling pressures were set to overlap. The experiment is shown in Fig. 7-9.
Shunt 83 at flow meter 51 was closed and shunt 84 at flow meter 52 opened while partitioning volume SV was simultaneously deactivated. This implies that the right pump half, had to work against a higher pressure than the left, because the flow meters had considerable flow resistance.
It was observed that the filling pressure at chamber R increased whereas it decreased correspondingly at chamber L. This is due to the fact that both expansion volumes RV and LV will through the flexible partition-volume SV expand into the chamber with the lowest output pressure; in this case it is chamber L.
After one or two beats the membranes 2 and 3 bulge over towards chamber L so much, that they will meet with membrane 12. This will result in an outflow even from chamber R.
The elasticity of the membranes 2 and 3 forced them back towards chamber R, though there was a higher filling pressure on that side. In that way chamber L could be filled.
The experiment had to be stopped when the fluid reached the full height of the open expansion vessel 21.
The outflow from chamber R and L is misleading, because of the closing and opening of the shunts 83 and 84.
The constant outflow, though increasing filling pressure at chamber R and decreasing filling pressure at chamber L, indicates that no regulation function is left.
When the experiment was stopped, by opening shunt 83 and closing shunt 84 plus activating PAC-SV, pressures and flow were rapidly normalised.
This experiment had to be repeated several times, since membranes 2 and 3 were damaged as the copper gauze 61 had no effect, when chamber R had to pump against a high pressure.
A total septal infarction, where septum does not restrict the systolic pressure of left ventricle in an active way, implies serious damage to the regulation function of the heart. That is the fact, even if a sufficient portion of its pumping capacity is preserved. As long as the pulmonary pressure, together with the passive resistance forces of the damaged muscle, cannot withstand the pressure generated in left ventricle during systole, following happens; a paradoxical movement of ventricular septum will result in an enhanced flow of blood to the pulmonary circuit, with a correspondingly rising pulmonary blood pressure propagated to the left atrium. Pulmonary edema will soon ensue, in combination with a reduced left ventricular output. Molaug et al. [118] found during experiments with intermittent ischemia of ventricular septum in anaesthetised open-chest dogs, paradoxical movement of ventricular septum, contributing to right ventricular ejection. (In these experiments, the heart is allowed to work as a displacement pump).
The pump was provided with an arrangement that allowed the insertion of a tiny wedge into the mitral valve, to prevent its complete closure. This corresponds to mitral valve insufficiency.
In this simulation (Fig. 7-10), flow and pressure at the outflow side of chamber L decreased, and the left filling pressure rose. The increased left filling pressure made the partition-wall deviate in the direction of chamber R, which thereby received a decreased stroke volume. The decreased filling pressure on the right side was partly due to fluid building up in the circulatory loop originating at the outlet of the right chamber. That was effecting the increased filling pressure of the left pump half. The higher left side filling pressure was needed, to provide for its complete filling. This was necessary in order to overcome resistance of the membrane material, due to bulging in the direction of chamber R. The filling pressure at chamber L had a tendency to rise during the simulation. That was due to the experimental insufficiency chosen happened to be a little to large; the expansion range of the volumes of partitioning wall S and the lateral walls could not cope with it. Flow from chambers L and R remained low during the simulation. When the wedge was removed, the system quickly returned to normal.
A mitral stenosis (Fig. 7-10) was similarly effected by introducing a V-shaped part, which prevented the mitral valve from fully opening. At first, flow on the left side was reduced, until the filling pressure had risen to a level compensating for the stenosis. Both pressure and flow were thereby restored on the left side. The filling pressure on the right side decreased and the right efflux pressure increased. That implied that the partition-wall effectuated a transmission of this change, in increased flow and pressure on the right side, until the system reached equilibrium again. Mitral stenosis does not reduce pumping capacity, once the pump is in balance. These results can be compared with the echocardiographic registration of a mitral stenosis (Chapter 6).
Experimental tricuspid valve insufficiency (Fig. 7-11) was achieved in the same manner as mitral insufficiency. Under these conditions, outflow from R decreased immediately. The filling pressure on the same side increased, pushing the partition-wall in the direction of the left chamber. This also caused a decrease in outflow on the left side. Left filling pressure decreased in proportion to the rising filling pressure on the right side. This means, that there was a redistribution of fluid accumulating in the open expansion vessels 21, 23 and 25. After a time, flow rates reached a new equilibrium. The slow equilibration compared with equilibration in other experiments, is due to the difficulty in producing an insufficiency smaller than membranes 2 and 3 can compensate for (50 cm3). Because of their configuration stiffness, a minimum pressure difference, provided by increasing filling pressure to chamber R, is needed to make them bend over towards chamber L. By removing wedge 67 in the tricuspid valve, the system returned quickly to normal.
