How brain science will be flipped on its head with this innovative headset

29 Jun 2017

A new EEG headset could turn the world of neuroscience upside down. Image: Marius Pirvu/Shutterstock

TechWatch’s Emily McDaid gets down to nitty-gritty of neuroscience data with Brainwavebank founder Brian Murphy.

Belfast start-up Brainwavebank is applying machine learning to the science of your thoughts.

The company was started in 2015 by a computational neuroscientist, who had previously co-founded a web company. He saw that brain data, if collected en masse, had many possible applications – not least in assessing cognitive health.

Future Human

Founder Brian Murphy said: “We wanted to see if we could produce an EEG headset for under $100, to start collecting more data from more people. Brain data is traditionally noisy and expensive to collect. The more data we have, the more we know about the brain.”

In terms of technology, a brainwave or EEG detector is essentially a very responsive voltmeter. The brain produces tiny electrical signals that echo on the scalp, which highly sensitive equipment can detect.

“The ultimate goal of Brainwavebank is to gather the biggest collection of EEG data on the planet, and to apply machine learning to gain insights from that data. Hence, the hardware had to be inexpensive,” explained Murphy.

Customers will be able to track their daily brain health using the wireless headset, while playing mobile games for a few minutes a day. Machine learning and brain-reading technologies are used together to build a detailed record of the person’s cognitive health and how it varies over time. Insights and advice are fed back to the user, on how their individual lifestyle factors affect their day-to-day cognitive performance.

‘There will be nothing like this in the world of brain science’

I asked Murphy what stage the young company is at, knowing it has already secured an impressive £1m through private investors and competitive R&D grants.

“We’ve been running a field trial of the platform since late last year with a professional sports team, and we have advanced prototypes of the hardware, which we’ll be releasing to alpha testers later in the summer,” he said.

An important goal of the system is to flag early signs of abnormal brain function. Depressed cognitive function and performance could, for example, be an early sign of dementia.

“Once neurodegenerative diseases, such as dementia, have set in, there aren’t really any effective therapies. These diseases need to be caught early,” said Murphy.



EEG headset. Image: Brainwavebank

Is this is for everyone, or just for those with higher risk of developing of dementia?

The first people to use it will be people who already have reason to believe they could develop dementia and who want to track their cognitive performance. But, eventually, all people can use it.

Is this brain training?

Evidence shows that being cognitively challenged every day is critical. But perhaps brain training isn’t the best way to achieve that – even just a regular coffee morning with your friends can provide the right stimulus. Having an active, interesting and healthy lifestyle is more important.

How exactly is machine learning used on the data from your headsets?

We use a lot of signal processing to clean up the data. We filter out particular brainwaves that are markers for different cognitive functions like memory, decision-making, concentration or cognitive flexibility. Then we use time-series machine learning methods to extract the features that are of interest to us.

What’s the business model?

One company that inspired us was 23andMe, a consumer-focused DNA company, where people send in DNA samples to get insights into their own health, and they’re also asked if their data can be used for the greater good. So, people are tracking their personal health while also helping humanity. We’ll be doing the same for cognitive health. The pricing will be based on a headset purchase price, plus a subscription fee.

What’s the future for the company?

It’s important to remember that machine learning and AI can only show its prowess with huge datasets. This will be world-changing once we have built a database the size of what we’re attempting.

There will be nothing like this in the world of brain science.

By Emily McDaid, editor, TechWatch

A version of this article originally appeared on TechWatch

TechWatch by Catalyst covered tech developments in Northern Ireland