IL102100A - Method and apparatus for servo-controlling the motor of a peristaltic heart prosthesis - Google Patents

Abstract

The invention relates to a method of automatic control of a motor of a peristaltic cardiac prosthesis, characterised in that it comprises the following steps: - during an ejection phase, applying to the motor a current which causes its rotation in a first sense at an essentially constant speed, and - during a return phase, applying to the motor a current which causes its rotation in the opposite sense at a speed which varies according to a monotonic law with the value of a parameter obtained from the average current flowing in the motor and representing the average force provided by the motor. The invention also relates to a device implementing the method. <IMAGE>

Description

102100/2 >\3 \3^ a arnji ansa Ttt -nifc>ii ipnm ηυ ^ Method and apparatus for servo-controlling the motor of a peristaltic heart prosthesis LINIQUE DE LA RESIDENCE DU PARC incent POL ean-Claude DUMAS METHOD AND APPARATUS FOR SERVO-CONTROLLING THE MOTOR OF A PERISTALTIC HEART PROSTHESIS The present invention relates in general to prosthetic heart pumps, and more particularly to a method and apparatus "*'. for servo-controlling a peristaltic heart pump.

European patent application No. 0 148 661 in the name of the Applicants discloses an implantable blood pump of the peristaltic type which makes it possible to respond adaptively to a certain number of physiological requirements. Such a pump, like the natural heart, is a pump of the peristaltic type. By an appropriate selection of its operating point, its flow rate response is linear.

An essential characteristic of that known pump lies in that variations in flow rate do not depend on any external sensor, i.e. the operation of the pump is determined solely by using such information as is inherent to its connection to the network of blood vessels, i.e. the aorta pressure and the venous return volume. Thus a characteristic of the operation of that pump is that it never tends towards positive displacement operation.

A variation in aorta pressure occurs each time that vascular constriction is observed in the peripheral system. Thus, during contraction of the artificial ventricle, systolic ejection can take place only at the moment when the pressure in the bag has become at least equal to the aorta pressure.

At the end of the ejection stage, the pressure in the bag becomes very much greater than the aorta pressure, thereby building up reserve of energy that can be made use of. Thus, if the power supply to the motor is switched off at this instant, it is observed that the pump returns at a speed that is roughly proportional to the amount of energy that has been stored, i.e. to the pressure in the bag.

As a result, if the pressure is low, then the motor returns slowly, the frequency of the pump is low, and the blood flow rate is small. Conversely, high pressure in the bag gives rise to an increase in frequency and in flow rate. The known pump thus implements hydraulic servo-control that optimizes its operation as a function of external conditions.

Such hydraulic servo-control of the known pump is also advantageous in that the mean electricity consumption of the motor is reduced, with the motor being powered in some cases only during the systolic contraction phase, i.e. the shortest phase.

In order to ensure natural control over this phenomenon at all operating frequencies, it is advantageous to select a high contraction speed so that said phase lasts for about 200 ms, together with a lower return speed having a duration of about 1000 ms, giving a basal frequency of about 50 beats per minute.

When venous return increases, the bag is filled to a greater extent and the pressure therein also increases. The return speed therefore increases and the pump operates at a higher frequency to cope with the flow rate.

However, such hydraulic servo-control relies on the existence of high aorta pressure since otherwise the pressure in the bag is limited to a value that prevents the operating frequency increasing beyond a certain limit. Thus, the design of the pump described in the above-mentioned patent application does indeed make it possible to increase frequency as a function of filling volume; more precisely an increase in said volume at constant contraction speed gives rise to an increase in the quantity of blood injected per unit time and thus to an increase in pressure.

However, experience shows that this complementary effect is insufficient since with a natural heart variation in flow rate as a function of filling volume predominates.

To mitigate this difficulty, a known solution consists in using one or more pressure sensors and/or flow rate sensors in the auricular cavity of the heart, and. in controlling the return speed to increase the operating frequency in the event of low aorta pressure. However, that solution is complex and dangerous since the sensors require proper electrical powering and they must also be reliable and accurate.

