Artificial Organs
`26(5):420–429, Blackwell Publishing, Inc.
`© 2002 International Society for Artificial Organs
`
`Implications for the Establishment of Accelerated Fatigue
`Test Protocols for Prosthetic Heart Valves
`
`*Kiyotaka Iwasaki, *Mitsuo Umezu, *Kazuo Iijima, and †Kou Imachi
`
`*Department of Mechanical Engineering, School of Science and Engineering, Waseda University; and †Department of
`Biomedical Engineering, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
`
`Abstract: The goal of this research is to establish a reliable
`methodology for accelerated fatigue tests of prosthetic
`heart valves. A polymer valve was the subject, and the
`influence of various drive parameters on durability was
`investigated in three different machines. Valve lifetime
`was notably shortened by increasing the cyclic rate or
`stroke even though the maximum pressure difference at
`valve closure was maintained at 120 mm Hg. These results
`demonstrate that adjustment of the maximum transvalvu-
`lar pressure is not sufficient to ensure tests are conducted
`
`under the same conditions and indicate that measurement
`of the dynamic load would be more efficacious. Moreover,
`the locations of tears sustained in the accelerated tests
`differed from those encountered in an animal experiment
`although in both cases the locations were entirely consis-
`tent with the areas of strain concentration revealed by
`finite element analysis. These findings should be discussed
`during a revision of ISO 5840. Key Words: Accelerated
`fatigue test—Durability—Fracture—ISO 5840—Polymer
`valve—Jellyfish valve.
`
`Accurate estimation of durability in a timely man-
`ner is one of the most important unresolved issues in
`the basic research of artificial organs. ISO 5840 (Car-
`diovascular Implants) prescribes guidelines for ac-
`celerated fatigue test methods applied to heart
`valves (1), the assessment of durability by acceler-
`ated cycling having been widely accepted as an es-
`sential component in the developmental stage of
`prosthetic heart valves. However, in bioprosthetic
`heart valves especially, investigators have reported
`varying degrees of success in obtaining a correlation
`between the tears and perforations observed in clini-
`cal cases with those failure modes observed during in
`vitro accelerated fatigue tests (2–4). Furthermore, in
`the case of polymer valves that have been developed
`as alternatives to mechanical and bioprosthetic
`valves for use in artificial hearts, the optimal proto-
`col for accelerated fatigue tests has yet to be estab-
`lished. The authors are accumulating fundamental
`data on heart valve durability by means of three dif-
`ferent types of accelerated fatigue testers. The ulti-
`
`Received October 2001.
`Address correspondence and reprint requests to Dr. Kiyotaka
`Iwasaki, Department of Mechanical Engineering, Waseda Uni-
`versity, 3-4-1 #58-322 Ohkubo, Shinjuku, Tokyo 169-8555, Japan.
`E-mail: iwasaki@umezu.mech.waseda.ac.jp
`
`mate goal of this research is to establish a reliable
`methodology for accelerated fatigue testing of pros-
`thetic heart valves. The aim of the current study is to
`investigate the influence of drive parameters on du-
`rability under the test conditions recommended by
`ISO 5840 and also to compare the fracture patterns
`of heart valves subjected to in vitro accelerated fa-
`tigue tests with an animal model.
`
`MATERIALS AND METHODS
`Valves used in this study
`A polymer valve known as the Jellyfish valve,
`which has been developed for use in artificial hearts
`(5–6), was used as the test subject throughout this
`research. The main reasons why this valve was em-
`ployed are as follows. First, assessments of durability
`and the locations of membrane fracture have been
`obtained in experiments in goats in which Jellyfish
`valves were incorporated into blood pumps as shown
`in Fig. 1 (7,8); these results provide benchmarks for
`the accelerated fatigue tests. Second, the durability
`of the Jellyfish valve in these goat experiments was
`312 days, or approximately 10 months. This lifetime
`was considered to be more convenient, with regard
`to the prospective duration of the accelerated fatigue
`tests, than the lifetime that is typical of currently
`
`420
`
`WATERS TECHNOLOGIES CORPORATION
`EXHIBIT 1013
`
`PAGE 1 OF 10
`
`

