New MIT soft prosthetic gives users tactile feedback and control

17 Aug 2021

The prosthetic hand holding a cupcake. Image: Zhao et al/MIT

The inflatable prosthetic limb is designed to provide the same flexibility, durability and control as a traditional one – but at a fraction of the cost and weight.

Engineers at MIT have developed a new soft, stretchy artificial hand that reimagines traditional prosthetics and aims to bring advanced and versatile technology to anyone who has undergone amputation.

The artificial limb is made up of five balloon-like fingers, with fibre replicating bones in human hands. These bendy fingers are then connected to a palm that is made with a 3D printer. The resulting hands weigh just above 200 grams and could cost as little as $500 – a significant saving over existing neuroprosthetics.

Xuanhe Zhao and his colleagues have published the work around their prosthetic in Nature Biomedical Engineering. Co-authors include MIT postdoc Shaoting Lin, as well as Guoying Gu, Xiangyang Zhu and collaborators at Shanghai Jiao Tong University in China.

Most neuroprosthetics rely on electrical motors to control fingers but this can add to the weight of the limbs and makes them cumbersome. Other research has highlighted solutions to create lightweight and realistic systems, but the results are often expensive.

This new limb instead uses a pneumatic system to precisely inflate the fingers and cause them to bend into the desired position, while saving on costs.

“This is not a product yet, but the performance is already similar or superior to existing neuroprosthetics, which we’re excited about,” said Zhao, professor of mechanical engineering and of civil and environmental engineering at MIT.

“There’s huge potential to make this soft prosthetic very low cost, for low-income families who have suffered from amputation.”

From the lab

Lin created a computer model that is able to relate the finger’s intended position with a corresponding pneumatic pressure that would create that bend. By using this model, the researchers made a controller that guides the system to inflate the fingers into different common positions, including pinching fingers together, making a fist and cupping a palm.

The pneumatic system is then hooked up to get signals from an electromyography sensor – a sensor that measures signals generated by motor neurons to control muscles. These sensors are embedded at the joining between the user’s limb and the prosthetic. By doing this, the artificial limb can mimic what would happen if human muscle and bone were present.

The team used an algorithm to decode muscle signals that correspond to different grasps. This algorithm is able to tell that an amputee is imagining balling their fist and can create the pneumatic pressure needed for that gesture.

Researchers also decided to incorporate tactile feedback. By stitching a pressure sensor into each fingertip, the limb sends back a signal that is in proportion to the detected pressure. These sensors are also connected to different parts of the limb, so the user is able to tell which part of the prosthetic is being touched.

Into volunteer’s hands

The team worked with two volunteers, which involved a 15-minute training session to learn how to make the programmed grasps. Afterwards, the participants tried turning pages, writing with a pen and picking up delicate objects such as strawberries.

In one exercise, researchers blindfolded the participants and found the volunteer was able to tell which finger was being touched and brushed. He was also able to judge the size of bottles in the prosthetic hand and lift the bottles in response.

This demonstrated that the limbs were of practical use, and so the team has now filed a patent on the design and plans to work on further improvements.

“We now have four grasp types. There can be more,” said Zhao. “This design can be improved, with better decoding technology, higher-density myoelectric arrays, and a more compact pump that could be worn on the wrist. We also want to customise the design for mass production, so we can translate soft robotic technology to benefit society.”

Sam Cox is a journalist at Silicon Republic covering sci-tech news

editorial@siliconrepublic.com