U.S. 4,884,013, published November 28, 1989, discloses a motor unit for a mechanical fluid delivery system, which includes pump operating means and pump control means which comprise means for sensing the condition of the pump operating means.

U.S. 5,054,353, published November 12, 1991, discloses a pressure responsive linear motor driven pumping apparatus. The apparatus comprises an electric linear motor, pressure detecting means,, energization means, current detecting means and current control means.

The present invention seeks to mitigate these drawbacks of the prior art and to provide a method and an apparatus for controlling the motor in a heart pump of the above-mentioned type without requiring special sensors to be implanted and without requiring any change in the physical structure of the pump.

To this end, the present invention provides a method of servo-controlling a motor of an electronically switched peristaltic heart prosthesis, characterized in that it comprises the following steps: applying current to the motor during an ejection stage that causes the motor to rotate in a first direction at an essentially constant speed, the torque developed by the motor being proportional to the amount of blood ejected per unit of time, and determining the torque developed by the motor to obtain information concerning the blood flow required by the human body by measuring the mean current consumed by the motor, and during a return stage, applying to the motor a variable current, by changing the width of the current pulses applied to the motor, to cause it to rotate in the opposite direction at a speed proportional to the required blood flow rate as represented by the measured mean current.

Given constant speed during the ejection stage, and consequent constant duration for said stage, the torque developed by the motor is proportional to the quantity of blood e ected per unit time, i.e. to the filling volume. By measuring the force or the torque delivered by the motor, information is thus obtained concerning the blood flow rate that is physiologically desirable, and on the basis of this information, the return speed and thus the beat rate of the prosthesis is adjusted to the appropriate value.

Other aspects, objects, and advantages of the present invention appear more clearly on reading the following detailed description of a preferred embodiment thereof given by way of non-limiting example and made with reference to the accompanying drawings, in which: Figure 1 is a diagrammatic perspective view of a motor for a heart prosthesis controlled in accordance with the present invention; Figure 2 is an electrical circuit diagram representative of the, motor in Figure 1; Figure 3 is a block diagram of the motor and of its control s stem; 4 Figure 4 is a circuit diagram of a first portion of the control circuit of the invention; Figure 5 is a circuit diagram of a second portion of said circuit; and Figures 6a, 6b, 7, 8a, and 8b are waveform diagrams illustrating the operation of a portion of the circuit of the invention.

As a preliminary point, it may be observed that portions or elements that are identical or similar from one figure to another are designated therein by the same reference symbols.

With reference initially to Figures 1 and 2, the motor to which the control method of the present invention applies is a motor without a commutator, slip rings, or brushes, for the purpose of ensuring the high degree of reliability and long life that are necessary for the application in question. The motor is controlled by single- or multi-phase electronic switching. In the description below, a two-phase motor is considered, and all the angles given are electrical phase angles.

The motor comprises a rotary portion or rotor RT, and a non-rotary portion or housing HA.

The rotor RT includes a set of permanent magnets (not shown) secured to a rotary shaft AR. The magnets define a magnetization vector NS extending along the shaft perpendicularly to its rotary axis, as shown in Figure 1.

The rotary shaft is used as the drive shaft and it is supported about a stationary axis that is centered relative to the housing HA e.g. by means of two ball-bearings in other bearings (not shown) which provide it with complete freedom in rotation. For further details concerning the installation of . the motor in a blood pump, reference should be made to European patent application No. 0 148 661 in the name of the Applicants, the contents of which is incorporated into the present description by reference.

The housing HA comprises two subassemblies that are fixed relative to each other and that are completely symmetrical about the axis of rotation of the rotor. 5 A first one of the subassemblies constitutes a stator ST which is itself constituted by a set of copper wire windings designed in such a manner as to be equivalent to a pair of coils Ml and M2 suitable for producing a magnetic field that rotates as a function of the polarity of the current flowing through one and/or the other of said coils, said rotating field always being perpendicular to the axis of rotation of the rotor.

