Trending Articles
Trending Topics
Trending
IBM has teamed up with neurotech firm Inclusive Brains to help bring brain-machine interfaces closer to the mainstream.
The pair have announced they’re working together on a study that will use artificial intelligence and quantum computing to improve the performance of brain-machine interfaces (BMIs).
BMIs, sometimes known as brain-computer interfaces or BCIs, are exactly what they sound like: electronic systems that convert the electrical signals in the human brain into electrical signals that can be understood by software and hardware.
While Elon Musk’s Neuralink is probably the best known of the BMI companies, BMIs have been used experimentally in humans for decades.
BMIs’ most interesting, and potentially life-changing, use cases lie in helping people with disabilities.
Instructions from the brain are normally transmitted to the body by means of impulses carried by nerve pathways. When these pathways are disrupted, those impulses can’t be carried as they would normally.
Left without instructions, parts of the body stop functioning. For example, a spinal injury might stop electrical signals getting from the brain to the legs, causing a person to become paralysed.
BMIs aim to gather the signals from the brain that would normally convey instructions to the body and route them into peripherals. Instructions intended to go from the brain to the arm could instead be used to instruct a robotic arm, a keyboard or smart home controls, allowing people with disabilities to have more control over their environment.
Such systems often require electrodes to be installed on the surface of the brain. However, other companies – including Meta – have investigated whether external hardware such as headsets could gather enough meaningful information about the brain’s activity to control peripherals.
The IBM-Inclusive Brains study will use a non-invasive “AI system trained with brainwaves, facial movements and expressions, eye movements, and other physiological signals to control connected devices or digital environments”, according to the companies.
The partnership between IBM and Inclusive Brains aims to make the way BMI systems interpret patterns of brain activity more reliable using AI. IBM’s Granite foundation models will help find the most efficient algorithms for interpreting and classifying brain activity, by building code that can create benchmarks to test the algorithms.
Each person’s brain activity is different, so BMIs must be customised to their user. Granite will help select the best algorithms for each individual, which can then be used to operate workstations.
IBM and Inclusive Brains will also investigate the potential of quantum computing to classify brain activity, and plan to publish the results of their study in open research journals.
