| A Single-Pump, with Ability to Control and be Controlled by Filling Pressure |
 |
|
|
| | Clear Image Window | |
In Chapter 7 it was possible (using the double-pump) to simulate the role of left and right filling pressures in the control of output balance; the balance is mediated by the ventricular septum.
The double-pump does not, however, embrace the other of the two essential working principles of the heart outlined earlier. That is the "gripping" pumping mode, attained by a moving valve plane connecting two chambers. This pumping mode is probably important for several reasons: it saves energy and requires low filling pressure at both low and high stroke rates.
The balancing action of the ventricular septum is activated by force components acting in the direction of the minor heart axis. The "gripping" mode pumping action though, occurs perpendicularly to it. The complex nature of the heart muscle enables it to exert its force in three dimensions.
Moreover, the process of polarisation makes it change from a material with high compliance in diastole to material with low compliance in systole.
Timing of the polarisation-de-polarisation process regionally within the heart is also of major importance. It is, therefore, not surprisingly, that the construction of a single-pump displaying all the attributes of the living heart is hardly feasible.
Left Ventricular Assist Devices, LVAD's, are increasingly gaining acceptance as means of circulatory support, but their clinical use has been delayed by problems. They are related to pumping characteristics (impaired filling at higher stroke rates and with inlet tubes of normal width), control of pumping rate, power supply and to blood compatibility.
There is evidence for the feasibility of artificial maintenance of systemic circulation, by an LVAD without the participation of RV [53, 86, 117, 121, 142].
In the present context, it might be of interest to transfer the "gripping" heart working mode into a mechanical pump [109]. It should be operating at essentially low and constant inflow pressures over a wide range of stroke rates. It should also be giving pulsed outflow pressures, similar to that of the left ventricle of the human heart.
A rather surprising solution for the control problems mentioned above was also found. By incorporating certain features into its design related below, it was possible to achieve stroke and thereby output, control by filling (inflow) pressure.
Its design also incorporates a basic principle for the shape of the inner surfaces of the heart postulated for the first time by Puff [135], implying that the surface area of the endocardium should remain essentially constant over the entire cardiac cycle.
The pump is shown in Fig. 8-1 and Fig. 8-2.
It is based on a tube with two bulbs made of flexible (but essentially non-extensible) material (e.g., certain types of polyurethane or silicone rubber, possibly reinforced by non-extensible fibres).
The tube with two bulbs, 6, is mounted in a casing 1, composed of parts 1A and 1B. The tube has a smaller bell-shaped bulb 6A and one that is larger, 6V. This can be seen in Fig. 8-1.
It has been produced in the following way: first a paraffin form was turned on a lathe and then polished by heat input; it was coated with several layers of an elastomer (which in this case was silicone rubber) with reinforcing material inlaid (gauze bandage in laboratory device), and finally allowed to set.
In the constriction 9, between the bulbs 6A and 6V, a dish-like rigid drive ring 10, is mounted; for the sake of light weight, it consists of two thin sheet-metal dishes fastened together and provided with holes. The drive ring is mounted in the casing 1 together with valves 4 and 5, in this case heart valves of a well known construction (Björk-Shiley valves).
As is evident from the drawings, the tube with two bulbs is fastened to other parts of the assembly in three places, namely to the valve 5 in the constriction 9 and also at openings 7 and 8 in the casing 1.
The whole arrangement is shown in mounted condition in Fig. 8-1.
Drive ring 10 runs freely up and down in the rigid casing 1. It has grooves 15 on its inner surface, so that air can pass freely between the sections of the casing on either side of the drive ring 10.
The smaller bulb 6A defines an atrium chamber (A), and the larger bulb 6V a ventricle chamber (V). The inlet opening to the atrium chamber A is joined to the casing at the opening 7.
The constriction 9 is a passage through which blood can flow only from the atrium chamber A to the ventricle chamber V, through the one-way valve 5.
The opening 8 (containing the one-way valve 4) is the outlet from the pump, through which blood is delivered in pulses under pressure.