`

`ACCELERATED FATIGUE TESTS FOR HEART VALVES
`
`421
`
`outer diameter of the valve seat was widened from
`20 mm to 28 mm to provide chuck-area for mounting
`in the test chambers; however, the flow field of this
`modified valve is the same as that of the normal (20
`mm) valve. The thickness of the Jellyfish valve mem-
`brane can be varied by changing the insertion vol-
`ume of liquid K-III that enters the casting mold. The
`thickness of the membrane in the Jellyfish valve that
`was fractured in the animal experiment was around
`60 ␮m. Thus, to ensure correlation between the ani-
`mal model and the in vitro accelerated fatigue tests,
`60 ␮m membranes were fabricated. And, further-
`more, only membranes with a maximum thickness
`deviation of ±10 ␮m at 24 measuring points (8 points
`each in the inner, middle, and outer areas of the
`membrane) were chosen in an effort to eliminate
`membrane thickness as an independent variable in
`the durability tests. The typical behavior of the Jel-
`lyfish valve in a pulse duplicator under physiological
`conditions is shown in Fig. 2.
`
`Accelerated fatigue testers for heart valves
`Three different types of accelerated fatigue testers
`were employed in this study as described below. A
`Helmholtz-type accelerated fatigue tester (Helm-
`holtz-Institute for Biomedical Engineering, Aachen,
`Germany) (10) is shown in Fig. 3. In this tester,
`transvalvular flow is produced by the combination of
`a rotary pump, located below, and a rotating disk
`
`FIG. 2. Typical behavior of the Jellyfish valve membrane in the
`outlet position of a pulse duplicator is shown. Mean transvalvular
`flow rate was adjusted to 5 L/min against a mean aortic pressure
`of 100 mm Hg.
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`FIG. 1. A membrane fracture in the Jellyfish valve incorporated
`into a paracoporeal total replacement-type, pneumatically driven
`blood pump after 312 days of pumping is shown. The valve was
`explanted from the outlet position of the left-side pump.
`
`available clinical heart valves. Therefore, one experi-
`mental plot can be made available within 1 month at
`a cycling rate of 20 Hz whereas, generally speaking,
`a valve with an expected clinical durability of 10
`years would require about 6 months to yield terminal
`results under the same accelerated conditions. Third,
`thanks to well-established fabrication techniques
`and quality control procedures, these valves can be
`produced with a high degree of uniformity. Fourth,
`the polymer leaflet material exhibits viscoelastic be-
`havior that is not unlike that of the tissues used in
`bioprosthetic leaflets (9); therefore, it is reasonable
`to assume that data obtained from these polymer
`valves would be useful for understanding fatigue test
`results on bioprosthetic heart valves. For these rea-
`sons, the Jellyfish valve is seen not only as suitable
`but ideal in the context of this study.
`The Jellyfish valve consists of a flexible membrane
`and a rigid valve seat, both of which are fabricated
`by casting techniques. The membrane is made of a
`copolymer of segmented polyurethane named K-III
`(Nippon Zeon Co. Ltd., Tokyo, Japan) that pos-
`sesses excellent blood compatibility. The valve seat
`is made of a two-component, room-temperature vul-
`canizing urethane (Quinnate CR330, Ciba Specialty
`Chemicals K.K., Tokyo, Japan) and then coated with
`K-III. Finally, the membrane is bonded to the valve
`seat at a central location, again using K-III. The 20
`mm size (which refers to the diameter of the mem-
`brane as well as the valve seat) was chosen, this hav-
`ing been used in the animal experiments. In the
`valves for use in the accelerated fatigue tests, the
`
`PAGE 2 OF 10
`
`

