s

m
m
Ti
a n iel
D
a nk Nest ler &

Peter Adams
THE MA

e r, F r
hy by
hey
GL

rap
ein
tog
EV

l
sK

ho

G
ia

OR
H

th

E.
Th

EE
at

R
.I
M
e

UM

T x
de

re
TR

de t
ca

EC

s- ea
SP

ol s Gr
| 

d ola 19
rn
 

qu h 20
Nic tu


es y g EP
tf B S
or sin | 
an pri
 
r 22
artif u
icia l hea r t t a ke s a s

Authorized licensed use limited to: Universidad de Talca. Downloaded on April 29,2021 at 00:49:41 UTC from IEEE Xplore. Restrictions apply.
 THE INVENTOR: Daniel Timms,
founder of Bivacor, holds the
650-gram machine that could
save the lives of men, women,
and children with heart failure.

Authorized licensed use limited to: Universidad de Talca. Downloaded on April 29,2021 at 00:49:41 UTC from IEEE Xplore. Restrictions apply.
 l

For more than 50 years, cardiac surgeons and biomed- Our company is now working toward human trials of our
ical engineers at the Texas Heart Institute (THI) have been device. It relies on a dramatic new approach: Rather than
questing for an artificial heart that can fully replace natu- using a mechanical pump that mimics the structure and
ral ones, which are in terribly short supply for transplant. actions of the four-chambered human heart, it uses a spin-
They’ve seen their share of metal and plastic contraptions ning disk, suspended in a magnetic field. With just one moving
that used a variety of pumping mechanisms, but none of part, the Bivacor heart is able to send oxygen-rich blood out
these machines could match the astounding performance to the body and return oxygen-depleted blood to the lungs.
of the human heart. We had to overcome many technical challenges to make an
In April 2019, the possible culmination of that long quest artificial heart that’s small, biocompatible, energy efficient,
was inside a shaggy brown cow, which stood peacefully chew- and durable. Consider that the human heart beats about
ing its cud at a THI research facility in Houston. The animal 112,000 times a day, which adds up to 42 million times a
was part of a 90-day trial in which it lived its life powered by year, and you’ll understand the magnitude of the challenge.
an implanted artificial heart made by our company, Bivacor. We’ve tested the Bivacor heart in 15 cows so far. While
Throughout the trial, the calf stayed healthy and energetic, the need for animal testing is unfortunate, it’s the only way
and gained weight at a normal rate. It even jogged on a tread- to prove the device’s safety and move forward to clinical
mill for 30-minute stretches. trials in humans. These Corriente calves, which are rela-

24  | SEP 2019 | SPECTRUM.IEEE.ORG PHOTOGRAPH BY Peter Adams

Authorized licensed use limited to: Universidad de Talca. Downloaded on April 29,2021 at 00:49:41 UTC from IEEE Xplore. Restrictions apply.
 One of us, Bivacor founder Timms, began working on this
project 18 years ago. [For more about his personal story, see
“Taking Lessons to Heart,” p. 2.] Through the years, the cen-
tral concept of the maglev heart has stayed the same, though
our engineering team has greatly refined the technology.
We’re thrilled that the device is nearly ready for human tri-
als because we think the Bivacor heart is the solution that
severely ill heart-failure patients have long been waiting for.

In heart failure, the heart becomes unable to pump

enough blood to keep the body healthy and strong. At least
26 million people around the world are living with the dis-
ease, and the number is rising as populations age. Patients
with severe heart failure have a bleak outlook: Their best
option is a heart transplant, but the limited number of donor
hearts means that only about 5,000 patients around the
tively small, are the right size to serve as analogues for adult world receive transplants each year. Thousands more patients
patients. We’ve also implanted the Bivacor heart in several are eligible for transplants, and some die while waiting for
sheep, which are more representative of patients with smaller a donor organ.
­bodies, including children. Our tests have shown that the Cardiologists have long dreamed of a mechanical replace-
heart holds up well: With its one moving part levitating in a ment. In 1969, THI physician Denton Cooley implanted the
magnetic field, there’s no worry that friction and mechani- first “total artificial heart” in a patient who was awaiting a
cal wear will cause the machine to give out. Our tests have transplant, keeping him alive for the 64 hours it took for a
also shown that the device can adapt to the user’s cardio- donor organ to arrive. However, that patient died shortly
vascular requirements. after the transplant surgery from an infection, and Cooley’s
LEFT: ISTOCKPHOTO; RIGHT: BIVACOR

