Brain-Computer Interfaces (BCIs) for Communication and Control
Jonathan R. Wolpaw · 2007 · Proceedings of the 9th International ACM SIGACCESS Conference on Computers and Accessibility (Assets '07) · doi:10.1145/1296843.1296845
Summary
This keynote paper from Jonathan Wolpaw at the Wadsworth Center (New York State Department of Health) provides an overview of brain-computer interface (BCI) research aimed at developing augmentative communication and control technology for people with severe neuromuscular disorders. The target users include those with amyotrophic lateral sclerosis (ALS), brainstem stroke, and spinal cord injury — people who may be totally paralysed or "locked in" but cognitively intact. BCIs work by detecting the user's intent from brain signals recorded either non-invasively via scalp-recorded EEG, or from electrodes surgically implanted on the cortical surface (electrocorticography, or ECoG) or within the brain. These signals are translated in real time into commands for operating computers, communication devices, or neuroprostheses. A fundamental principle of BCI operation is mutual adaptation: the user must learn to encode commands in their brain signals while the BCI system must learn to derive commands from those signals, requiring both to continually adapt to each other.
Key findings
The Wadsworth Center's research demonstrated that patients with motor disabilities can learn to control EEG sensorimotor rhythm amplitudes and use this control to move a cursor rapidly and accurately in one or two dimensions. By 2007, their EEG-based multidimensional control was comparable in speed and accuracy to that achieved with implanted electrodes, and they were developing sequential "reach and grasp" movement control. Parallel ECoG studies suggested even faster and more precise communication potential. The team had developed BCI2000, a general-purpose BCI software platform provided to over 120 labs worldwide. Crucially, the paper describes an early home-deployment effort: a simplified BCI home system using sensorimotor rhythms or P300 evoked potentials was being provided to severely disabled users. The first home user, a scientist with ALS retaining only eye movement, found the BCI superior to his eye-gaze system and used it 6-8 hours daily for email and other tasks. The team planned to expand to 20-30 home users and establish a clinical network for BCI deployment.
Relevance
This paper captures an important moment in BCI development — the transition from laboratory research to real-world home deployment for people with severe disabilities. For accessibility practitioners, it demonstrates that BCI technology is not merely theoretical but was already being used daily by people with ALS as a practical communication tool as early as 2007. The mutual adaptation principle (user and system must adapt to each other) has implications for how assistive technologies are designed and evaluated — it argues against one-size-fits-all approaches and for systems that learn from their users over time. The BCI2000 open platform model, shared with 120+ labs, also exemplifies how open-source approaches can accelerate assistive technology development. While BCI remains a specialised technology, it represents a critical last-resort communication channel for people who cannot use any conventional input method.
Tags: brain-computer interface · augmentative communication · neuromuscular disorders · ALS · EEG · electrocorticography · assistive technology · locked-in syndrome