`

`422
`
`K. IWASAKI ET AL.
`
`FIG. 3. A Helmholtz Institute-type accelerated fatigue tester for prosthetic heart valves is shown.
`
`with a flow passage comprising one-third of the total
`cyclic area and located in the casing. When the ro-
`tating flow passage encounters the test chamber,
`flow rapidly opens the valve membrane, and it closes
`as soon as the flow passage rotates beyond the test
`chamber. Accordingly, the effective systolic fraction
`is approximately 33%. The maximum pressure gra-
`dient following valve closure was adjusted to 120 mm
`Hg as recommended by ISO 5840 (1). In addition to
`this condition, the mean outlet pressure of the valve
`was adjusted to 115 mm Hg because this fatigue
`tester allowed a variety of dynamic conditions to be
`obtained under a given maximum pressure drop.
`Moreover, the mean transvalvular flow rate was ad-
`justed to 5.0 L/min to ensure that the opening be-
`havior of the membrane was similar to that at physi-
`ological cycle rates. The temperature of the working
`fluid was maintained at 37°C throughout the study.
`Typical pressure waveforms in this machine are
`shown in Fig. 4. In this tester, durability tests were
`conducted under the cycle rates of 400, 500, and 600
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`bpm. Thus, the influence of cyclic rate on durability,
`for a fixed maximum pressure difference at valve
`closure, was examined.
`Figure 5 shows a modified version of the commer-
`cially available Rowan Ash accelerated fatigue tester
`(Rowan Ash Ltd., Sheffield, England). The original
`tester was designed without a compliance element.
`The upper casing was flat, thus failing to model the
`elastic effects of the aorta. The preliminary experi-
`mental study indicated that this limitation resulted in
`an insufficient opening motion of the Jellyfish valve
`membrane. Therefore, the upper casing of the tester
`was modified to incorporate an air compliance
`chamber as shown in Fig. 5. Changes in pressure
`waveforms as a function of air volume under the
`same maximum pressure difference of 120 mm Hg at
`valve closure are shown in Fig. 6. In order to keep
`the same maximum pressure difference of 120 mm
`Hg, the amplitude of the sinusoidal stroke was con-
`trolled as shown in Fig. 7. The drive amplitude had
`to be increased after the inclusion of the air compli-
`
`PAGE 3 OF 10
`
`

`

`ACCELERATED FATIGUE TESTS FOR HEART VALVES
`
`423
`
`FIG. 4. Simultaneous pressure waveforms are shown in the
`Helmholtz-type accelerated fatigue tester under a cycle rate of
`500 rpm. The maximum pressure difference at valve closure was
`adjusted to 120 mm Hg under a mean flow of 5.0 L/min. Mean
`outlet pressure was adjusted to 115 mm Hg.
`
`ance element. Movement of the membrane is in-
`duced by the inertia of the fluid as the valves are
`sinusoidal displaced, on hollow pistons, by the stroke
`of a linear motor. The temperature of the test fluid
`did not need to be controlled because it was in a
`closed circuit and remained at room temperature
`(around 20°C). The cycle rate was adjusted to 1,200
`bpm, and the maximum pressure gradient at valve
`closure was maintained at 120 mm Hg. Then, the
`influence of air compliance on durability was inves-
`tigated.
`The Tsinghua-type accelerated fatigue tester
`(Tsinghua University, Beijing, China) is shown in
`Fig. 8. The motion of the valve membrane is ensured
`by the sinusoidal stroke of the linear motor located
`on the upper side of the test chamber. The valve is
`mounted in a holder and connected to the axially
`vibrating rod. The Tsinghua University machine has
`an open loop (the working fluid in the inflow side of
`the valve is open to atmosphere) unlike the Rowan
`Ash system. The Tsinghua University machine has
`the important advantage that the dynamic load act-
`ing on the valve can be measured by installing a load
`cell into the oscillating rods. The authors have been
`conducting preliminary experiments to investigate
`the influence of dynamic load on durability, the de-
`tails of which will be discussed elsewhere. In this
`paper, the maximum pressure difference at valve clo-
`sure was adjusted to 120 mm Hg to maintain equiva-
`lence among the three different accelerated test sys-
`tems. The temperature of the circulating water was
`
`FIG. 5. A modified version of the Rowan Ash accelerated fatigue
`tester for prosthetic heart valves is shown. A compliance element
`was added to the original system.
`
`maintained at 37°C by a heater. The test was con-
`ducted under a drive rate of 1,200 bpm. Simulta-
`neous pressure waveforms are shown in Fig. 9.
`
`RESULTS
`Influence of drive parameters on lifetime
`Accelerated fatigue tests were conducted by three
`different machines under the same maximum pres-
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`PAGE 4 OF 10
`
`