An experimental tricuspid valve stenosis (Fig. 7-11) was effected as related for "mitral stenosis"; the V-shaped part is shown in Fig. 7-2. The filling pressure began to rise on the right side and to decrease on the left. The lower pressure in chamber R caused the partition-wall to move over in the direction of chamber R. Outflow from the left side increased initially. Once the pressure gradient in the filling phase had been equalised through the rise of the filling pressure on the right side, the system was in balance again. Pressure and flow on the output side was the same as before the artificial stenosis was induced. Filling pressures, however, stabilised at different levels.
Aortic valve insufficiency (Fig. 7-12) was induced by the insertion of a wedge (as above). Immediately after the experimental insufficiency had been established, forceful pulsation's, reduced pressure and flow out from chamber L were observed. With influx pressure building up on the left side, flow and pressure at the left chamber outlet rose slightly. From chamber R, there was a rather fast decrease in both outflow pressure and flow, due to the partition-wall moving toward the right side in the filling phase. The rigid position of gauze 61, and the limit of 50 cm3 for the pumping volume contribution of the partition-wall, sets a limit. It implies that the pump can only compensate for an insufficiency up to that volume. The amount of fluid in the inflow expansion vessels on the left side increased and the amount of fluid decreased in those at the right side. After a short period of time, the system was again in balance, displaying reduced cardiac output because of the compensatory role of the partition-wall. These results should be compared with the authentic M-mode EC image of a patient with an aortic valve insufficiency (Chapter 6). The removal of the wedge resulted in a return to the situation at the start of the experiment.
The pulmonic valve insufficiency also resulted in pronounced pulsation's (Fig. 7-12), and in reduced outflow and pressure from chamber R. Filling pressure on the right side increased and made the partition-wall move towards chamber L in the filling phase. This resulted in a reduction in flow and pressure from the left side. The levels of fluid in the inflow expansion vessels on the left and right sides changed accordingly. A new equilibrium was soon established. The removal of the wedge returned those parameters to normality.
In order to simulate atrial septum defect (ASD) (Fig. 7-13), the experimental pump had to be provided with an additional valve 64. It connects the inlet tubes just above valves MV and TV.
By opening valve 64, an atrial septum defect (ASD) was simulated. No change in circulation could, however, be observed. This was due to the system being in a state of equilibrium, in which both chambers experienced the same filling pressure. The partitioning wall thereby assumed a neutral position and no flow through valve 64 ensued.
In order to simulate ventricular septum defect (VSD) (Fig. 7-13), the experimental pump had to be provided with an additional valve 65. It connects chambers R and L.
Opening valve 65 resulted in a ventricular septum defect (VSD). A sudden increase of pressure and flow in the right circulation system was observed. The left side circulation flow and pressure were reduced. Increased left filling pressure pushed the partition-wall toward chamber R in diastole, and thus changed the stroke volume on the left side, so that a new balance was reached. Closing valve 65 brought the system back to normal.
In order to simulate ductus arteriosus (DA) (Fig. 7-13), the experimental pump had to be provided with an additional valve 66. It connects the outlets just below valves AV and PV.
Opening valve 66 simulated an open ductus arteriosus (DA). The observed effect was the same as for a VSD. The response was, however, larger with DA (at same opening degree of the valves), due to a pressure gradient working over the entire diastole.
The healthy heart has its ventricular septum in the same position both in systole and diastole, except when reconstituting equilibrium because of re-partition of blood volumes. The ventricular septum thus has a convex shape against right ventricle also in diastole. In contrast to the experimental pump, the ventricular septum in the healthy human heart, does not normally contribute to the pumping of right ventricle by moving in the direction of right ventricle in systole. To maintain this septal position, a slightly higher filling pressure on the left side is required. An ASD implies, that this difference in filling pressure and the pumping action by the Atrio Ventricular-plane is counteracted by flow through the ASD. This causes a paradoxical movement of ventricular septum [179], giving right ventricle a higher stroke volume and left ventricle a lower one.
A self-regulating double-pump, operating in accordance with the basic balance principle suggested for the human heart in situ, was constructed (Concept 2). Two circulatory loops, arranged in the same way as the circulation of blood in man, have been shown to be kept in balance by the double pump. The similarity in function between this pump and the human heart goes even further; it allows simulation of a number of in vivo heart defects, including several types of myocardial infarction. It also demonstrates that as long as the partition wall is working properly, it will keep the system in balance as it does in man.
These findings applied to the human heart would suggest, that the regulating function of the ventricular septum is similar to that of the common wall in the double pump. That is on the assumption that the human heart pumps with displacement of the Atrio Ventricular-plane at a relatively constant outer contour (Chapter 3, Chapter 4, Chapter 5). This means that different ventricular stroke volumes do not jeopardise the regulation function of the ventricular septum. In fact, in contrast to the double-pump, an increasing stroke volume of the total heart (or a single ventricle) would also improve the regulation capacity of the ventricular septum. That is due to, that the wall between the two ventricles in the heart is more displaceable than the wall in the double-pump.
The pump is inferior to the heart in one major respect, and that is that it requires high filling pressure at high heart rates.
| Copyright © 1999 Inovacor AB. |