A Cap Instead of a Brain Implant?

A Cap for Reading Brain Activity

A team of biomedical engineers from Texas has developed a special cap that reads brain activity and translates it into commands executed by a computer. Training such a brain-computer interface (BCI) typically requires calibration with extensive offline data collection to create an individualized decoder. This process is highly time-consuming. Moreover, the initial decoder may not be very effective because users do not receive real-time feedback during data collection to help maintain an optimal sensorimotor rhythm during calibration.

The Same Training for Everyone?

To address this issue, researchers developed a new machine-learning program capable of identifying individual needs and adjusting training based on repeated activity. Continuous self-calibration allows users to train the interface without supervision or guidance from a research team. It also eliminates the need for a complex medical procedure to install a brain implant.

The initial decoder was trained by an expert using a simple task of “balancing” the right and left sides of a digital bar on a computer screen. This decoder can then be used by other untrained users without lengthy calibration. Users simply put on a device resembling a swimming cap covered in electrodes (see Figure). These electrodes detect electrical signals from brain activity and transmit them to decoding software. During technology testing, 18 previously untrained users learned how to play a car-racing simulation game with the BCI over five days. The decoder was so effective that users could train on both the simpler bar-balancing task and the more complex racing game simultaneously.

Applications in Rehabilitation

In this preliminary research, the BCI was used by healthy volunteers without motor impairments. The research team now plans to adapt the technology for people with disabilities. They are, for example, working on developing a wheelchair controlled using this technology. At the South by Southwest Conference and Festivals in Austin, participants had the opportunity to try controlling two rehabilitation robots for arm and hand exercises. Several volunteers learned to operate the robots within a few minutes.

Non-Invasive Brain Support

The head of the research team, Prof. José del R. Millán, compares non-invasive BCIs to a sort of “brain exoskeleton” that transfers information from the user’s brain to the computer and back. Compared to brain implants, this technology is limited by the relatively low resolution of signals obtained through electroencephalography. In the future, this problem could potentially be solved by untested sensing technologies based on ultrasound or microwaves. The head of the scientific team believes that, with enough time and resources, systems could be developed where the BCI exoskeleton almost becomes part of the brain. “On one hand, we want to bring our BCI into the clinical field to help people with disabilities. On the other hand, we need to improve our technology to make it easier to use and more effective. The purpose of this technology is to assist people with disabilities in their daily lives,” said Professor Millán about the current research results.

Editorial Team, Medscope.pro

Sources:

1. Universal brain-computer interface lets people play games with just their thoughts. The University of Texas at Austin 2024 Mar 29. Available at: www.news.utexas.edu/2024/03/29/universal-brain-computer-interface-lets-people-play-games-with-just-their-thoughts

2. Kumar S., Alawieh H., Racz F. S. et al. Transfer learning promotes acquisition of individual BCI skills. PNAS Nexus 2024 Feb 16; 3 (2): pgae076, doi: 10.1093/pnasnexus/pgae076.

3. Paul A. This cap is a big step towards universal, noninvasive brain-computer interfaces. Popular Science 2024 Apr 1. Available at: www.popsci.com/technology/bci-wearable-cap

4. Webster P. The future of brain-computer interfaces in medicine. Nat Med 2024 Jun; 30 (6): 1508–1509, doi: 10.1038/d41591-024-00031-3.