Disposable magnetically levitated centrifugal blood pump: design and in vitro performance

A magnetically levitated (MagLev) centrifugal blood pump (CBP) with a disposable pump head has been designed to realize a safe, easy-to-handle, reliable, and low-cost extracorporeal blood pump system. It consisted of a radial magnetic-coupled driver with a magnetic bearing having a two-degree freedom control and a disposable pump head unit with a priming volume of 24 mL. The easy on-off disposable pump head unit was made into a three-piece system consisting of the top and bottom housings, and the impeller-rotor assembly. The size and weight of the disposable pump unit were 75 mm x 45 mm and 100 g, respectively. Because the structure of the pump head unit is easily attachable and removable, the gap between the electromagnets of the stator and the target material in the rotor increased to 1.8 mm in comparison to the original integrated bearing system of 1.0 mm. The pump performance, power requirements, and controllability of the magnetic bearing revealed that from 1400 to 2400 rpm, the pump performance remained fairly unchanged. The amplitudes of the X- and Y-axis rotor oscillation increased to +/- 24 microm. The axial displacement of the rotor, 0.4 mm, toward the top housing was also observed at the pump rpm between 1400 and 2400. The axial and rotational stiffness of the bearing were 15.9 N/mm and 4.4 Nm/rad, respectively. The MagLev power was within 0.7 Watts. This study demonstrated the feasibility of a disposable, magnetically suspended CBP as the safe, reliable, easy-to-handle, low-cost extracorporeal circulation support device.     

A compact highly efficient and low hemolytic centrifugal blood pump with a magnetically levitated impeller

A magnetically levitated (maglev) centrifugal blood pump (CBP), intended for use as a ventricular assist device, needs to be highly durable and reliable for long-term use without any mechanical failure. Furthermore, maglev CBPs should be small enough to be implanted into patients of various size and weight. We have developed a compact maglev CBP employing a two-degree-of-freedom controlled magnetic bearing, with a magnetically suspended impeller directly driven by an internal brushless direct current (DC) motor. The magnetic bearing actively controls the radial motion of the impeller and passively supports axial and angular motions using a permanent magnet embedded in the impeller. The overall dimensions of the maglev CBP are 65 mm in diameter and 40 mm in height. The total power consumption and pump efficiency for pumping 6 L/min against a head pressure of 105 mm Hg were 6.5 W and 21%, respectively. To evaluate the characteristics of the maglev CBP when subjected to a disturbance, excitation of the base, simulating the movement of the patient in various directions, and the sudden interception of the outlet tube connected with the pump in a mock circulatory loop, simulating an unexpected kink and emergent clamp during a heart surgery, were tested by monitoring the five-degree-of-freedom motion of the impeller. Furthermore, the hemolytic characteristics of the maglev CBP were compared with those of the Medtronic Biomedicus BPX-80, which demonstrated the superiority of the maglev CBP.     

Development of a disposable magnetically levitated centrifugal blood pump (MedTech Dispo) intended for bridge-to-bridge applications--two-week in vivo evaluation

Last year, we reported in vitro pump performance, low hemolytic characteristics, and initial in vivo evaluation of a disposable, magnetically levitated centrifugal blood pump, MedTech Dispo. As the first phase of the two-stage in vivo studies, in this study we have carried out a 2-week in vivo evaluation in calves. Male Holstein calves with body weight of 62.4–92.2 kg were used. Under general anesthesia, a left heart bypass with a MedTech Dispo pump was instituted between the left atrium and the descending aorta via left thoracotomy. Blood-contacting surface of the pump was coated with a 2-methacryloyloxyethyl phosphorylcholine polymer. Post-operatively, with activated clotting time controlled at 180–220 s using heparin and bypass flow rate maintained at 50 mL/kg/min, plasma-free hemoglobin (Hb), coagulation, and major organ functions were analyzed for evaluation of biocompatibility. The animals were electively sacrificed at the completion of the 2-week study to evaluate presence of thrombus inside the pump,together with an examination of major organs. To date, we have done 13 MedTech Dispo implantations, of which three went successfully for a 2-week duration. In these three cases, the pump produced a fairly constant flow of 50 mL/Kg/min. Neurological disorders and any symptoms of thromboembolism were not seen. Levels of plasma-free Hb were maintained very low. Major organ functions remained within normal ranges. Autopsy results revealed no thrombus formation inside the pump. In the last six cases, calves suffered from severe pneumonia and they were excluded from the analysis. The MedTech Dispo pump demonstrated sufficient pump performance and biocompatibility to meet requirements for 1-week circulatory support. The second phase (2-month in vivo study) is under way to prove the safety and efficacy of MedTech Dispo for 1-month applications.     