`

`424
`
`K. IWASAKI ET AL.
`
`FIG. 7. Influence of inclusion of the air-compliance element on
`drive amplitude under three typical transvalvular pressures at
`closure are shown. In order to maintain the specified pressure
`drop, the drive amplitude was increased.
`
`creased stroke amplitude shortened valve lifetime
`despite inclusion of the compliance element in the
`outflow section which should have approached a
`more realistic simulation of normal valve motion. In
`
`FIG. 6. Changes in pressure waveforms are shown as a function
`of air-compliance volume at the outlet position of the valve. The
`maximum pressure difference at valve closure was maintained at
`120 mm Hg under a drive rate of 1,200 bpm. Inclusion of the
`air-compliance element increased the pulse pressure in the inlet
`position and also decreased that in the outlet position.
`
`sure difference (120 mm Hg) at valve closure. In
`addition, in the Helmholtz tester, the influence of
`cyclic rate on lifetime was investigated while, in the
`Rowan Ash tester, the influence of drive amplitude
`on lifetime was investigated. The results in the
`Helmholtz tester are shown in Fig. 10. The repetition
`numbers to fracture were 7.3 × 106 cycles, 7.9 × 106
`cycles, and 13.8 × 106 cycles under cycle rates of 600
`rpm, 500 rpm, and 400 rpm, respectively. These re-
`sults indicate that increased cycle rate shortens valve
`lifetime. Figure 11 shows the results in the Rowan
`Ash tester. The repetition numbers to fracture were
`8.3 × 106 cycles and 3.1 × 106 cycles under drive
`amplitudes of 0.4 mm (without air compliance) and
`0.9 mm (with air compliance), respectively. In-
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`FIG. 8. A Tsinghua University type accelerated fatigue tester for
`prosthetic heart valves is shown.
`
`PAGE 5 OF 10
`
`

`

`ACCELERATED FATIGUE TESTS FOR HEART VALVES
`
`425
`
`FIG. 9. Simultaneous pressure waveforms are shown in the
`Tsinghua-type accelerated fatigue tester under the drive rate of
`1,200 bpm. Sampling time was 100 µs.
`
`the Tsinghua University tester, the repetition num-
`ber to fracture was 8.7 × 106 cycles.
`
`Membrane fractures in accelerated fatigue tests
`A typical fracture pattern in the accelerated fa-
`tigue tests is shown in Fig. 12. Tears always occurred
`alongside the edge of the spoke in the three types of
`accelerated fatigue testers. Figure 13 shows the frac-
`ture surface at the locations of A, B, and C in Fig. 12
`as observed by a scanning electron microscope.The
`fracture surface indicated that the crack initiated in
`location A in the outer area of the membrane.
`The fracture locations in the accelerated fatigue
`tests did not coincide with those in the animal ex-
`periment.
`
`FIG. 11.
`Influence of drive amplitude on durability under the
`specified maximum pressure difference of 120 mm Hg in the
`Rowan Ash accelerated fatigue tester is shown.
`
`DISCUSSION
`
`Differences in durability under pressure condition
`recommended by ISO
`The durability of polymer valves of consistent
`quality was compared under the same maximum
`pressure difference at valve closure as recommended
`by ISO 5840. The durability tests conducted using
`the Helmholtz-type tester indicated that an increase
`in drive rate shortens the lifetime of the valve. In the
`
`FIG. 10. Influence of cycle rates on durability under the specified
`maximum pressure difference of 120 mm Hg in the Helmholtz-
`type accelerated fatigue tester is shown.
`
`FIG. 12. Shown is a typical fracture pattern in the accelerated
`fatigue testers. Tears always occurred alongside the edge of the
`spoke.
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`PAGE 6 OF 10
`
`