Each portion of the bulbs is engaged between complementary, generally dish-shaped, surfaces of the housing and the drive ring. The respective volumes in the atrial and ventricular chambers of the pump are controlled by engagement of bulbs 6A and 6V. They are located between the lower and upper walls 25 and 26 of the casing, and lower and upper surfaces 28 and 27 of the drive ring 10. In particular, the casing surface 25 is concave while surface 28 of the drive ring is convex.
Similarly, the bulb 6A is engaged during a part of each cycle of the pump between convex surface 26 of the casing and concave surface 27 of the drive ring.
The pump may be driven by any of a variety of electrical or pneumatic drive devices, as represented schematically in Fig. 8-1. It may even be driven by muscular force.
The unilateral driving force is applied to drive ring 10, via push ring 12B attached to a pair of diametrically located push rods 12A. These project out of the casing through openings in the top wall and are sealed by suitable sliding seals (not shown), so that the casing is hermetically sealed. An electric motor acting through the linkage, pushes the push ring 12B down into engagement with the drive ring 10, on the driven stroke of each pump cycle. At the end of the driven stroke, the push rods and push ring disengage from the drive ring and are retracted to the top of the casing by a return spring.
During each down stroke of the push ring and drive ring, the volume of the ventricle chamber is reduced and its pressure is increased. That causing the valve 5 to close and the outlet valve 4 to open, so that blood is ejected from chamber V.
Meanwhile, the volume of chamber A increases, so that blood continues to flow into it during the driving stroke of the pump.
At the end of the down-stroke, push ring 12B is retracted, so that pressure is no longer applied to chamber V.
When the momentum that sustains flow through outlet valve 4 subsides, the valve will close. The pressure of the incoming blood, together with the momentum of blood then passing from A to V through the valve 5, will produce net upward forces. The forces are exerted by the ventricle chamber bulb 6V against the bottom surface 28 of the drive ring 10. One area of engagement is between the bulb 6V and the under surface of the drive ring. It is normalised by projection on an imaginary plane, perpendicular to the directional axis of movement of the drive ring 10. That area is being larger, than the area of engagement (normalised as above) between the bulb 6A and the upper surface 26 of the casing. Hence, the drive ring is lifted upward, and part of the blood in 6A together with incoming blood passes into the ventricle chamber. The volume of V increases as the drive ring rises.
When the normalised areas of engagement between surfaces 26 and 28, and bulbs 6A and 6V respectively become equal, the pump has reached its maximum volume. No more blood can then flow into the pump.
Accordingly, the frequency of the drive pulses of the drive device are established to ensure, that the chambers of the pump do not reach maximum volume between the driving strokes.
Is there something more that decides the extent, to which the chambers of the pump are filled during each operating cycle of the pump? The filling is also influenced by the pressure of the gas within the casing and outside the pump chambers.
During each driving stroke of the pump, the volume occupied by the gas in the casing increases, and the pressure of the gas drops accordingly.
During the return stroke of the pump, the total volume of the chambers increases, the volume of the gas decreases; the pressure of the gas in the casing increases accordingly. As the pressure in the gas approaches the pressure of the incoming blood, the rate of filling of the chambers decreases. It is therefore apparent, that the pressure changes that occur in the gas have a regulating effect on the filling of the pump throughout each cycle.
The gas pressures prevailing in the casing are determined (amongst other things) by the relationship between the displacement volume of the pump, and the volume occupied by the gas at any given point in the operating cycle. This is a matter of the geometric design of the pump.
The amount of the gas in the casing can be regulated by a pressure control valve, composed of two one-way valves, set to provide high and low limits on the gas pressure.
The operating cycle of the pump is schematically shown at four points in Fig. 8-3A-D.
Fig. 8-3A, Fig. 8-3B, Fig. 8-3C and Fig.
8-3D put together as a small animation
A driving stroke is commenced by a downward movement of the push ring 12B, by the force of the drive device, as shown in Fig. 8-3A.
Fig. 8-3B shows the pump at the end of the driving stroke.
In Fig. 8-3C, the drive device retracts the push ring 12B, at the end of the driving stroke. The hydrostatic pressure in the ventricle chamber will drop abruptly, and valve 5 will open due to both the hydrodynamic and hydrostatic pressures of blood entering the atrium chamber.