Dynamic characteristics of a magnetically levitated impeller in a centrifugal blood pump

Centrifugal blood pumps that employ hybrid active/passive magnetic bearings to support noncontact impellers have been developed in order to reduce bearing wear, pump size, the power consumption of the active magnetic bearing, and blood trauma. However, estimates made at the design stage of the vibration of the impeller in the direction of passive suspension during pump operation were inaccurate, because the influence of both the pumping fluid and the rotation of the impeller on the dynamic characteristics was not fully recognized. The purpose of this study is to investigate the dynamic characteristics in a fluid of a magnetically levitated rotating impeller by measuring both the frequency response to sinusoidal excitation of the housing over a wide frequency range and the displacement due to input of a pulsatile flow during left ventricular (LV) assist. The excitation tests were conducted under conditions in which the impeller was levitated in either air or water, and with or without rotation. The experimental and analytical results indicate that vibration of the impeller due to the external force in water was decreased, compared with that in air due to the hydraulic force of water. The axial resonant frequency rose quadratically with rotational speed, and the tilt mode had two resonant frequencies while rotating due to the gyroscopic effect. With the pump inserted into a mock systemic circulatory loop, the dynamic stability of the impeller when pulsatile pressure was applied during LV assist was verified experimentally. The amplitudes of vibration in response to the pulsatile flow in the passively constrained directions were considerably smaller in size than the dimensions of initial gaps between the impeller and the pump housing.     

Magnetically suspended centrifugal blood pump with an axially levitated motor

The longevity of a rotary blood pump is mainly determined by the durability of its wearing mechanical parts such as bearings and seals. Magnetic suspension techniques can be used to eliminate these mechanical parts altogether. This article describes a magnetically suspended centrifugal blood pump using an axially levitated motor. The motor comprises an upper stator, a bottom stator, and a levitated rotor-impeller between the stators. The upper stator has permanent magnets to generate an attractive axial bias force on the rotor and electric magnets to control the inclination of the rotor. The bottom stator has electric magnets to generate attractive forces and rotating torque to control the axial displacement and rotation of the rotor. The radial displacement of the rotor is restricted by passive stability. A shrouded impeller is integrated within the rotor. The performance of the magnetic suspension and pump were evaluated in a closed mock loop circuit filled with water. The maximum amplitude of the rotor displacement in the axial direction was only 0.06 mm. The maximum possible rotational speed during levitation was 1,600 rpm. The maximum pressure head and flow rate were 120 mm Hg and 7 L/min, respectively. The pump shows promise as a ventricular assist device.     

HeartQuest ventricular assist device magnetically levitated centrifugal blood pump

Improvements in implantable ventricular assist device (VAD) performance will be required to obtain patient outcomes that are comparable with those of heart transplantation. The HeartQuest VAD (WorldHeart, Oakland, CA, U.S.A.) is an advanced device, with full magnetic suspension of the rotor, designed to address specific clinical shortcomings in existing devices and to maximize margins of safety and performance for an implantable assist device. The device dimensions are 35 x 75 mm, with a total weight of 440 g. The system was designed using extensive computer modeling of device function; a total of two iterations of device prototypes were built before building the clinical version. Animal study results have been very promising, with over 30 calf studies completed. Plasma-free hemoglobin levels returned to preoperative levels, and other hematology results were in the normal ranges. Highlights include clean surfaces seen in a 116-day experiment with no anticoagulation after day 43. Feasibility clinical trials are planned to start in 2006.     