`

`426
`
`K. IWASAKI ET AL.
`
`FIG. 13. Scanning electron microscope views of the fracture sur-
`faces are shown. The crack initiated at location A and propagated
`toward locations B and C.
`
`Rowan Ash tester, it was shown that the increasing
`of the drive amplitude of the axially oscillating rod
`also shortens the lifetime. Figure 14 shows changes
`in maximum velocity during the closing phase of the
`Jellyfish valve membrane as a function of drive rate
`in the Helmholtz tester. The motion of the mem-
`brane was captured by a laser displacement sensor
`(LB-01, LB-60, Keyence Corporation, Osaka, Ja-
`pan). The maximal velocity during the closing phase
`was increased by increasing the drive rate. Figure 15
`shows a comparison of membrane surfaces along the
`spoke edge between two different test conditions in
`the Rowan Ash tester. Distinctly deeper abrasion
`was observed in the membrane surface that was
`tested under the drive amplitude of 0.9 mm (with
`compliance element) as compared with the surface
`that was tested under the drive amplitude of 0.4 mm
`(without compliance element). This difference could
`be explained by the fact that the maximal velocity of
`the rod under the drive amplitude of 0.9 mm was 9.3
`m/s which is 2.2 times faster than that under the
`
`FIG. 14. The graph shows changes in maximum velocity of the
`Jellyfish valve membrane during the closing phase as a function
`of drive rate in the Helmholtz-type accelerated fatigue tester. The
`maximum pressure difference at valve closure was maintained at
`120 mm Hg.
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`FIG. 15. Comparison of membrane surfaces of the Jellyfish
`valve under the two different drive amplitudes in the Rowan Ash
`accelerated fatigue tester are shown. The right side represents a
`drive amplitude of 0.4 mm while the left side represents 0.9 mm.
`The maximum pressure difference at valve closure was main-
`tained at 120 mm Hg.
`
`drive amplitude of 0.4 mm. Apparently, the actual
`dynamic load conditions were different even though
`the pressure difference at valve closure was adjusted
`to the same value. This implies that different dura-
`bility results can be obtained by variation of drive
`conditions within the guidelines of ISO 5840. In
`other words, specification of only the maximum
`pressure difference drop at valve closure is not suf-
`ficient to ensure that tests are conducted under
`equivalent conditions. The results indicate that mea-
`surement of the dynamic load acting on heart valves
`would be of importance for the establishment of a
`reliable, standard test methodology. To this end, the
`authors have developed a technique for measuring
`dynamic loads in the Tsinghua tester (11), and the
`influence of dynamic load on durability is now under
`investigation.
`
`Comparison of fracture locations between the
`accelerated fatigue tests and animal experiment
`If in vitro accelerated fatigue testing could reliably
`predict in vivo failure patterns, it would indeed be a
`very powerful tool. In this study, the fracture loca-
`tions of Jellyfish valves obtained in three different
`accelerated fatigue-tests machines were compared
`with those observed in an animal experiment. All the
`membrane fractures occurred along the spoke edge
`in the accelerated fatigue tests while the membrane
`was fractured in the region between adjacent spokes
`in the animal experiment. The results showed that
`the fracture locations in the accelerated fatigue tests
`were not consistent with those in the animal model.
`In this study, the question of reproducibility of the
`single animal experimental result is of much impor-
`tance. However, accumulation of long-term animal
`experimental data (over 300 days, for example) is
`
`PAGE 7 OF 10
`
`