From now on (Fig. 8-3D) the two chambers act as a simple unit, confined between two stationary contact surfaces (25, 27) and two variable contact surfaces (26, 28). As long as there is a difference between the areas of the variable surfaces, the combined volume of 6A and 6V increases, when the drive ring moves upwards.
The force acting on the drive ring, is equal to the pressure difference multiplied by the differential area. The pressure difference is between inside and outside of the bulb assembly. The differential area, is the difference between the variable contact-surface areas 26 and 28. While the push ring 12B is retracted, the force acting on the drive ring is negligible, therefore liquid pressure on the inside must equal gas pressure around the bulbs. The filling thus depends on the difference between liquid pressure outside the inflow opening and gas pressure outside the bulb. A pressure difference of a few mm Hg is sufficient for adequate filling rate.
The single-pump is a new kind of displacement pump, not earlier found to be described.
A suitable name for this kind of pump can be differential-displacement pump (or DeltaV pump).
For in vivo application, the casing and the drive device may be mounted in a bag preferable of silicone rubber. It should be of such a volume, that the whole equipment will reach the same density as the replaced volume in the body.
The pressure control valve e.g., two one-way valves, one in each direction, provides communication between the interior and exterior of the casing 1. They have predetermined opening pressures for pressure regulation. Thereby, gas pressure within casing 1 can be maintained at a predetermined level. If the filling pressure is lower than the predetermined level of the gas pressure in the casing, drive ring 10 will not be displaced. The extent of filling of the ventricle chamber thus depends upon the pressure of the incoming blood.
At a certain position, the volumes of chambers A and V become equal. Upward movement of drive ring 10 (and thus the regulating function) then ceases. It does not matter how large the difference in pressure are (between the chambers and that prevailing outside these chambers).
If the pump is working beyond its regulating range, we will have a severe imbalance in the system, and a pulsating inflow to right ventricle. To prevent that, a sensor device may be provided for monitoring the highest position of the drive ring 10 during a cycle.
If the venous return increases, this will become noticeable because the drive ring 10 rises faster, towards the maximum volume. It is then possible to arrange a control circuit, that would increase the stroke frequency of the drive device. That should provide increase in stroke rate (and hence in output) similar to the so-called Bainbridge reflex in man.
An experimental model of the pump is shown in Fig. 8-4 and Fig. 8-5.
It is mounted in an aluminium case, for demonstration purposes.
It is equipped with instruments monitoring inflow and outflow pressure, flow, stroke rate and power consumption.
The pump is driven by a high-efficiency DC motor, provided with a gear box.
A tachometer is attached to the free end of the motor shaft and calibrated to display actual strokes per minute.
The outgoing shaft of the gear box drives a rotor disk, which effects the push rods unilaterally as described above.
The pump outlet is connected to a sealed polymethacrylate container, with silicone rubber bellows. The bellows are connected to the short outlet pipe, as a direct continuation thereof. This container with bellows serves as a capacity load (Windkessel) simulating device [139].
A pressure sensor is also attached between the outlet of the pump and the Windkessel unit.
At the outlet of the Windkessel unit, an electromagnetic type flow meter is arranged. The tube to which this flow meter is attached is flexible and may be constricted e.g., by a clamp to simulate a peripheral-type resistance analogous to in vivo circulation.
From this flexible tubing, the loop goes further to the inlet of a cylindrical container, the "waterfall vessel".
The inlet pipe of the container enters at the bottom, and continues upwards in the inside, concentric to the outer wall. This central pipe is shorter than the cylinder. The fluid flowing through the central pipe falls down into the space between the pipe and the outer wall. The outlet of the container is at the lower end of the outer wall. This arrangement transforms pulsating into non-pulsating flow i.e., it assumes part of the function of the arteriolar-alveolar system, with an outflow rate independent of inflow rate ("waterfall" concept [130]).
The outlet of the "waterfall vessel" is connected to a third polymethacrylate container, which in turn is joined (by short pieces of flexible tubing) to the inlet of the pump.
The third container is a substitute for the venous system (including atrium and auricle), furnishing the pump with a smoothing means (in a sense, a negative Windkessel type).
A sensor for monitoring inflow pressure is arranged at the inflow.
