A Soft Robot with a Lot of Heart

An innovative design reduces the risk of stroke and bleeding in patients awaiting a new heart.

by Jeff O'Heir
April 10, 2017

Soft robotics can mimic the look and basic motion of human muscles. Replicating the twisting, extending and flexing of the tongue, diaphragm, or stomach is much more difficult. Duplicating the twisting motion of the heart, given its complexity and vital function, is especially tough.

But Ellen Roche, a biomedical and mechanical engineer who has developed stents and aortic valve replacements for medical device manufacturers, had a few ideas about how to do it.

The result – a soft robotic sleeve that fits around the heart and uses actuators to mimic its twisting and compressing movement to pump blood – could help save lives.

Implanting VADs

A few years ago, Roche was earning her doctorate at Harvard University’s Biodesign Lab, one of the world’s top soft robotic labs, when Dr. Frank Pigula reached out to the group.

Pigula, the former clinical director of pediatric cardiac surgery at Boston Children’s Hospital, was looking for engineers to create a new product that could replace the ventricle assist devices (VADs) used to pump blood through the body of people suffering end-stage heart failure while they await transplants.

A new device, Pigula believed, could have a big impact on the 2,000 people who receive heart transplants each year. VADs are effective in keeping patients alive, but their design poses risk and can cause complications.

When surgeons implant the device in a patient, they connect a tube from the patient’s heart to the VAD, which pumps the blood through an outflow tube and into a major vessel. This risks infection. Also, the blood that comes in contact with the tubes and pumps can trigger clotting and increase the likelihood of stroke by 20 to 30 percent. Anticoagulants used to prevent clotting may cause additional bleeding where the VAD was implanted and require constant monitoring.

“Our ability to control coagulation is like using a sledgehammer on a mosquito—it’s not very precise,” said Pigula, who recently joined the cardiovascular and thoracic surgery department at the University of Louisville and is a coauthor on the paper, recently published in Science Translational Medicine journal. “It is a pretty serious problem.”

To help save more lives, Pigula wanted a device that did not touch the blood. Roche, who was looking for a thesis topic at the time, took up the challenge.

Mimicking the Heart

Through her work on medical devices and soft robotics, Roche knew she could combine the nonrigid silicone that makes up most soft robots with soft actuators to create an inexpensive device that mimics the precise twisting and recoiling movement of the heart’s left ventricle. The left ventricle moves most of the blood throughout the body and is much stronger than the right ventricle.

The researchers, including Biodesign Lab founder Conor Walsh, took several approaches. First, they created a silicone dome that could cover the heart from its base to its apex. The engineers 3D-printed a two-part mold to create two layers of silicone and sandwiched the actuators between the layers.

The team based its actuators on the McKibben pneumatic artificial muscle (PAMs), an established class of soft actuators originally designed for artificial limbs. Easy and inexpensive to make, they consist of an inflatable bladder surrounded by a soft braided mesh. They are strong, responsive, operate under low pressure, and behave like a real muscle when powered pneumatically. But they are limited in that they can only expand and contract. Roche needed something that could also twist.

To solve the problem, the team arranged the PAMs in a helix, a slightly curved pattern that produces a twisting motion by rotating the device’s base and apex in opposite directions, just like a left ventricle. Each PAM is tethered to a compressor by tubes. Individually powered and controlled, the PAMs increase or decrease force on the heart, depending on the need.

“We can change the pressure or timing sequence to make the device do whatever we want,” Roche said. “We have complete control over it.”

Working with Engineers

The device worked well enough on bench-top heart models. But testing it on a beating pig’s heart, which closely resembles a human heart, posed several challenges. That’s where the engineers turned to the doctors for help.

“It was a great example of how engineers and surgeons can work closely together,” said Roche, who will join the faculty at MIT in September.

First, the doctors helped adjust the dome so that the actuators provided pressure on exactly the right spot of the heart. Relying on years of experience, the doctors also intuited the rate of the heartbeat and helped to synch the rhythm of the actuators to it.

A few other problems persisted. The dome often slipped off the heart and proved to be too stiff in some areas. To solve those challenges, the engineers fabricated thin sheets of silicon, sandwiched the actuators between them, and formed a sleeve.

That sleeve design allowed the surgeons to attach the device more snugly around the heart and to give it more flexibility. The team also used gel to help the device adhere to the heart, something the engineers are still trying to improve. The gel also reduces friction between the silicon and the muscle.

Studies in pigs and in lab simulations showed that the sleeve could restore the output of an acutely failing heart to 97 percent of its original capacity. The team will continue to improve the accuracy and portability of the control system, as well as the sleeve’s adhesion. Researchers also envision using the device to deliver a variety of mechanotherapy applications inside and outside of the body.

Although the device won’t hit the market anytime soon, Harvard is seeking to license the technology. Pigula said its future is promising.

“The proof of principle provides important information that this has real value,” he said. “It can open up new avenues that we haven’t explored during the last two decades. But we still have lots of testing to do.”


Jeff O’Heir is a technology writer based in Huntington, New York.