Since the motor does not include a commutator, special means are provided to provide information at all, times concerning the angular position of the rotor, i.e. the direction in which its magnetic field vector is pointing relative to the housing. To this end, the second subassembly of the housing consists in a resolver RE. It is constituted by Hall effect coils or cells EH, and in the present case there are two of them.

The principle on which electronic control of the motor is based is described below with reference to Figure 3.

In a first block Fl, the angular position of the rotor is measured on the basis of the signals delivered by the cells of the resolver RE. The function of the block F2 is to determine the configuration of electric currents to be caused to flow through the coils Ml and M2 to obtain simultaneously the desired motor torque, the desired direction of rotation, and the desired speed of rotation, which parameters are provided in the form of reference values applied to three corresponding inputs of the block F2.

The block F3 serves to deliver electrical currents to the coils Ml and M2 of the stator. To this end,' it includes a power circuit controlled by the block F2 and receiving power from a suitable DC power supply unit referenced AL.

The block F2 determines the currents to be passed through the stator on the basis of three parameters PI, P2, and P3. PI is a logic signal specifying the direction of rotation of the motor: forwards or backwards. P2 is a signal whose value is representative of the maxiiTtum speed authorized in forwards rotation, and also the associated maximum torque which is 6 likewise a function of the electrical power delivered by the block F3. P3 is a signal whose value is representative of the reference speed of backwards rotation.

The motor torque delivered to the shaft AR is a direct function of the electrical power delivered to the block F3 by the power supply unit AL. By including the efficiencies of the motor, of the associated pump mechanism, and of the power, supply block F3 in the calculation, it is possible to obtain an equation which expresses the torque delivered as a function of the current drawn by the block F3.

A specific example of an electric circuit implementing the functions of the blocks Fl, F2, and F3 is described below with reference to Figures 4 and 5.

The power supply unit AL is constituted, for example, by one or more batteries of appropriate capacity, e.g. delivering DC at +12 volts and referenced +V.

A low value resistance constituting a measuring shunt SM is interposed between the negative terminal of the unit AL and ground .

The terminals of the shunt SM are connected to the non-inverting input, and to the inverting input respectively of an operational amplifier Al connected as an amplifier/integrator and having a capacitor CI connected between its output and its inverting input.

A block DET.S combines an integrator having a long time constant (e.g. about 10 seconds) with a threshold detector situated upstream from the integrator, for purposes explained below.

The output signal from DET.S is amplified by the operational amplifier A2 whose gain determines the slope of the automatic control characteristic, as described below. The output of A2 is connected to the non-inverting input of another operational amplifier A3 via a resistor Rl. A manual adjustment potentiometer PI has its slider connected via a resistor R2 and a diode Dl to the non-inverting input of A3. A limiter circuit LIM is also connected via a diode D2 to said input . 7 The operational amplifier A3 is connected as a high gain integrating amplifier, with a capacitor C2 being connected between its output and its inverting input. Its output is connected to the base of an NPN bipolar transistor Tl via a resistor R3. The emitter of Tl is connected to ground via an emitter resistor R4 and it is also connected to the non-inverting input of an operational amplifier A4. The inverting input of A4 is connected to ground via a resistor R5 and to its output via a resistor R6. The output of A4 is also connected via a resistor R7 to the inverting input of A3.

The collector of Tl is connected to an output terminal Bl of the circuit, and also to the inverting input of an operational amplifier A5. The non-inverting input of A5 receives a determined fraction of the voltage +V via a divider bridge, and a capacitor C3 is connected between said non-' inverter and the output of A5, which output is connected to an output terminal Q' . As explained below, this output gives an indication of the position of the bracket in the heart pump.

The circuit includes two other output terminals B2 and BO respectively connected to +V and to ground.

The output terminals BO to B2 are for connection via appropriate cabling to the portion of the circuit shown in Figure 5 which is integrated in the prosthesis.

The terminal Bl constitutes a "third wire" for conveying a current written i in Figure 4 that constitutes a return speed reference signal as generated on the basis of the mean value of the current drawn by the motor.

