Using video game techniques to optimise brain research

A video game algorithm is assisting brain researchers gain new insights into brain cell activity, with fascinating results from the Queensland Brain Institute now published in Nature Communications.

The ultimate goal is to learn more about how the brain works.

Lead author and brain researcher Dr Tristan Wallis studies neurons in the brain, how they are connected and interact with one another.

Brain neurons “communicate with chemicals”, he explains. 

“So a signal passes down one nerve, and then when it gets to the end of the nerve, the nerve squirts some chemicals across a little gap to the next nerve, which then picks up the signal and keeps on going.”

These interactions are occurring at a sub microscopic level and happening lightning fast.

Which is why researchers, like Wallis, use high tech super-resolution microscopes to record those molecules moving around at nanometre scale inside the brain. 

The movies or data from these microscopes look like “little fluorescent bulbs moving around, hanging around together, bumping into each other, bouncing off each other.”

To Wallis, it looked a bit like an old-school computer game.

Qbi tristanwallis 21062021 21 copy
Dr Tristan Wallis / Credit: Supplied by University of Queensland

That thought — along with being trapped at home during COVID-19 lockdowns and unable to visit the lab — gave him the idea to draw on a computer game concept to better analyse the  data from the super-resolution microscopes.

“I am enough of a programmer to know that there’s these algorithms computer games employ, to very quickly be able to determine — if you’ve got hundreds or thousands of objects moving around on a screen at the same time  — which of them are bumping into each other. 

“It’s the core of most video games — which things are interacting, and when,” he says.

Wallis specifically wanted an algorithm used in computer games for a task called, ‘spatial indexing’, where programming and equations divide up the 3D virtual space in the game, and determine interactions, like collisions, that happen during the game play.

Working in a computer coding language Python, Wallis was able to find a suitable algorithm and apply it to the data coming from the microscopes.

Once he got the program working with artificial data, he started using real experimental data. “To everyone’s great joy, it worked for experimental data as well,” he says.

A key piece of information the video game algorithm adds is the timing of these chemical interactions within the brain — when molecules in the brain collide, or cluster together, knowing how fast they’re moving  — allows researchers to better analyse and gain deeper insights into what brain neurons are doing.

Wallis says, applying the algorithm is really just the starting point.

His work on the software and publication, supported by colleagues in the lab of Professor Frederic Meunier from the Queensland Brain Institute, means researchers can feed in brain cell activity data from the super-resolution microscopes, and very quickly generate results to analyse what’s happening, gaining more insights about the time component of molecular interactions. 

The ultimate goal for researchers is to better understand one of the biggest mysteries and challenges in science – how the brain works.

It demonstrates the value of scientists understanding computer programming, and even video game techniques, Wallis says.

“Some of the geniuses programming games, they’re pushing computers as far as they can possibly go, so they know all about what’s called code optimisation, [to] make something run as quickly and as well as possible on all the hardware that you can give it.”

For the brain researchers, the video game algorithm means researchers can “ask more questions and get more answers”.

Subscribe to ultramarine from riaus

Do you care about the oceans? Are you interested in scientific developments that affect them? Then our new email newsletter Ultramarine, launching soon, is for you. Click here to become an inaugural subscriber.

Please login to favourite this article.