Centrifugal blood pump with a magnetically suspended impeller

A centrifugal blood pump with a magnetically suspended impeller has been developed. It has a single inlet and outlet, and it generates centrifugal forces by the rotating impeller. The fluid-dynamical design for inflow and outflow through the impeller leads to elimination of the axial force and unbalanced radial force acting on the impeller. Consequently, three-component control systems, instead of five-component ones, are enough to position the impeller. The magnetically suspended impeller rotates by the magnetic coupling with the permanent magnets embedded in the outer rotator of the motor. This pump has enough performance to function as a blood pump. Further research on the null-power magnetic suspension and the generation of an efficient rotating magnetic field is in progress.     

Detection of left ventricle function from a magnetically levitated impeller behavior

The magnetically levitated (Mag-Lev) centrifugal rotary blood pump (CRBP) with two-degrees-of-freedom active control is promising for safe and long-term support of circulation. In this study, Mag-Lev CRBP controllability and impeller behavior were studied in the simulated heart failure circulatory model. A pneumatically driven pulsatile blood pump (Medos VAD [ventricular assist device]-54 mL) was used to simulate the left ventricle (LV). The Mag-Lev CRBP was placed between the LV apex and aortic compliance tank simulating LV assistance. The impeller behavior in five axes (x, y, z, theta, and phi) was continuously monitored using five eddy current sensors. The signals of the x- and y-axes were used for feedback active control, while the behaviors of the other three axes were passively controlled by the permanent magnets. In the static mock circuit, the impeller movement was controlled to within +/-10-+/-20 microm in the x- and y-axes, while in the pulsatile circuit, LV pulsation was modulated in the impeller movement with the amplitude being 2-22 microm. The amplitude of impeller movement measured at 1800 rpm with the simulated failing heart (peak LV pressure [LVP] = 70 mm Hg, mean aortic pressure [AoP(mean)] = 55 +/- 20 mm Hg, aortic flow = 2.7 L/min) was 12.6 microm, while it increased to 19.2 microm with the recovered heart (peak LVP = 122 mm Hg, AoP(mean) = 100 +/- 20 mm Hg, aortic flow = 3.9 L/min). The impeller repeated the reciprocating movement from the center of the pump toward the outlet port with LV pulsation. Angular rotation (theta, phi) was around +/-0.002 rad without z-axis displacement. Power requirements ranged from 0.6 to 0.9 W. Five-axis impeller behavior and Mag-Lev controller stability were demonstrated in the pulsatile mock circuit. Noncontact drive and low power requirements were shown despite the effects of LV pulsation. The impeller position signals in the x- and y-axes reflected LV function. The Mag-Lev CRBP is effective not only for noncontact low power control of the impeller, but also for diagnosis of cardiac function noninvasively.     

Miniaturization of a magnetically levitated axial flow blood pump

This article introduces a unique miniaturization process of a magnetically levitated axial flow blood pump from a functional prototype to a pump suitable for animal trials. Through COMSOL three-dimensional finite element analysis and experimental verification, the hybrid magnetic bearings of the pump have been miniaturized, the axial spacing between magnetic components has been reduced, and excess material in mechanical components of the pump was reduced. Experimental results show that the pump performance was virtually unchanged and the smaller size resulted in the successful acute pump implantation in calves.     

Flow rate and pressure head estimation in a centrifugal blood pump

Considering a miniaturization of percutaneous cardiopulmonary support (PCPS) system, we have developed a pump system that can estimate flow rate and pressure head in the circuit. A commercially available centrifugal blood pump (C1E3, Kyocera Co., Ltd., Kyoto, Japan) with an exclusive motor driver was used in the experiment. In this system, the motor revolutions per minute and torque were used to derive an estimation equation. Taking into consideration fluid viscosity, the accuracy of the estimation was increased. Bovine blood was used to evaluate the system in vitro. The average maximum difference between the estimated flow rate and the measured flow rate was 0.56 L/min (for 98% accuracy). The average difference between the estimated pressure head and the measured pressure head was 30.7 mm Hg (for 94.2% accuracy). Because of the stability of the estimations, we believe this system has the possibility of use for real-time monitoring and miniaturization of PCPS system.     