The portion of the circuit situated in the prosthesis is described below with reference to Figure 5. Terminals ΒΌ, B'l, and B*2 are respectively connected via suitable flexible conductors to the terminals BO, Bl, and B2 of Figure 4.

Terminals BIO & Bll and B12 & B13 receive output signals from the two Hall effect probes EH of the motor. These signals are applied via resistors RIO, Rll, R12, and R13 to the input terminals of two operational amplifiers A10 and All each associated with a respective one of the probes. These amplifiers are connected as differentiators, with feedback 8 resistors R14 and R15 connected between their outputs and their inverting inputs. The outputs of A10 and All are respectively connected to the non-inverting inputs of two comparators A12 and A13 whose inverting inputs are connected to a mid voltage generator circuit that is described below. The outputs of A12 and A13 are respectively connected to two inputs of a four-way decoder DEC, and also to two inputs of a first EXCLUSIVE-OR gate PIO . The output of P10 is applied directly to a first input of a second EXCLUSIVE-OR gate Pll, and to the second input of said gate via a resistor R16. A capacitor Cll is provided between said second input and ground. The output of Pll is connected via a forwards-connected diode D10 to a first input of a third EXCLUSIVE-OR gate PI2 whose other input is connected to ground. Said first input of P12 is connected to ground via a resistor R25 and a capacitor C13. The output of P12 is connected to the control input of the decoder DEC. The four parallel outputs of DEC are applied to a power circuit PC, for example a conventional type of hybrid circuit having the function of applying appropriate currents in the appropriate directions to the windings Ml and M2 with timings defined by the states of its inputs.

The circuit powering the Hall effect probes and generating the above-mentioned mid voltage is constructed around two controlled changeover switches, e.g. C-MOS type switches, given respective references 110 and 111, which switches are capable of generating a voltage that operates between positive and negative between two terminals B14 and B15. Two resistors R18 and R19 connected in series between these two terminals serve to apply one-half of said voltage to the inverting inputs of A12 and A13, thereby forming a comparator threshold.

The terminal B'l is connected to a first fixed contact of another controlled switch 112 via a resistor R22. The othe fixed contact is connected to ground via a resistor R21. The moving contact of 112 is connected to the common point between R25 and CI 3. A resistor R23 and a capacitor CIO are connected between ground and the first fixed contact of 112. An PN type bipolar transistor T10 has its collector also connected to said 9 first fixed contact, while its emitter is connected to ground and its base is connected via a resistor R24 to the Q output of a rising edge D-type bistable referenced BS, which output also controls simultaneous switchover of the three switches 110, in, and 112.

The set and reset inputs (S, R) of the bistable BS are connected to input terminals B16 and B17 for receiving signals delivered by two end-of-stroke sensors of the heart pump, said sensors being described in the above-mentioned European patent application.

The operation of the apparatus of the present invention is described below.

It is assumed initially that the pump is in its return stage. According to an essential aspect of the present invention, the return speed must be based on a reference speed which is itself set as a function of the energy consumed by the motor during the ejection stages.

The value of this energy is obtained by integrating the current i drawn by the motor. More precisely, the voltage across the terminals of the resistor SM is proportional to said current. This voltage is amplified by Al and the integrator function thereof serves to eliminate meaningless transients, such that its output produces an instantaneous voltage Vint which is proportional to the current.

The circuit DET.S is a detector/integrator having a long time constant, e.g. of the order of 10 seconds. The purpose of the detector (threshold circuit) is to prevent the integration taking account of meaningless residual currents (in particular the current drawn, while the pump is performing return motion, i.e. is rotating backwards ) . The integrator enables the output of DET.S to produce a voltage which is representative of the mean force over some number of heart beats. More precisely, given that the systolic contraction speed has been selected to be constant, as explained below, the torque developed by the motor during said phases is proportional to the quantity of blood ejected per unit time, i.e. to the filling volume. An indication of the physiologically desirable flow rate is thus obtained in this manner. 10 The gain of A2 which may be modifiable, for example by means of an appropriate variable resistor (not shown), serves to determine the slope of the automatic control by varying the feedback loop gain.