`

`ACCELERATED FATIGUE TESTS FOR HEART VALVES
`
`427
`
`extremely difficult. Therefore, in order to predict
`possible fracture sites in the Jellyfish valve from me-
`chanical considerations, the distributions of strain,
`deflection, and stress in the membrane were ana-
`lyzed using the finite element method. The analysis
`was performed on the closed phase of the Jellyfish
`valve. Because the membrane of the Jellyfish valve is
`not restrained during the opening and opened
`phases except at the central attachment to the valve
`seat, as shown in Fig. 2, the mechanical restraint of
`the membrane by the supporting valve seat was as-
`sumed to be the major factor for inducing the mem-
`brane fracture. Taking advantage of the axial sym-
`metry of the valve, the analysis was performed on
`only one of the twelve radial segments of the Jelly-
`fish valve. Figure 16 shows the analytical model with
`the applied boundary conditions. The model radial
`segment of the valve seat is contained within the
`area surrounded by bold lines. As indices of predic-
`tion of possible fracture sites in the Jellyfish valve,
`the distribution of equivalent elastic strain (␧e), de-
`flection, and equivalent Von Mises stress (␴e) were
`analyzed by the following equations:
`
`␴e =
`
`关共␴x − ␴y兲2 + 共␴y − ␴z兲2 + 共␴z − ␴x兲2
`
`公2
`2
`(2)
`+ 6共␶xy2 + ␶yz2 + ␶zx
`
`
`2 兲兴
`where ␧ is tensile strain, ␥ is shear strain, ␴ is tensile
`stress, and ␶ is shear stress. The results are shown in
`Fig. 17. Strain concentrations were observed in two
`distinct areas of the membrane: midway between ad-
`jacent spokes and adjacent to the spoke edges. The
`maximum strain occurred in the former region, cor-
`responding to the location of maximum deflection.
`The latter region was consistent with that of maxi-
`mum stress concentration.
`The fracture location in the animal experiment
`was clearly coincident with that of maximum strain
`concentration caused by the deep deflection of the
`membrane at valve closure. Therefore, it was con-
`firmed that the fracture location observed in the ani-
`mal experiment was predictable from a mechanical
`point of view. Furthermore, the fracture locations in
`the accelerated fatigue tests were consistent with the
`regions associated with stress concentration. Al-
`though the positions of fracture were different be-
`tween the accelerated fatigue tests and the animal
`experiment, they were, in both situations, entirely
`consistent with the regions of high strain concentra-
`tion. The difference in fracture locations could have
`been the result of changes in the viscoelastic prop-
`erties of the membrane in the physiological environ-
`
`1 2
`
`␧e =冋4
`
`9
`+ 1
`3
`
`
`
`
`
`
`
`共␧x2 + ␧y2 + ␧z2 − ␧x␧y − ␧y␧x − ␧z␧x兲
`
`2 兲册1
`
`
`
`共␥xy2 + ␥yz2 + ␥zx
`
`
`
`2
`
`(1)
`
`FIG. 16. The analysis model and boundary conditions for finite element analysis of the Jellyfish valve membrane are shown. Only one
`of twelve radial segments was treated, taking advantage of the axially symmetric shape.
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`PAGE 8 OF 10
`
`

`

`428
`
`K. IWASAKI ET AL.
`
`ment. In vitro reproduction of the in vivo fracture
`locations is indispensable for the establishment of a
`reliable protocol for accelerated fatigue tests. There-
`fore, a method for compensating for the hypoth-
`esized difference in viscoelastic performance of the
`polymeric material in the accelerated fatigue tests
`could be one of the essential steps in the reproduc-
`tion of in vivo fracture modes. Whereas this study
`dealt with the durability of a polymer valve in the
`accelerated fatigue environment, the similarity in
`viscoelastic behavior of clinical bioprosthetic heart
`valves begs that further research should be con-
`ducted to extend the current protocol to biopros-
`thetic valves.
`
`CONCLUSIONS
`
`The durability of a polymer valve was significantly
`altered by employing different drive conditions in
`accelerated fatigue tests even though all of these
`tests complied with the recommendations of ISO
`5840 (Cardiovascular Implants). Moreover, fracture
`locations between the accelerated fatigue tests and
`the animal experiment were not coincident although,
`in all cases, the fracture locations were consistent
`with the areas of high strain concentration as com-
`puted by finite element method. Further research
`should be conducted into the influence of the dy-
`namic load acting on valves and the differences in
`viscoelastic behavior between the physiological and
`accelerated-rate environments, especially for poly-
`mer and bioprosthetic heart valves. Discussion of
`these basic data would be useful in a revision of the
`durability test protocol prescribed by ISO 5840.
`
`Acknowledgments: This research was conducted with
`the aid of the following research funds: The Program for
`Promotion of Fundamental Studies in Health Science of
`the Organization for Drug ADR Relief, R&D Promotion
`and Product Review of Japan (No. 96-12); Grant-in-aid for
`Scientific Research of Japan (No. 09470288); Health Sci-
`ences Research Grants, Research on Pharmaceutical and
`Medical Safety (No. H-11-006); and Waseda University
`Grant for Special Research Projects, Individual Research
`(No. 2001A-865).
`
`REFERENCES
`
`1. International Standards Organization. ISO5840: Cardiovascu-
`lar implants—Cardiac valve prosthesis (Committee draft).
`Geneva, Switzerland: International Standards Organization,
`1994;18–24.
`2. Clark RE, Swanson WM, Hagen RW, Beauchamp RA. Du-
`rability of prosthetic heart valves. Ann Thorac Surg 1979;26:
`323–35.
`3. Gabbay S, Bortolotti U, Wasserman F, Factor S, Strom J,
`Frater R. Fatigue-induced failure of the Ionescu-Shiley peri-
`
`FIG. 17. Distribution of the equivalent elastic strain, equivalent
`Von Mises stress, and deflection of the Jellyfish valve membrane
`under a pressure load of 90 mm Hg are shown.
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`PAGE 9 OF 10
`
`