Influence of radial clearance and rotor motion to hemolysis in a journal bearing of a centrifugal blood pump

Hemolysis due to narrow clearance of noncontact bearings is a critical problem for rotary blood pumps. We developed a centrifugal blood pump with a magnetic and hydrodynamic hybrid bearing, and found that the hemolysis in the narrow clearance depends not only on the clearance size, but also on the rotor stability. In this study, we quantified the relation between the hemolysis, radial clearance (c), and rotor stability through the measurement of the rotor motion and hemolysis. As a result, it was confirmed that the rotor of the current pump is stabilized within the oscillation of 20 microm in blood, and the hemolysis decreases with increase in the c, which is the opposite in the unstable rotor motion with the previous pump. In order to theoretically discuss this hemolysis tendency, we implemented hemolysis estimation in the c according to hydrodynamics and hemodynamics. This estimation can represent the measured hemolysis tendency, and revealed that the flow rate has large influence on the hemolysis in the c.     

Measurement of the rotor motion and corresponding hemolysis of a centrifugal blood pump with a magnetic and hydrodynamic hybrid bearing

This study proposed a centrifugal blood pump with a novel magnetic and hydrodynamic hybrid passive bearing, which consisted of a plain journal bearing for radial stability and a permanent magnetic bearing for axial and tilting stability. We measured the radial motion of the bearing and performed hemolysis tests for the different radial clearance sizes. In the results, it appeared that the radial motion had two modes: the stable center mode, in which the radius of the radial motion rapidly converged to less than 20 microm; and the unstable circle mode, in which the rotor suspension linearly increased with the rotation speed. It also appeared that the pumps with the radial clearance of 80 microm caused more hemolysis than with the smaller clearance sizes in the circle mode. The circle mode was avoidable by the higher rotation or the asymmetric pump structure, but the mechanism of hemolysis in this mode was still unclear.     

An ultradurable and compact rotary blood pump with a magnetically suspended impeller in the radial direction

A magnetically suspended centrifugal blood pump has been developed with a self-bearing motor for long-term ventricular assist systems. The rotor of the self-bearing motor is not only actively suspended in the radial direction, but also is rotated by an electromagnetic field. The pump has a long lifetime because there are no mechanical parts such as seals and motor bearings. An outer rotor mechanism was adopted for the self-bearing motor. The stator was constructed in the central space of the motor. The rotor shaped thin ring was set at the circumferential space of the stator. Six vanes were extended from the upper surface of the rotor toward the center of the pump to construct an open-type impeller. The outer diameter and the height of the impeller are 63 mm and 34 mm, respectively. The magnetic bearing method and the servomotor mechanism were adopted to levitate and rotate the rotor. Radial movements of the rotor and rotation are controlled actively by using electromagnets in the stator. Axial movement and tilt of the rotor are restricted by passive stability to simplify the control. The radial gap between the rotor and the stator is 1 mm. A closed-loop circuit filled with water was used to examine basic performance of the pump. Maximum flow rate and pressure head were 8 L/min and 200 mm Hg, respectively. Maximum amplitude of radial displacement of the impeller was 0.15 mm. The impeller could be suspended completely without touching the casing wall during the entire pumping process. Power consumption of the pump was only 9.5 W to produce a flow rate of 5 L/min against a pressure head of 100 mm Hg. We conclude that the pump has sufficient performance for the implantable ventricular assist system.     