A3 adds and integrates the voltages from A2 and from PI, with said potentiometer constituting a manual control member. The circuit LIM co-operates with the diode D2 to fix an upper limit that is not to be exceeded for the voltage applied to A3, for purposes explained below.

The output of A3 biases the base of Tl, thereby serving to control the current i that can flow through said transistor from the third wire terminal Bl to ground via the resistor R4.

The essential function of the circuit of Figure 4 is thus to allow current to flow through Tl in a manner that increases with the torque produced by the motor in compliance with a specific curve that is approximately a straight line. The general slope of said curve is fixed by the gain of A2 while the amplitude of said curve is fixed by the potentiometer PI. Finally, the limit circuit prevents the high end of the curve exceeding a specified ceiling, by forcing it to said ceiling.

The operation of the portion of the circuit shown in Figure 5 is described below with reference to the timing diagrams of Figures 6b, 7, 8a, and 8b. Figure 6a shows the angular distribution of equilibrium positions for the rotor relative to the two windings Ml and M2^when powered for forwards operation (VI and V2, respectively) and for backwards operation (Rl and R2, respectively). In addition, in the timing diagrams, the left-to-right direction corresponds to the motor operating forwards (ejection stage), while the right-to-left direction corresponds to the motor operating backwards (return stage) .

The controlled switches 110 and 111 serve to apply bias voltages to the Hall effect cells alternately in one direction and in the opposite direction. The voltage taken from the common point between R18 and R19, both of which have the same resistance, constitutes the mid voltage Vc of said bias. 11 The outputs of the differential amplifiers A10 and All deliver the sinewave voltages produced by the Hall effect sensors, as represented by solid lines designated RESl and RES2 in Figures 5 and 6a. The inverses of the voltages RESl and RES2 are represented by dashed lines. x The amplifiers A12 and A13 are connected as comparators, having the above-mentioned mid voltage as their common threshold. Their respective outputs therefore deliver squarewave signals referenced SA1 and SA2 which take up a high level (+12 volts) during positive half-cycles in the signals RESl and RES2 , and a low logic level (0 volts) during negative half-cycles.

These signals are applied to two signal inputs of the decoder DEC, and also to the two inputs of the gate P10.

The decoder DEC has an appropriate combinatorial logic circuit for the purpose of generating logical position signals Q0, Ql, Q2, and Q3 (Figure 6b) which are respectively associated with the four equilibrium positions of the motor.

In compliance with another advantageous aspect of the present invention, it can be seen that the positions of the sensors EH are chosen so that the position indications can be derived from signals that relate solely to the signs of the signals produced by the sensors. This is achieved by positioning the sensors EH at an offset relative to the coils Ml and M2 such that the sensors (and thus the extremes of the signals that they produce) are angularly halfway between pairs of equilibrium positions of the rotor.

In the present example of a two-phase motor, the value of the offset is 22.5°.

Conventionally, such sensors "used to be positioned overlying the windings and as a result the edges of the generated logic position signals were derived from the relatively flat peaks of the substantially sinusoidal signals delivered by the sensors. The resulting accuracy was poor.

Because of this aspect of the invention, the edges of the position signals are obtained with very good accuracy, and in a manner that is particularly appropriate for controlling a motor 12 as described below, they are situated at angular positions that are exactly halfway between the equilibrium positions of the motor. More precisely, each signal QO to Q3 passes to the high level at a mid position between an adjacent equilbrium position and the associated equilibrium position, and it returns to the low level at a mid position between the associated equilibrium position and the following equilibrium.

The output of the gate PIO delivers a squarewave signal SB as shown in Figure 7. The purpose of the gate Pll is to produce a narrow pulse at its output whose width is determined by the values of R16 and Cll, on each rising edge and on each falling edge in SB (signal SC in Figure 7). As a result, one such pulse occurs at each angular position halfway between two equilibrium positions of the motor (which positions are referenced VI and V2 for forwards operation and Rl and R2 for backwards operation), with the pulses being separated by electrical angles of 45° that are 22.5° in advance of each equilibrium position.