HEART TECH: In Bivacor’s engineering lab, founder and CEO/CTO Daniel Timms tinkers with the parts of the latest prototype of the artificial heart
[above left]. Its centrifugal pump [above right] uses a single rotating disk, which is suspended in its casing via magnetic levitation. The left side of the
pump stands in for the anatomical heart’s left ventricle, doing the job of sending oxygenated blood [red] out to the farthest reaches of the body; it has
wide impeller vanes that produce enough pressure for the task. The right side of the pump substitutes for the heart’s right ventricle; its narrower impeller
vanes return deoxygenated blood [blue] to the lungs. The Bivacor device doesn’t need components to stand in for the heart’s left and right atrium, which
take in blood and send it to the ventricles.

SPECTRUM.IEEE.ORG | SEP 2019 |  25
Authorized licensed use limited to: Universidad de Talca. Downloaded on April 29,2021 at 00:49:41 UTC from IEEE Xplore. Restrictions apply.
 l
The Maglev Heart

team shelved the device due to concerns about reliability

and compatibility with the human body.
Since then, a handful of other total artificial hearts have
been developed, and a few have made it into human trials.
But these devices were large, heavy, and prone to mechani-
cal failure, and only two gained regulatory approval in the
United States. The SynCardia Systems artificial heart was
approved in 2004 as a “bridge to transplant” and is currently
being tested as a permanent replacement. However, users
must carry a 6-kilogram case containing a loud air com-
pressor that’s attached to tubes penetrating the abdomen,
where the air drives the device’s two pneumatic pumps.
A second artificial heart, the AbioCor, gained approval in
2006, but it was discontinued almost immediately, when
the company behind the device decided it wasn’t com-
mercially viable.
The most common mechanical support now in use is what’s
called a left ventricular assist device (LVAD). This machine
adds its pumping power to an ailing heart, focusing on the left
ventricle, which pumps oxygen-rich blood through branch-
ing arteries that reach up to the brain and down to the toes.
LVADs are now being used as both temporary aids for patients
awaiting transplants and, in cases where transplantation
isn’t an option, as permanent additions to patients’ chest
cavities. However, LVADs can cause the right ventricle to fal-
ter, requiring intensive drug treatments and sometimes the
implantation of a right ventricular assist device.
At Bivacor, we wanted a true long-term mechanical replace-
ment for the entire heart. Unlike most earlier attempts, our
effort didn’t set out to mimic the heart’s natural pulsatile
pumping mechanism, with valves that open and close as
the left ventricle pushes blood out to the body and the right
ventricle pushes blood to the lungs. Most prior artificial
hearts used positive displacement pumps to achieve this
effect: Their two artificial chambers were bifurcated with
membranes that flexed forward to push blood out through
mechanical valves.
We opted instead for a centrifugal pump that propels a
continuous stream of blood into the arteries. Such a pump
has no valves, so a patient using the Bivacor heart in its most The Bivacor heart would fit in the palm of your hand—
basic mode would have no pulse. But we recently adapted our it’s about 650 grams, slightly heavier than an adult human
device to give it a pulsatile outflow option, as we’ll explain heart. Its shell is made of titanium, a noncorroding material
below. We want clinicians to look at the cardiac monitors for that almost never triggers an immune response. Patients
patients with implanted Bivacor hearts and see the familiar will wear a 4-kg external controller pack that contains
rhythmic readouts they’ve seen since medical school. two rechargeable batteries (providing about 5 hours of

IN THE LOOP: To rapidly test their artificial heart prototypes, and to reduce the need for animal testing, the Bivacor team built an apparatus that
replicates the human circulatory system [above], simulating the flow of blood through the body and the lungs. By connecting a prototype to the testing
rig, Bivacor engineers can watch how it propels a synthetic bloodlike liquid through the tubes and valves. The engineers test a prototype’s sturdiness
[opposite, left and middle] by keeping it hooked up to a special durability rig for several months. The device’s casing [opposite, right] is made of titanium,
which is biocompatible. The parts that look golden have a coating of titanium nitride.

