A soft exoskeleton could restore hand function in people with motor impairments

Researchers at the Technical University of Munich and the Center for Rehabilitation Passauer Wolf have developed a lightweight soft hand exoskeleton intended to help people with severe motor impairments regain grasping ability. Reported in Nature Machine Intelligence, the system combines a flexible mechanical glove design, sensing of electrical activity in hand muscles, and machine learning.

The exoskeleton is made from fabric using a standard sewing machine and is powered by air. Its sensors detect weak muscle activity from the wearer’s hand. A machine-learning model interprets those signals to predict when the person intends to grasp an object, after which the glove assists the required hand movement. The researchers emphasized that predicting the wearer’s own intention is important, rather than making the glove move automatically.

The team evaluated the device with one participant with amyotrophic lateral sclerosis who had been unable to move his right hand for years, along with six people who had lost movement in one hand following strokes. Participants completed the Box and Block Test and the Action Research Arm Test, measures commonly used to assess hand movement in clinical settings.

According to the reported results, people with the most severe impairments received the greatest benefit. Participants who retained partial hand control did not consistently perform better while wearing the device. In one example, the participant with ALS used the system to pick up a fork and feed himself cake after about four years of paralysis. He also helped modify the device and its controls, preferring to actively instruct the glove rather than have it operate independently.

The researchers plan to refine and assess the hand system in additional clinical settings. They are also developing a lower-body exoskeleton.