Design considerations of volute geometry of a centrifugal blood pump

This article compares two different design techniques that are conventionally used in the design of volutes for centrifugal pumps. The imbalanced forces due to the geometry of the volute need to be taken into consideration especially in centrifugal blood pumps with magnetically suspended impeller. A reduction of these forces can reduce the instability of the impeller motion as well as the power needed to counteract its influence. Volutes using the constant angular momentum (CAM) and the constant mean velocity (CMV) methods were developed and modeled numerically. The computational results on the effect of volute geometry on the performance of a centrifugal blood pump impeller for six different volutes are presented here. For volutes designed using the CAM method, model B (volute expansion angle of 3 degrees ) had the lowest radial force of 0.26 N while the pressure head generated was 12,900 Pa. For volutes designed using the CMV method, model F (1.6 m/s) had the lowest imbalanced force of 0.45 N. However, the pressure developed by this pump was also one of the lowest at 10,652 Pa. Furthermore, when the peak scalar stresses and the mean exposure time of particles for all designs were determined using Lagrangian particle tracking method, it was observed that in general, the peak scalar stresses in CAM designed volutes are lower than those designed using CMV method. The mean exposure time of particles in the pump ranged from 400 to 500 ms. The simulation results showed that the volute designed using CAM method was superior to that of a CMV volute in terms of the magnitude of the radial force and the peak scalar stresses for the same pressure head generated. Results show that the design of volutes for blood pumps should go beyond conventional empirical methods to obtain optimal results.     

A hydrodynamically suspended, magnetically sealed mechanically noncontact axial flow blood pump: design of a hydrodynamic bearing

To overcome the drive shaft seal and bearing problem in rotary blood pumps, a hydrodynamic bearing, a magnetic fluid seal, and a brushless direct current (DC) motor were employed in an axial flow pump. This enabled contact-free rotation of the impeller without material wear. The axial flow pump consisted of a brushless DC motor, an impeller, and a guide vane. The motor rotor was directly connected to the impeller by a motor shaft. A hydrodynamic bearing was installed on the motor shaft. The motor and the hydrodynamic bearing were housed in a cylindrical casing and were waterproofed by a magnetic fluid seal, a mechanically noncontact seal. Impeller shaft displacement was measured using a laser sensor. Axial and radial displacements of the shaft were only a few micrometers for motor speed up to 8500 rpm. The shaft did not make contact with the bearing housing. A flow of 5 L/min was obtained at 8000 rpm at a pressure difference of 100 mm Hg. In conclusion, the axial flow blood pump consisting of a hydrodynamic bearing, a magnetic fluid seal, and a brushless DC motor provided contact-free rotation of the impeller without material wear.     

Computational fluid dynamics investigation of a centrifugal blood pump

Abstract:  In the development of a ventricular assist device, computational fluid dynamics (CFD) analysis is an efficient tool to obtain the best design before making the final prototype. In this study, different designs of a centrifugal blood pump were developed to investigate flow characteristics and performance. This study assumed the blood flow as being an incompressible homogeneous Newtonian fluid. A constant velocity was applied at the inlet; no slip boundary conditions were applied at device wall; and pressure boundary conditions were applied at the outlet. The CFD code used in this work was based on the finite volume method. In the future, the results of CFD analysis can be compared with flow visualization and hemolysis tests.     

Mechanical antithrombogenic properties by vibrational excitation of the impeller in a magnetically levitated centrifugal blood pump

Mechanical circulatory support devices have been used clinically for patients with heart failure for over 10 years. However, thrombus formation inside blood pumps remains a risk to patient life, causing pump failure and contributing to neurological damage through embolization. In this article, we propose a method for preventing thrombus formation by applying vibrational excitation to the impeller. We evaluate the ability of this method to enhance the antithrombogenic properties of a magnetically levitated centrifugal blood pump and ensure that the impeller vibration does not cause undue hemolysis. First, 3 vibrational conditions were compared using an isolated pump without a mock circulation loop; the vibrational excitation frequencies and amplitudes for the impeller were set to (a) 0 Hz‐0 μm, (b) 70 Hz‐10 μm, and (c) 300 Hz‐2.5 μm. The motor torque was measured to detect thrombus formation and obtain blood coagulation time by calculating the derivative of the torque. Upon thrombus detection, the pump was stopped and thrombi size were evaluated. The results showed an increase in the blood coagulation time and a decrease in the rate of thrombus formation in pumps with the impeller vibration. Second, an in vitro hemolysis test was performed for each vibrational condition to determine the effect of impeller vibration on hemolysis. The results revealed that there was no significant difference in hemolysis levels between each condition. Finally, the selected vibration based on the above test results and the non‐vibration as control were compared to investigate antithrombogenic properties under the continuous flow condition. The blood coagulation time and thrombi size were investigated. As a result, vibrational excitation of the impeller at a frequency of 300 Hz and amplitude of 2.5 μm was found to significantly lengthen clotting time, decreasing the rate of pump thrombus compared to the non‐vibration condition. We indicate the potential of impeller vibration as a novel mechanical antithrombogenic mechanism for rotary blood pumps.     