26  | SEP 2019 | SPECTRUM.IEEE.ORG
Authorized licensed use limited to: Universidad de Talca. Downloaded on April 29,2021 at 00:49:41 UTC from IEEE Xplore. Restrictions apply.
PHOTOGRAPHS BY Peter Adams SPECTRUM.IEEE.ORG | SEP 2019 |  27
Authorized licensed use limited to: Universidad de Talca. Downloaded on April 29,2021 at 00:49:41 UTC from IEEE Xplore. Restrictions apply.
EVOLUTION OF AN IDEA: Timms came up with the original concept of an artificial heart with a centrifugal pump in 2001 and has been developing the
technology over the past 18 years. The first concept model [left] tested the hydraulic feasibility of the design, with a single rotor pumping fluid in two
directions. Later prototypes [middle left and right] were used for the crucial tests in cows. The final design [right] is more tailored for human anatomy.

operation each), although they can also plug in directly goes about daily activities. It’s important to keep the rotor
to a power outlet. properly suspended and to prevent it from bumping into the
Throughout our design process, we used 3D printers to sides of the casing, which could damage the components
make both titanium and plastic parts for our prototypes, and smash blood cells.
allowing us to rapidly experiment with different geometries. This positional control system works as follows: Tiny con-
For testing, we built a hardware simulation of the human tactless sensors send out magnetic fields that interact with the
circulatory system in our engineering office in Los Angeles; rotor, determining its exact location many times per second.
this mock-up allows us to validate a device’s function thor- If the rotor is moving in one direction or another, the control
oughly and repeatedly in a controlled environment, and system puts electrical energy into electromagnetic coils within
reduces the need for animal testing. several actuators, causing them to cancel out that movement.
The primary design innovation in the Bivacor heart is its The design of the magnetic bearing actuators was critical,
simple construction, with one motor and one rotating disk particularly because they had to be small, light, and energy
that simultaneously supports the pumping of blood to both efficient, yet strong enough to compensate for all the ­jiggles
the body and the lungs. The rotating disk is completely sus- and joggles created by a person on the move. We used com-
pended in a magnetic field, operating under the same ­“maglev” puter simulations of the magnetic field to optimize the design,
principle that has been used by high-speed trains. The disk experimenting with different materials and geometries to find
has open impeller vanes on both sides, one larger set that the configuration that would provide sufficient force within
pumps blood at the high pressure necessary to send blood a small space. Keeping our device small and light means that
throughout the body, and another smaller set that pumps it will fit inside people with small bodies, including women
blood at lower pressure to the lungs. Each side of the disk and children.
can pump more than 12 liters per minute, more than enough To improve the efficiency of our artificial heart, we also
output for patients engaged in moderate exercise. integrated a “zero power” controller into the suspension sys-
While some blood from the two sides mixes around the tem. This controller monitors the additional electro­magnetic
edges of the suspended disk, this isn’t a concern, because power used by the bearing as it responds to external forces
of the direction of the flow. Some oxygenated blood leaks acting on the rotor, then moves the rotor to a position where
from the high-pressure side to the low-pressure side, which the permanent magnets of the magnetic bearing system
means that some already-oxygenated blood returns to the can provide a balancing force. This system doesn’t produce
lungs. And this leakage is actually a design feature, not a bug. instant adjustments—that’s the job done by the main stabil-
The wash of blood around the disk cleans out the casing and ity controller—but it does reduce the amount of power used
ensures that there are no areas where stagnant blood can by the magnetic bearing when exposed to external forces.
form into dangerous clots. A unique feature in the Bivacor heart is that the rotor can
The motor’s stator provides rotational torque by coupling to shift along its axis of rotation to change the amount of blood
a set of permanent magnets in the rotor disk. During normal moved by the left and right sides of the pump. When the rotor
operation, it spins the rotor at speeds of between 1,600 and moves toward the left side of the casing, it brings the impeller
2,700 rotations per minute. The attractive forces between the vanes closer to the casing wall. In this narrow space, most of
motor and the rotor are counteracted by the magnetic bear- the blood is whirled around by the vanes and little flows over
ing on the opposite side of the rotor, which actively ­controls the vanes’ tips, thus increasing the left pump’s efficiency and
the rotor’s position within the casing. This active control sys- consequently its output of blood to the body. This feature is
tem is necessary because the rotor naturally moves around useful for quick adjustments necessary in transitions, such
as the patient walks, climbs stairs, jumps, and generally as when a patient stands up.