It is now assumed that the motor is in a return stage. According to an essential aspect of the invention, the return speed is determined by the heart force to be delivered, which force is represented, as explained above, by the value of the current i that can flow through the transistor Tl. In this situation, the bistable BS delivers a low level logic signal Q, and the controlled switches 110, 111, and 112 take up the positions shown by dashed lines in Figure 5. The transistor T10 is off.

Figure 8a shows the case where the mean force delivered by the motor during the preceding cycles has been low, such that the current that can flow through through Tl is also low.

Under such circumstances, the signal SD at the connection between D10 and R25 and as applied to the second input of the gate P12 has the appearance of a sawtooth voltage with a gentle descent slope. More precisely, when a pulse of the signal SC appears at the output of Pll, the capacitor CI3 charges quickly via the resistor R25 to a relatively low value. When the pulse disappears, CI 3 discharges via the only available circuit, i.e. 13 via 112, R22, Tl, and R4, and it therefore discharges at a rate determined by the current that can flow through Tl, and in the present case that means it discharges slowly.

Since the other terminal of P12 is grounded, the output of P12 delivers a logic signal SE whose duty ratio varies as a function of the rate at which the signal SD decreases. In the present case, broad pulses are observed at high logic level, corresponding to instants during which the voltage SD is greater than the threshold voltage Vs at which the output of PI2 switches over from one level to the other.

The voltage SE is applied to a control input of the circuit DEC in which it is combined using an appropriate logic function with each of the signals QO to Q3 so as. to delay the rising edge in the signal Qi under consideration to the instant at which the falling edge occurs in SE. Figure 8a shows the signal QO at instants to and to' for its rising edge and its falling edge. The signal R0 generates from QO and SE is a narrower pulse whose rising edge is at instant tO" that corresponds to the falling edge in SE, while its falling edge remains at the instant to ' .

In the case shown in Figure 8b, the mean force delivered by the motor during the preceding cycles was large. The current i that can flow through Tl is therefore large, and the rate at which the signal SD falls is therefore fast. The pulses in SE are therefore shorter, such that the pulses of the signals Ri (and in particular of the signal R0) begin much sooner than in the case shown in Figure 8a, and are therefore much wider.

Thus, in the case of Figure 8b, the mean current delivered to the motor is large because of the high duty ratio of the signals Ri as applied to the power circuit CP.

Starting from the situation shown in Figure 8b, when the mean force delivered falls off, i.e. when the current drawn by the motor during the ejection stages decreases, then the voltage at the input of A3 also decreases, as does the current i. The situation thus tends towards that shown in Figure 8a, thereby slowing down the return speed and decreasing the operating frequency of the pump. 14 It may be observed at this point that, in particular during low force stages, current is delivered to the motor during periods that correspond to positions that are intermediate between successive equilibrium positions. In this way, the motor operates efficiently since it operates as far as possible from its equilibrium positions.

An essential advantage of the circuit described lies in the fact that it automatically regulates the effective speed of the motor to the reference speed. More precisely, consider the case where the motor is operating under the conditions shown in Figure 8 and the speed of rotation of the motor begins to increase. Assume that the mean force delivered by the motor has not changed and that the rate at which the signal SD decreases remains the same.

Under such circumstances, the duration of each cycle in the signal SE decreases, but the duration of the high level stage in the cycle remains the same since the slope of the signal SD remains unchanged. The duty ratio of SE therefore increases, and as a result the duty ratio of RO (and of the other signals Rl to R3) decreases, thereby reducing the amount of current delivered to the motor, thus reducing the torque it delivers, and thereby reducing its speed, assuming that external conditions remain identical.

The opposite regulation phenomenon occurs when the effective speed of the motor falls relative to the reference speed as determined by the current i.

The operation of the circuit in Figures 4 and 5 is described below during a systolic ejection stage. The beginning of such a stage is triggered by the action of the back end-of-stroke detector which applies a set pulse to the bistable BS. The output Q of the bistable BS takes up a high level, e.g. 12 volts.

