A game of virtual strength

Photo via Thinkstock.

Photo via Thinkstock.

There are two things that govern how we move: our brains and the mus­cles them­selves.  Every time we pick up a cup or wave to a friend, neu­ro­log­ical stimuli must make their way from the brain to the mus­cles involved to gen­erate a con­trac­tion or relax­ation in those mus­cles and a sub­se­quent move­ment. But local prop­er­ties of the mus­cles also play a role.

How much force a muscle pro­duces in response to neu­ro­log­ical stimuli depends on its length, velocity, and stiff­ness, in addi­tion to its strength. The same signal from the brain could gen­erate very dif­ferent move­ment depending on these prop­er­ties. North­eastern assis­tant pro­fessor of phys­ical therapy, move­ment and reha­bil­i­ta­tion sci­ences Christo­pher Hasson is inter­ested in the inter­play between those two phe­nom­enon and wants to know whether the brain is aware of—and com­pen­sates for—variations in those properties.

The way to test this is to per­turb the muscle prop­er­ties and see how the brain responds – but this is rather dif­fi­cult to do in living humans – at least without doing some rad­ical surgery.

To decouple neural activity from muscle activity he cre­ated a com­puter game that asks sub­jects to hold their arm in a sta­tionary posi­tion and simply flex their tri­ceps or biceps to con­trol the vir­tual arm they see on screen. Sen­sors stuck to the person’s actual muscle read its activity and that is sent to the com­puter to acti­vate the vir­tual mus­cles. It’s an accu­racy game and suc­cess is mea­sured by how close they can posi­tion the vir­tual arm near a fixed target.

But here’s the cool part: the on-​​screen arm con­tains vir­tual mus­cles that may or may not func­tion like the prop­er­ties of the actual arm. Hasson first gives sub­ject time to per­form the task and adapt to the vir­tual arm. Once the sub­ject becomes pro­fi­cient, he can instantly alter the muscle prop­er­ties by pro­gram­ming in things like reduced strength or increased stiff­ness. He hypoth­e­sized that if the brain “knows” what a par­tic­ular muscle’s prop­er­ties are, then it will more readily learn how to maneuver the vir­tual arm if it’s mus­cles are pro­grammed to behave like real human mus­cles, rather than in a more arti­fi­cial way.

He’s done some ini­tial tests on his hypoth­esis and the work is under review for pub­li­ca­tion. You’ll have to stay tuned to find out the results!