An open fluid expansion vessel is also connected to the system near the inflow opening. This open expansion vessel may be raised or lowered, in order to increase or decrease the circulating fluid volume. It can also be used to raise or lower fluid pressure at the inlet of the pump. It provides a substitute for the capacity-changing capabilities of the venous system in vivo.
The two Windkessel units are provided with valves enabling their inner air pressure to be changed.
This experimental pump has not been optimised. Its casing has not been sealed against ambient pressure. The model thus does not make full use of the regulation capabilities of the basic pump design, with respect to variable filling pressures (see above). The present pump in all experiments will thus have its regulation level at atmospheric pressure, i.e., a filling pressure of 0 mm Hg.
The inflow and outflow pressure curves at a low and high frequency (100 and 300 beats/min), were recorded (Fig. 8-6A, Fig. 8-6B).
The inflow pressure curves were transferred to an EchoCardiographic (EC) right side registration, of the Atrio Ventricular(AV)-plane of a healthy subject Fig. 8-7. The event marks a'-h' described in Chapter 3 were added.
The EC registration was done with simultaneous ECG and phonocardiogram registrations, together with jugular vein pressure pulse tracing during a slight Valsalva manoeuvre, to raise jugular vein pressure.
At systole, all pressure traces in Fig. 8-7 show a decrease in pressure (x-wave).
The x-wave is followed by a v-wave in the jugular vein pulse curve, and in the inflow pressure curve of the single-pump at high frequency. There is no v-wave recorded for the single-pump at low stroke rates. The reason for this is that the profile of the cam disk, transmitting power from the motor to the push rods, determine the slope of the x-wave. Thereby it also determine the size of dynamic forces generated.
In the heart it is the velocity of ventricular muscle contraction that accomplish that.
The single-pump, in contrast to the heart, has no forced atrial contraction raising the valve plane in further atrially. In other words, it can not raise the valve plane in the pump in the direction of the inflow. Therefore no extra pressure rise in the form of an a-wave is seen, in the inflow pressure curve for the single-pump.
The major difference at low frequencies between the heart and the single-pump with respect to atrial/inflow pressure, is thus the absence of pressure transients in the latter.
The effect of the inflow pressure on the minute volume was monitored (Fig. 8-8).
Before the experiment was started, filling pressure was set to zero. That means, that the fluid level in the adjustable expansion vessel 7 (Fig. 8-5) was placed at the same level as the drive ring 10 (Fig. 8-1). It was in a position halfway between its maximum diastolic and maximum systolic position (Fig. 8-3A, Fig. 8-3B, Fig. 8-3C, Fig. 8-3D).
The stroke length of the drive ring was 18 to 20 mm. This implies, that stroke length due to the influence of gravity deviated up to about 1 mm Hg from the preset value.
The stroke rate was set at 100 beats/min.
Then the experiment was started, and the adjustable expansion vessel was raised step-wise, until pump regulation capacity was fully utilised.
Then, the expansion vessel was again brought back to zero level and stroke rate was increased to 200 beats/min.
The expansion vessel was again raised step-wise.
In connection with this experiment the following observations were made:
The pump can be provided with a sensor monitoring the position of the drive ring 10. The signal of this sensor is used to control the stroke rate, in order to prevent the pump from working outside its regulating range.
The effect of increasing stroke rate on the performance of the pump was tested (Fig. 8-9).
The same preparations were made as in the preceding experiment. The expansion cylinder was then raised, so as to achieve an inflow pressure of 3.5 mm Hg.
A constant peripheral resistance was introduced by clamping the outflow tube connecting the electromagnetic flow sensor and the "waterfall" cylinder.
The pump was then started with an initial frequency of 50 beats/min and the stroke rate was increased by increments of 25 beats/min.
The following observations were made:
At a stroke rate of between 250 and 275 beats/min, a renewed increase of efflux rate and pressure is noticed. Filling pressure then reaches a new constant level, near the extreme systolic position of the drive ring 10. Inflow is almost completely non-pulsating, as can be seen from the very small pressure fluctuations on the inflow side.
A time-expanded recording (Fig. 8-6B) at this frequency range, shows the appearance of a v-wave in the inflow pressure trace. The inflow pressure trace is now almost identical to the jugular vein pulse pressure curve (presented in Fig. 8-7).
