Mechatronic Artificial Heart Doesn't Beat

For all their success at extending the lives of a few heart-failure patients, artificial hearts still suffer from a couple of design flaws that have limited their use. These electromechanical hearts currently rely on positive displacement blood pumps, which tend to be bulky. So bulky, in fact, that the most recent self-contained heart designs simply won't fit in smaller chest cavities. Positive displacement pumps also have more moving, cyclically-loaded parts than engineers like to see from a reliability standpoint. A new kind of mechatronic heart under development at the Texas Heart Institute

uses a much simpler pump and sophisticated control algorithms to address both issues.

Rather than positive displacement pumps which mimic the pulsations of a natural heart, the Texas Heart Institute's total artificial heart (TAH) design employs a pair of continuous-flow axial pumps. A pulmonary-loop pump carries oxygen-depleted blood away from the heart to the lungs and returns oxygenated blood to the heart. And a systemic-loop pump carries oxygenated blood away from the heart to the body and returns deoxygenated blood back to the heart. A controller allows the pumps to work in tandem, adjusting their outputs relative to one another and to changing physiological demands.

According to Steve Parnis, assistant technology director for the Texas Heart Institute's Center for Cardiac Support, these continuous flow pumps are essentially repurposed DeBakey ventricular assist devices (VADs) from MicroMed Cardiovascular. Usually a VAD does what its name suggests – it assists the natural heart with its pumping duties. "In this case, the two VADs would completely replace the natural ventricles," Parnis says.

It's a novel idea that's been around for a few years now. Dr. Bud Frazier, the Texas Heart Institute's director of research and chief of cardiopulmonary transplantation, published papers on a continuous flow TAH back in 2006.  His idea took a step closer to clinical reality this year, when the National Institutes of Health awarded a $2.8 million grant to the Texas Heart Institute to fund the development of the continuous flow heart design.

VADs have a lot going for them in a total heart application. For one, they're each about the size of a C-cell battery, versus a 2-lb chunk of titanium and plastic for a self-contained pulsatile pump. "VADs will fit in the majority of patients versus a minority of patients for the pulsatile pumps currently in use," says Parnis.

For another, VADs have a proven clinical track record. About 500 of the current generation of DeBakey VADs are in use right now, according to Bob Benkowski, MicroMed's chief operating officer and one of the engineers who helped develop the original DeBakey VAD model. "VADs have run in patients for as long as eight years," he says. And he attributes that reliability in part to the simplicity of MicroMed's axial pump, whose single moving part, the impeller, is directly driven by the electrical windings.

Parnis puts the lifecycle expectation for even the most modern positive displacement blood pumps, whose pulsations create cyclical loading conditions not seen by the axial models, at two years. Continuous flow pumps will also likely require less power and cost less than the more complex pulsatile models, he says.

So if a couple of continuous flow VADs would make such a great TAH, why aren't they in use yet? It turns out these devices need a significant amount of controls engineering to make the jump from heart helper to total heart replacement.

And that's where Matthew Franchek and Ralph Metcalfe, both Ph.D mechanical engineers and professors at the University of Houston's Cullen College of Engineering, enter the picture. As part of the NIH grant, they're working on a feedback controller that will allow two VADs to work together as a TAH. Franchek and other university researchers have developed similar auto-regulating control systems for automotive applications, most recently working on a diesel engine governor for Cummins Engine.

In some ways, Franchek and Metcalfe have had a head start in the controls development work thanks to the use of the proven VAD technology. MicroMed's VADs already have their own controls. Benkowski describes them as feedback controllers, which take an actual flow measurement from an ultrasonic sensor, compare it with a desired flow output and generate an appropriate PWM control signal to regulate the impeller speed.

Yet the two engineering professors still have their work cut out for them. VADs normally operate individually as support for a remaining natural heart. In the TAH, they have to operate in close coordination to emulate the balanced flow provided by a natural heart's left and right ventricles. "Pairing the pumps creates a complex multivariable control problem," Franchek says. "Each pump's loading conditions and flow output affects the loading conditions and flow output of the other pump."

The TAH controller also has to tie these interrelated flow and loading conditions – which include both inlet pressure and outflow resistance – back to the changing needs of the human body. Franchek says everyday activities such as standing or walking change flow and loading conditions. So do cardiovascular events such as vascular restrictions, hypertension or changes in blood viscosity. And so do intrinsic physiological differences between individual patients. "Our challenge is to maintain a steady-state cardiac output as physiological conditions fluctuate for whatever reason," says Franchek.

Axial flow pumps inherently lend themselves to meeting this challenge. They can auto-regulate transient events because their flow output is sensitive to both inlet pressure and outflow resistance. And Benkowski says the VAD pump's impeller geometry and flow passages can be tweaked to come up with optimized flow-pressure behavior for the TAH application. "We can alter the pressure sensitivity of the pumps to make it a little easier for the control algorithms to do their thing," he says.

Those algorithms, meanwhile, will be based on an analog integral controller which measures actual output flow, compares it to the desired value and adjusts the voltage to the pumps accordingly. Franchek and Metcalfe picked a seemingly simple integral control strategy for this application because it does a good job at maintaining steady-state conditions in systems whose dynamic behavior is both well understood and characterized by cooperative transients. Understanding that dynamic behavior given the influence of physiology on pump conditions is not so simple. And a large part of the control development work under the NIH grant involves the creation of a lumped parameter mathematical model of the human circulatory system. According to Franchek, this model will ultimately be incorporated into the TAH control algorithms (see block diagram below).

Franchek expects the first pass at the TAH control algorithms won't be ready until this summer. "Right now, we're at the very beginning of the controls engineering," says Franchek. And there are still some fundamental decisions to be made about how the pumps will operate. For example, the researchers have yet to decide whether one or both of the pumps should be operated in a quasi-pulsatile mode. Franchek says a repetitive control strategy would let the pump motors "whirl up and whirl down" to emulate the pulsating action of the natural heart if need be.

Other development work includes the possible addition of blood-viscosity monitoring to the system. "We believe we'll be able to infer the effective viscosity of the blood from our flow measurements and voltage signals," Franchek says.

He and Metcalfe are using a variety of simulation tools to do their development work, including MATLAB and Simulink to develop the mathematical models. They're simulating the resulting control algorithms and prototyping the controller hardware in dSPACE, a set of development tools for mechatronic systems.