`

`ACCELERATED FATIGUE TESTS FOR HEART VALVES
`
`429
`
`cardial xenograft in the mitral position. J Thorac Cardiovasc
`Surg 1984;87:836–44.
`4. Nugent AH, Scotten LN, Walker DK, Brownlee RT. Accel-
`erated fatigue testing of heart valves. In: Schwartz MD, ed.
`Proceedings of the 37th Annual Conference on Engineering
`Medicine and Biology. Bethesda, MD: Alliance for Engineer-
`ing in Medicine and Biology, 1984;149.
`5. Imachi K, Mabuchi K, Chinzei T, Abe Y, Imanishi K, Yon-
`ezawa T, Maeda K, Suzukawa M, Kouno A, Ono T, Fujimasa
`I, Atsumi K. In vitro and in vivo evaluation of a jellyfish valve
`for practical use. Trans Am Soc Artif Intern Organs 1989;35:
`298–301.
`6. Imachi K, Mabuchi K, Chinzei T, Abe Y, Imanishi K, Su-
`zukawa M, Yonezawa T, Kouno A, Ono T, Nozawa H, At-
`sumi K, Fujimasa I. Blood compatibility of the jellyfish valve
`without anticoagulant. Trans Am Soc Artif Intern Organs
`1991;37:220–2.
`7. Abe Y, Chinzei T, Mabuchi K, Snyder AJ, Isoyama T, Imani-
`shi K, Yonezawa T, Matsuura H, Kouno A, Ono T, Atsumi K,
`Fujimasa I, Imachi K. Physiological control of a total artificial
`
`heart: conductance- and arterial pressure-based control. J
`Appl Physiol 1998;84:868–76.
`8. Imachi K, Chinzei T, Abe Y, Mabuchi K, Matsuura H, Karita
`T, Iwasaki K, Mochizuki S, Son YP, Saito I, Kouno A, Ono T.
`A new hypothesis on the mechanism of calcification formed
`on a blood-contacted polymer surface. J Artif Organs 2001;4:
`74–82.
`9. Vesely I, Boughner DR, Dietrich JL. Bioprosthetic valve tis-
`sue viscoelasticity: implications on accelerated pulse duplica-
`tor testing. Ann Thorac Surg 1995;60:S79–83.
`10. Reul H, Eichler M, Potthast K, Schmitz C, Rau G. In vitro
`testing of heart valve wear outside of the manufacturers labo-
`ratory—requirements and controversies. J Heart Valve Dis
`1996;5(suppl 1):105–110.
`11. Iwasaki K, Umezu M, Wakui H, Kawada H, Fujimoto T, Ima-
`chi K. Improvement of dynamic conditions in the accelerated
`fatigue testing for prosthetic heart valves. In: Goh JCH,
`Nather A, eds. Proceedings of the 9th International Confer-
`ence on Biomedical Engineering. Singapore: National Univer-
`sity of Singapore, 1997;342–44.
`
`Artif Organs, Vol. 26, No. 5, 2002
`
`PAGE 10 OF 10
`
`