Further increase of the stroke rate, raises output due to these dynamic forces, until the decrease of filling pressure exerts a limiting effect on minute volume (Fig. 8-9).
This increased flow at high stroke rates shows that dynamic forces dominate. Outflow through valve 4 (the "aortic" valve) continues even after the drive ring 10 has reached its extreme systolic position. This implies that valve 5 (the "mitral" valve) opens before the closure of valve 4.
The dynamic forces also push back the drive ring. The v-wave in the filling pressure trace of the single-pump, is probably due to these dynamic effects.
The single-pump does not employ any elastic forces, to bring the drive ring 10 to its end diastolic level.
There are two kinds of forces which bring the drive ring back to its end-diastolic level:
The heart effects the movement of the AV-plane towards apex, by shortening and thickening the muscles within a practically constant outer contour of the ventricles (see Chapter 3, Chapter 4 and Chapter 5).
At the end of systole and closure of the aortic- and pulmonic valves, atria and ventricular chambers (as in the single-pump) act as a single unit. They have a common volume, that is embedded in the surrounding tissues of the heart (including the pericardium).
The dynamic forces that have been generated by movement of the AV-plane, will be directed towards the apex. The ventricular part of the common volume (which is moving in the direction of the dynamic forces) will change direction, due to resistance of the surrounding tissues. As in the single-pump, the dynamic forces will reverse and the AV-plane will be pushed away from the apex.
It is obvious, that if the ventricles at the end of systole were allowed to expand their outer contour, less force would be left to move the AV-plane. Furthermore, the volume expansion of the ventricles would (through the pericardium) reduce the possibility of the atria to expand. This would result in a reduced movement of the AV-plane, and another type of pump (a common displacement pump).
When no more expansion of the total heart volume is possible (due to restriction of the pericardium and the surrounding tissues), the following happens; any remaining dynamic forces will be converted to a pressure gradient (Bernoulli's Theorem, water-hammer effect), and movement of the AV-plane will cease. An example of this is shown in Fig. 8-10A, Fig. 8-10B, Fig. 8-10C, where the third heart sound coincides with the valve plane reaching event g'.
The third sound is thus produced by the sudden retardation of the dynamic forces of diastolic inflow of blood (cf. Ishimitsu et al. [82]).
There are some indications, that larger stroke volumes can be achieved under some conditions, for instance in aortic valve insufficiency (Chapter 6) and at inspiration. The course is, that when the pressure within the thoracic cavity is negative (compared with the pressure within the heart), the following happens; the AV-plane will be pushed further away from the apex, resulting in a thinning of the walls of the ventricles, and thus larger stroke volumes are achieved.
This indicates, that when the heart in diastole expands its total volume by relatively long acting forces (and thus low support by the surrounding tissues), the following happens; the imaginary surface perpendicular to the directional movement of the AV-plane is larger on the ventricular side, than on the atrial side. The net forces are then pushing the AV-plane towards the atria (as in the single-pump).
With rising heart rate, the length of systole increasingly encroaches on ventricular filling.
At a heart rate of 60 beats/min, the length of ventricular diastole is about two-thirds of the cardiac cycle.
At a rate of 180 beats/min, the situation is reversed, with systole now accounting for approximately two-thirds.
It should also be noted that in absolute terms, diastole is reduced in length from about 300 milliseconds to about 100 milliseconds [56]; this is in agreement with earlier observations of intact dogs [57, 58].
It is obvious that with increasing heart rate, dynamic forces must dominate diastolic filling of the heart, in the same way as they do in the single-pump.
It would not be surprising if the same phenomenon that has been observed in the single-pump was found; in the normal heart working at high rates and low peripheral circulatory resistance, the mitral- and tricuspid valves open before the closure of aortic- and pulmonary valves.
The "gripping heart" working mode, Concept 1, served as a model for making a mechanical pump operating at essentially low and constant inflow pressures. That was relevant for a wide range of stroke rates, giving pulsed outflow pressure similar to that of the left ventricle of the human heart.
The following conclusions were made, in testing this pump (which can be used e.g., as a Left Ventricular Assist Device, LVAD):
| Copyright © 1999 Inovacor AB. |