28  | SEP 2019 | SPECTRUM.IEEE.ORG PHOTOGRAPHS BY Peter Adams

Authorized licensed use limited to: Universidad de Talca. Downloaded on April 29,2021 at 00:49:41 UTC from IEEE Xplore. Restrictions apply.
 Our device is elegant in its cooperation with the human hope that our device, which can be used in either continu-
body, adapting automatically to a person’s activity levels. ous flow or pulsatile mode, can contribute to the scientific
When the patient is exercising, the mechanical heart will investigation of this important topic.
pump more blood out to the muscles, just as a biological
heart does. This adjustment is accomplished by attending l
to the body’s feedback: When a person starts to run on a
treadmill, the hardworking muscles in the legs and elsewhere The Bivacor heart has undergone numerous iterations
use up oxygenated blood faster and squeeze the deoxygen- since coauthor Timms began working on the original concept
ated blood back into the circulatory system. That increased 18 years ago. Timms began the project in his native A ­ ustralia,
blood flow into the Bivacor heart causes both sides of the and since those days, the development team has benefitted
pump to move more blood, without requiring an increase from a network of passionate collaborators in Germany, Japan,
in rotor speed. This seemingly basic functionality is actually Taiwan, and the United States. During various stages of the
the result of careful optimization of the hydraulic design. project, the team relocated to international laboratories to
We relied on our 3D printer to make make use of our collaborators’ exper-
numerous experimental prototypes tise. Currently, our team is based in
with casings and impeller vanes of the United States (in Los Angeles and
slightly different shapes. Houston) and Australia (in Melbourne
Biocompatibility is one of the big- and Brisbane), and we’re focused on
gest challenges in our field, because turning our prototype into a commer-
the interactions between a mechani- cial product.
cal device and biological systems are We’re quite satisfied with the design
so complex. For example, the deli- of our device. Now we’re standard-
cate blood cells and other blood com- izing the production process: Our
ponents can be damaged by rough 3D-printing techniques were well
transit through a device. As a key suited for the development phase,
design feature of the Bivacor heart, when we wanted to test prototypes
we ensured that the blood has plenty and iterate rapidly, but now we’re
of clearance between the levitating switching over to precision machin-
rotor and the casing and conduits. ing, which will give the device’s parts
All flow paths have clearance gaps of a smoother polish and more exact and
at least 240 micrometers during nor- consistent dimensions. These proce-
mal operation, which is more than dural changes will make larger-scale
20 times the size of a red blood cell. manufacturing possible. We’re also
This design reduces the shear forces documenting every step of both the
to which the blood is exposed, and SCRATCH-PAD BEGINNINGS: Timms first production and implantation process
also ensures that there’s no blood sketched out the basic idea that became to prepare for our first clinical trials.
stagnation within the casing. the Bivacor artificial heart in a notebook By the end of 2019, we’ll have a
while he was getting his Ph.D. in biomedical
The pump can run at constant engineering—shortly after his father was c
­ linical-grade system. Then we’ll sub-
speed, producing continuous blood diagnosed with heart failure. mit our request to the U.S. Food and
flow at a constant pressure, and in our Drug Administration for an early feasi-
early experiments we concentrated bility study in human patients, which
on testing this “pulseless” mode. But it’s easy to change its we hope will commence in 2020.
speed, and our later experiments proved that controlled We are already envisioning these first human trials. Gravely
speed changes could produce a wide range of flow and pres- ill patients will go into the operating room with failing bio-
sure characteristics. Running the pump first at high speed logical hearts beating feebly in their chests, and come out of
(sending out more blood) and then at low speed (sending out surgery with smoothly functioning Bivacor artificial hearts
less) creates something resembling a biological heart’s pulse; whirring away. As the devices send powerful streams of
rapidly alternating these two speeds creates something that ­oxygen-rich blood coursing through their bodies, we hope
looks like a normal heartbeat. the patients will quickly find their strength returning. If these
We’re now working primarily in that pulsatile mode. Within patients can rise from their hospital beds, hug their family
cardiology, there’s an open debate about whether a pulse members, and continue their lives for many years to come,
is necessary for good health. Some patients with implanted we’ll have taken a great step forward in the long quest for a
LVADs that produce continuous (and thus pulseless) flow total artificial heart. n
experience medical problems such as gastrointestinal bleed-
ing, yet it’s not clear whether their devices are the cause. We ↗ POST YOUR COMMENTS at https://spectrum.ieee.org/bivacorheart0919

PHOTOGRAPHS BY Peter Adams SPECTRUM.IEEE.ORG | SEP 2019 |  29

Authorized licensed use limited to: Universidad de Talca. Downloaded on April 29,2021 at 00:49:41 UTC from IEEE Xplore. Restrictions apply.