The controlled switches 110, 111, and 112 take up the positions shown by solid lines. Because the signals applied to the resolvers are inverted, the time sequencing of the signals Q0 to Q3 is inverted, thereby causing the motor to operate in the forwards direction. The gates P10 to P12 perform the same 15 function as before, except that CI3 no longer charges via Tl, but charges at constant speed as set by the resistance of resistor R21. The slope at which the signals SD fall off is thus invariable, as is the duty ratio of the signals SE, and thus of the signals RO to R3 applied to the power circuit. For example, these signals may begin in the mid position between the preceding equilibrium position and the associated equilibrium position ( as when operating backwards ) , and they may terminate in the vicinity of the associated equilibrium position.

This achieves a constant speed ejection stage that may occupy a duration of about 200 ms, for example.

During the ejection stage, the circuit of the invention performs another function whereby the position information delivered by the two end-of-stroke detectors of the pump as transferred to the output Q of the bistable BS is returned to the circuit of Figure 4.

More precisely, during the return stage described above, a certain amount of current flows through Tl and its collector is therefore at some positive voltage. The resistances of R8 and R9 are chosen so that regardless of the value of i, the collector voltage of Tl as applied to the inverting input of A5 is always greater than the voltage applied to the non-inverting input of A5. The output of A5 therefore delivers a low level logic signal on its output Q' indicating that the motor is operating backwards.

Conversely, during the ejection stage, the discharge current of C13 no longer flows through Tl so the terminals Bl and B'l are returned to a voltage close to zero since T10 is then switched on. This voltage is less than the voltage at the non-inverting input of A5 such that the output Q' now delivers a high level logic signal.

As a result, each time the signal Q' takes up a high value, that indicates the beginning of an e ection stage. This information delivered on an appropriate connection terminal external to the patient thus makes it easy to measure the frequency and the flow rate of the heart prosthesis.

Naturally, the present invention is not limited to the embodiment described above and shown in the drawings, and to the person skilled in the art will be able to apply any variant or modification thereto that complies with the spirit of the invention.

Claims (21)

- 17 - 102100/2 CLAIMS:

1. A method of servo-controlling a motor of an electronically switched "peristaltic heart prosthesis, characterized in that it comprises the following steps: - applying current to the motor during an ejection stage that causes the ; motor to rotate in a first direction at an essentially constant speed, the torque developed by the motor being proportional to the amount of blood ejected per unit of time, and determining the torque developed by the motor to obtain information concerning the blood flow required by the human body by measuring the mean current consumed by the motor, and - during a return stage, applying to the motor a variable current, by changing the width of the current pulses applied to the motor, to cause it to rotate in the opposite direction at a speed proportional to the required blood flow rate as represented by the measured mean current.

2. A method according to Claim 1, characterized in that the mean current is determined according to a plurality of operating cycles of the motor.

3. Apparatus for servo-controlling the motor of an electronically switched peristaltic heart prosthesis, characterized in that said apparatus comprises means (Fl, F2, F3) for applying to the motor: - current that causes it to rotate in a first direction at an essentially constant speed during an ejection stage, and - current that causes it to rotate in the opposite direction at a speed proportional to the mean current flowing through the motor and representative of the mean force delivered by the motor, during a return stage, and wherein it also comprises: - first sensor means (SM, A1-A4) for generating an electric signal proportional to the mean current flowing through the motor, - second sensor means (EH, A12, A13) for generating position signals (Q0-Q3) representative of the angular position of a rotor (RT) of the motor, - signal width control means (P10-P12, C13) capable: - 18 - 102100/2 during a return stage, of producing control signals (SE) of width varying as a function of said electric signal proportional to the mean current, and during an ejection stage, of producing control signals of fixed width, - a logic circuit (DEC) combining the position-indicating signals and the control signals to produce current switching signals (R0-R3), - a power circuit (CP) receiving said current switching signals and applying to at least one winding of the motor a current whose time sequencing depends on said switching signals and determines the speed of rotation of the motor.

