Students create brain-controlled prosthetic hand
For people with amputated forearms, the prospect of having a responsive prosthetic hand and wrist is right around the corner.
Mohit Chaudhary, Chris Bargoud, Julian Hsu, James Wong and Rebecca Wenokor, a group of five Rutgers biomedical engineering seniors in the School of Engineering, took on their senior project, “Brain-Controlled Dexterity Upper Extremity Prosthesis.”
The group worked to develop a prosthetic hand and wrist to operate according to brain signals for a current Rutgers student.
The team began the project last September with mentors Dr. William Craelius, a professor at the Department of Biomedical Engineering, and Dr. Kang Li, assistant professor at the Department of Industrial and Systems Engineering.
The uniqueness of the project lies in the wrist, Bargoud said. The goal is to give the wrist controllable freedom that allows different motions, instead of being as static as the wrists on the market are.
“[Wrists in the market] work like a pin [on the joint between hand and forearm]. You unlock it, rotate your wrist and then lock it,” Bargoud said. “You can’t actually control what you want to do.”
The team has seen tangible results, even with a limited amount of time and a budget. Using a 3-D printer, the seniors improved designs for the hand, printed out parts and had their prosthetic hand assembled.
The hand in the picture has strings attached inside. By attaching motors in the palm area and rotating and pulling the cables, fingers are able to move and bend according to user’s needs.
The team intends to use servos to perform the motions. A servo is a kind of motor that can rotate to an intended angle.
“We are focusing on the structure and how to motorize the structure right now,” Chaudhary said. “We’ll add on the thumb as well. After our hand is well implemented, we will add on the wrist.”
Through many trials of different designs and prototypes, the team strives toward enabling “three degree-of-freedom.”
The three degrees-of-freedom are flexion and extension, medial and lateral deviation and pronation and supination, Bargoud said.
Flexion and extension means reducing or increasing the angle between hand and forearm, Bargoud said.
Bargoud explained medial and lateral deviation means bending the hand toward and away from the medial line of the body with palm facing down, while pronation and supination means rotating the hand inward and outward while keeping the hand straight from the forearm.
BCI, which stands for brain-computer interface, is a device that reads brain signals and connects with a computer that processes the signals. The team is using a BCI device from the company OpenBCI, which is an open source project, he said.
“Even though the name of our project says ‘brain-controlled’, it is more accurately brain initiated,” Hsu said.
The team plans to program the functions into an Android phone, where Bargoud said the user can click and choose the kinds of motion he wants to perform.
Meanwhile, BCI electrodes attach to the brain to pick up a “start this action” command and then pass that signal along.
Both the BCI and the Android phone are connected to the Arduino board, an external device that receives both signals of “when to move” from BCI and “what to do” from the Android phone, Bargoud said.
From there, the Arduino board processes the signals and signals the servos to rotate at a specific moment to a certain angle. The servos drive the gears in the prosthetics to complete the action of the wrist or hand, he said.
“When [the servo] rotates one way, it will pull the string, and when it goes back, we will have rubber bands attached to the back so it will snap it back with the tension,” Bargoud said.
Besides achieving a three degree-of-freedom, Chaudhary also hopes to practice with the user once the wrist is in place by the end of this semester.
“For any user of brain computer interface, we have to train them in order to be able to attain a certain threshold of brainwave. That takes training and practice,” Chaudhary said.
The team plans to continue improving the design. Chaudhary, Bargoud and Hsu all agree the ideal prosthetic hand would be stable, strong, easy to manufacture, user-friendly and as lightweight and realistic as possible.
“Stability is one of our biggest concerns right now, as Dr. Lee mentions. When it is not used at all, we want to make sure it is stable and still, not just flapping around,” Chaudhary said, “We also want to make it anthropomorphic, as close to real hand as possible.”