Brain-computer interface (BCI) technology aims at helping severely motor disabled people like the physicist Stephen Hawking, due to diseases such as amyotrophic lateral sclerosis, or spinal cord injury, etc. By decoding the brain signals, BCI enables those disabled people to communicate with the external world when losing control of their peripheral nerves and muscles. With the help of BCI, the disabled people can control home appliances, prostheses, even type letters. However, clinical applications proposed new challenges for such a promising technology: BCI using non-invasive scalp electroencephalography (EEG) can only work for short term due to low signal quality; BCI using invasive microelectrode suffers from high surgery risk and long-term signal degeneration due to encapsulation. Recently, Dr. Bo Hong’s team proposed and implemented a novel brain-computer interface using neural signals recorded on the surface of human brain, which has achieved both long-term usability and high performances.

In collaboration with the neurosurgery team in Tsinghua Affiliated Yuquan Hospital and PLA General Hospital, Bo Hong’s team from School of Medicine conducted the research with epilepsy patients who were implanted with intracranial electrodes for diagnosis purposes. The patients were presented with a virtual keyboard with hidden moving visual patterns. The patients selected the intended letters using their attention. The attended letters were found to elicit enhanced neural activity over 60 Hz in a focused brain region over middle temporal cortex, constituting the basis for BCI application. Functional magnetic resonance imaging (fMRI) showed that the target brain region matched the functional brain region for visual motion processing. Hereby, the dedicated BCI design allows localization of the target brain region by fMRI.

Using one surface electrode selected by fMRI, the team achieved a novel BCI for letter typing. Compared with existing BCI technology, the proposed technology achieved long-term stable neural signal recording while having the surgery risk minimized. The idea of implementing a minimally invasive brain-computer interface guided by non-invasive fMRI is proposed for the first time. This study is recently published in the journal NeuroImage.

More details could be found here: http://www.sciencedirect.com/science/article/pii/S1053811913000086