4. Apparatus according to Claim 3, characterized in that the first sensor means comprises: - - a resistor (SM) connected in series in the path of current to the motor, and - amplification and integration means (A1-A3, DET.S) receiving the voltage across the terminals of the resistor.

5. Apparatus according to Claim 4, characterized in that the amplification and integration means comprise an integrator (DET.S) whose time constant is equal to the duration of a plurality of operating cycles of the pump.

6. Apparatus according to Claim 5, characterized in that the integrator (DET.S) is associated with a threshold detector.

7. Apparatus according to Claim 4, 5 or 6, characterized in that the first sensor means further includes manual adjustment means (PI) for varying the slope and/or the height of a monotonic curve relating said electric signal to the mean current.

8. Apparatus according to Claim 7, characterized in that the first sensors further include limiter means (LIM) for applying a ceiling to said monotonic curve. - 19 - 102100/2

9. Apparatus according to any of Claims 4 to 8, characterized in that the first sensor means are designed to vary the value of an electric current that is allowed to flow along a current path (Tl, R4).

10. Apparatus according to Claim 9, characterized in that the means for controlling the width of the signal include a capacitor (CI 3) capable during the return stages of being charged or discharged by a current (i) flowing along said path, and threshold means (P12) receiving the voltage across the. terminals of the capacitor.

11. Apparatus according to Claim 10, characterized in that it is provided with switching means (112) that are capable, as a function of the direction of rotation of the motor, of connecting the capacitor (C13) alternately to said current path and to a fixed value resistance (R21).

12. Apparatus according to Claim 11, characterized in that the switching means are controlled by a circuit (BS) connected to the end-of-stroke sensors of a moving element of the pump.

13. Apparatus according to any of the Claims 3 to 12, for a motor without a commutator, characterized in that the second sensor means of said apparatus comprise at least one magnetic sensor (EH) mounted on a stator of the motor and comparator means (A12, A13) for generating logic position signals (SA1, SA2) that vary with the sign of the signals provided by the magnetic sensor(s).

14. Apparatus according to Claim 13, as dependent on Claim 11, characterized in that the switching means (110, 111) of said apparatus are also connected to the second sensor means for modifying the time sequencing of the position signals (Q0-Q3) when the direction of rotation of the motor reverses.

15. Apparatus according to Claim 13 or 14, characterized in that the magnetic sensors (EH) of said apparatus are positioned on the stator in such a manner that the edges that delimit the position signals are situated at median angular positions between the equilibrium positions (VI, V2, Rl, R2) of the motor. 102100/2

16. Apparatus according to Claim 15, for a two-phase motor, characterized in that the magnetic sensors (EH) of said apparatus are positioned on the stator at an angular offset of 22.5° relative to the equilibrium positions of the motor.

17. Apparatus according to any of Claims 13 -to 16, characterized in that the magnetic sensors are Hall effect cells (EH).

18. Apparatus according to any of Claims 3 to 17, characterized in that the logic circuit (DEC) generates current switching signals (R0-R3) causing current to be applied to the winding(s) of the motor starting from or until an instant (tO1) corresponding to a median angular position between two equilibrium positions of the motor.

19. Apparatus according to any of Claims 3 to 18, comprising an external first portion housing the first sensor means, and an implanted second portion housing the second sensor means, the means for controlling the width of the signal, the logic circuit, and the power circuit, together with flexible cables interconnecting the two portions and constituting a link (Bl, B'l) between the first sensor means and the means for controlling the width of the signal.

20. Apparatus according to Claim 19, characterized in that, during ejection stages, said link (Bl, B'l) is used to convey information concerning the beginning of an ejection stage generated in the second portion of the apparatus to the first portion thereof.

21. Apparatus according to Claim 20, as dependent on Claim 12, characterized in that the second portion of said apparatus includes switching means (T10) controlled by the output of the circuit (BS) connected to the end-of-stroke sensors for the purpose of putting said link at a determined voltage during ejection stages, and in that the first portion of the apparatus includes comparator means (A5) receiving the voltage on said link. For the Applicants, HN AND PARTNERS