January 2022 issue
January 31, 2022 | The worldwide market for EEG systems, estimated at about $1.2 billion by Neurotech Reports, is currently dominated by a handful of large medtech vendors such as Natus Medical, who offer hospitals and other clinical settings standalone, multichannel neurodiagnostic systems for diagnosing disorders such as epilepsy, stroke, and Alzheimer’s disease. But the emerging market for portable and home-based EEG systems stands to alter the competitive outlook in this space.
Recently, a team of researchers from Russia and the U.S. published open-source instructions on GitHub for a 24-channel EEG device that could make it easier for upstarts to enter the market. Ildar Rakhmatulin of South Ural State University, along with colleagues from Brainflow, North Carolina State University, and Skolkovo Institute of Science and Technology collaborated on the project. For roughly $350 and a trip to your local electronics store, an entrepreneur can build a device that weighs in at about five ounces, is 25 mm in radius, and has a peak input noise below 0.35 µV.
The device off-loads data via the TCP-IP protocol, which is not encrypted, but builders can certainly find ways to bake in their own security. Hobbyists are unlikely to make security features a priority while it is recommended that professional researchers using the device deploy the security, compliance, and privacy practices of their home institutions.
Both the price-point to build the device and the notion of making it open-source are key points of differentiation for this project. The average EEG of similar quality is around $1,000 USD—not something that is a priority form many hobbyists and even young labs, depending on their research focus. The authors confessed to an altruistic motivation in making the device freely available, but also noted the benefits provided back to the field. “None of the products on the market are perfect,” noted Rakhmatulin. “Everything can be improved and open source is the way to go. Today, we are on the threshold of when brain interfaces can enter our lives.”
Rakhmatulin went on to say that advances in machine learning provide enough computing power to better seek, find, and understand correlations in signals, but the availability of datasets is currently limited. He hopes that an economical, high-quality device such as this one will help to generate more and bigger data sets.
Little electronics experience is needed to assemble the device and Rakhmatulin noted his plans to post companion instructions to help users make their own device. Potential hobbyist uses include everything from thought-driven typing, robot controls, game applications, or even a lie detector test. Research and clinical applications mirror those already addressed by commercially available EEG devices.
Collaborating with Rakhmatulin were Mikhail Lebedev at Skoltech, Zachary Traylor and Chang Nam at NC State, and Andrey Parfenov at Brainflow in Moscow. They published their work in Experimental Brain Research.
