Research offers hand to understanding motor control

Paul Mayne // Western News

Through his exploration of how the brain produces movement, particularly for the hand, Jörn Diedrichsen, Western Research Chair for Motor Control and Computational Neuroscience, hopes to unlock better treatments for stroke, spinal cord damage and other motor control disorders.

We can do many things with our hands – carry loads, manipulate fine objects, grasp items, even communicate with each other. Our hands, Jörn Diedrichsen says, are “the Swiss Army knife of our body parts.”

As the Western Research Chair for Motor Control and Computational Neuroscience, and part of the Brain and Mind Institute, Diedrichsen explores how the brain produces movement, particularly for the hand, in hopes of unlocking better treatments for stroke victims, spinal cord damage and other motor control disorders.

“Controlling the hand is a really difficult problem and the brain solves that by having a lot of different neurons talking to each other and sending out commands to the hands, getting sensory information back, and processing them very, very fast,” said Diedrichsen, who arrived from University College London (U.K.) in September. “We don’t understand the language by which these neurons talk to each other. For that, we really need a computational description of how big populations of neurons talk to each other and produce these amazing things we can do.”

Diedrichsen’s lab uses robotics to study human movement. By simulating objects or environments, he studies how the brain recalibrates well-learned motor skills, or acquires new ones, by developing computational models to understand the underlying control and learning processes.

“We are mostly interested in how and when the brain learns new movements. How do these neuro representations change and how can we understand them in the live human being?” said Diedrichsen, who uses fMRI technology to look at brain activity during such tasks. “It’s important to be able to bridge the gap between animal research and human research, and really transport these results from basic neuroscience into the human domains and then apply it in the clinical domain for stroke recovery and other disorders.”

Human motor control has become an exciting research field, added Diedrichsen, where neuroscientists, like himself, and roboticists are continually feeding off one another to further their individual research.

“As neuroscientists, we take the control mechanisms – the algorithms roboticists are developing – and can test if this gives us insight about what the human brain is doing,” he said. “The roboticists, and computer scientists, are taking the results of what we can produce about neuro networks – and what we can understand about the human brain – to write new algorithms to better control robotic devices.

“These two fields are coming closer and closer as we work forward.”


New developments in artificial intelligence are able to mimic the brain, to some degree, and reach levels of near-human performance in vision. However, Diedrichsen said, motor control is a far different scenario.

“If you watch 1970s sci-fi movies, you would think that in 2016 there should be robotic devices running around everywhere. You don’t see them,” he said. “There is a reason for this – motor control is a very hard problem to solve, and we haven’t quite understood how the brain actually solves it. We are not anywhere close to near human performance. Motor control is a really hard problem to crack for computational scientists.”

Everyone who is trying to build a robotic device that moves and acts like a human, specifically mimicking a human hand, walks away with an increasing admiration of how nature actually created such amazing appendage, Diedrichsen added. How we are able to control the spongy muscles, tendons and bones to do such amazing things is, in a way, a “real kind-of miracle.”

“Look at this,” said Diedrichsen, holding up his hand. “It’s a total engineering nightmare. It’s built out of squishy parts; it’s not very exact; the motor and muscles are really noisy; the linkage is all flexible. As an engineer, you would never build it like this. The brain actually controls this better than we can control devices that are a hundreds or thousands of times more accurate.”

A big reason Diedrichsen chose to come to Western, he said, was the strength of the university’s functional neuroimaging reputation worldwide.

“Not only the devices, but there is an amazing team who support this,” he said. “The Brain and Mind Institute has really developed over the last few years to become a great interdisciplinary team of scientists. It is a very collaborative place, and with different interests in the brain, we are all talking to each other. I came here as one of the computational people, and through that we really have developed a lot of collaborations.”

It is through these collaborations and partnerships Diedrichsen hopes to makes strides in an area he feels needs to make more progress.

“Dysfunctional motor control is one of the leading causes of disability in Canada. We really need to have better treatments, for example, for stroke or spinal cord damage,” he said. “Over the last few years, we have learned a lot about how the brain controls the hand. But we haven’t really been able to translate that into the clinical domain. Stroke recovery, and stroke rehabilitation, is really on the level of the 1920s or 1930s.

“We haven’t made any breakthrough movements there and translate that scientific insight into the clinical domain, and that really needs to happen. That needs to change.”