Evaluation of hydraulic radial forces on the impeller by the volute in a centrifugal rotary blood pump

In many state-of-the-art rotary blood pumps for long-term ventricular assistance, the impeller is suspended within the casing by magnetic or hydrodynamic means. For the design of such suspension systems, profound knowledge of the acting forces on the impeller is crucial. Hydrodynamic bearings running at low clearance gaps can yield increased blood damage and magnetic bearings counteracting high forces consume excessive power. Most current rotary blood pump devices with contactless bearings are centrifugal pumps that incorporate a radial diffuser volute where hydraulic forces on the impeller develop. The yielding radial forces are highly dependent on impeller design, operating point and volute design. There are three basic types of volute design--singular, circular, and double volute. In this study, the hydraulic radial forces on the impeller created by the volute in an investigational centrifugal blood pump are evaluated and discussed with regard to the choice of contactless suspension systems. Each volute type was tested experimentally in a centrifugal pump test setup at various rotational speeds and flow rates. For the pump's design point at 5 L/min and 2500 rpm, the single volute had the lowest radial force (～0 N), the circular volute yielded the highest force (～2 N), and the double volute possessed a force of approx. 0.5 N. Results of radial force magnitude and direction were obtained and compared with a previously performed computational fluid dynamics (CFD) study.     

A novel design of spiral groove bearing in a hydrodynamically levitated centrifugal rotary blood pump

Good washout is very important in spiral groove bearing (SGB) designs when applied to blood pumps due to the micrometer scales of lubrication films and groove depths. To improve washout, flow rate or leakage through SGBs should be as large as possible. However, this special goal violates conventional SGB designs in which no leakage is desired as the leakage would decrease load-carrying capacity significantly. So, a design concept is formed fulfilling the two goals of high load-carrying capacity and large flow rate: let groove width decrease along flow path and the mating surface of the rotor rotate with a direction facilitating the flow through the grooves. Under this concept, a novel SGB is designed, contrary to conventional ones, with groove width decreasing with increasing spiral radius. This SGB is mounted on the motionless upper plate of our designed centrifugal blood pump, with the mating surface of rotor rotating with a direction facilitating the outward flow. To assess SGB designs, a characteristic plane is originally presented relating to pressure-normalized load-carrying capacity and flow rate. Comparisons between various kinds of SGB designs are made, and computational fluid dynamics (CFD) results are plotted in this characteristic plane from which load/flow performances can be directly read out. CFD and comparison results show that the new designs have superior load/flow characteristics. However, the impact of SGB designs upon hemolysis/thrombus formation is still to be verified according to the concept presented.     

Experimental study on shear stress distributions in a centrifugal blood pump

Wall shear stress on the pump casing cover was measured using a surface-mounted hot-film sensor. In addition, the shear stress distribution in the pump was qualitatively investigated by means of oil-film visualization. The characteristics of shear stress in the pump are discussed, including the results on the oil-film visualization. The centrifugal blood pump used was a Nikkiso HPM-15 (Nikkiso Co., Ltd, Tokyo, Japan). The hot-film measurement indicated that the shear stress was approximately proportional to the rotating speed, and exceeded 300 Pa when r/R > 0.5 at 3000 rpm. The circumferential average shear stress on the casing cover was of the same order as the characteristic stress sigma obtained from the pump axial torque. These results suggest that the shear stress on the casing cover can be used to evaluate the characteristic shear stress in the pump.