Prosthetic Hand System Can Learn Movement Patterns
July 01, 2019
Researchers at the Biological Systems Engineering Lab at Hiroshima University, Japan, have developed a 3D-printed prosthetic hand combined with a computer interface that can react to a user's motion intent. Previous generations of the prosthetic system were made of metal and were heavy and expensive to make. An open-access article about the device was published June 26 in Science Robotics.
Toshio Tsuji, PhD, a professor in the university's graduate school of engineering, described the mechanism of the hand and computer interface using a game of rock-paper-scissors. The prosthesis user imagined a hand movement, such as making a fist for rock, and the computer attached to the prosthesis combined the previously learned movements of all five fingers to make the motion.
"The patient just thinks about the motion of the hand and then the robot automatically moves. The robot is like a part of his body. You can control the robot as you want. We will combine the human body and machine like one living body," said Tsuji.
Electrodes in the socket measured EMG signals from nerves through the skin. The signals were sent to the computer, which only takes five milliseconds to make its decision about what movement it should be. The computer then sends the electrical signals to the motors in the hand.
The neural network, named Cybernetic Interface, was trained to recognize movements from each of the fingers and combine them into different patterns to turn scissors into rock, pick up a water bottle, or control the force used to shake someone's hand.
"This is one of the distinctive features of this project. The machine can learn simple basic motions and then combine and then produce complicated motions," said Tsuji.
Seven male participants, including one person with an amputation, were recruited for the study. Participants were asked to perform a variety of tasks with the hand, such as picking up small items or clenching his fist. The accuracy of prosthetic hand movements measured in the study for a single simple motion was above 95 percent, and complicated, unlearned motions was 93 percent, according to the researchers.
The next step in the research is to create a training plan to reduce muscle fatigue for the user.