The National Institute on Deafness and Other Communication Disorders (NIDCD), one of the National Institutes of Health (NIH), held a programmatic workshop, Towards AAC-BCI Synergy, on September 17, 2015. The intent was to explore ways to move towards synergy between the Augmentative and Alternative Communication (AAC) and the Brain-Computer Interface (BCI) research fields. The virtual workshop used the WebEx platform to connect presenters and participants. Scot Ryder of the NIDCD Information Systems Management Branch (ISMB) served as the WebEx resource staff member.
Background
The NIDCD held its first workshop on AAC in 1994. With recent advances in BCI development, the NIDCD held a workshop in 2006 on AAC-BCI, focusing on BCI for speech synthesis. Despite growth in the NIDCD voice and speech research portfolio, there was a need to foster better collaboration across the fields of AAC and BCI and to work towards synergy between them.
Participants
The concept for the workshop emerged from professional discussions between Roger Miller, Ph.D., and Lana Shekim, Ph.D., of the NIDCD Division of Scientific Programs (DSP). They organized the workshop in consultation with Melanie Fried-Oken, Ph.D., of the Oregon Health and Science University, and Richard Ellenson, CEO of Cerebral Palsy International Research Foundation (CPIRF). Drs. Fried-Oken and Shekim co-chaired the workshop.
Pre-Workshop Discussion
The need for such a workshop became apparent to Drs. Miller and Shekim after they attended the 5th International BCI meeting, held in Asilomar, California, in 2013. In preparation for this workshop, Dr. Shekim visited the laboratory of Dr. Fried-Oken in Portland, Oregon, in May 2015. There she experienced their BCI system and met with potential users of the system (individuals with severe speech and physical impairments, or SSPI, and their caregivers). Dr. Fried-Oken, Dr. Shekim, and Dr. Miller further refined the plans for the workshop in consultation with Mr. Ellenson of CPIRF.
Workshop Format
The online workshop began at 11 a.m. (EST) with welcoming comments by Dr. Shekim. Sixteen researchers (participants list) were invited to participate in the workshop and offer their expertise in the discussion. There were five presentations organized into two major sessions, with discussion time after each session.
The first session started with a presentation from Dr. Fried-Oken, who gave an overview of the current relationship between the AAC and BCI fields. She began by reviewing recommendations from previous NIDCD workshops. Dr. Fried-Oken identified six challenges facing the fields. She then offered four charges to the BCI field from an AAC perspective.
Dr. Jane Huggins gave the second presentation, which focused on the engineering challenges for AAC and BCI. She reviewed seven challenges, some of which were: performance metrics, brain signals, intra-subject variability, calibration, and integration of AAC-BCI.
Dr. Huggins’ presentation was followed by a presentation from Dr. Leigh Hochberg on neurosciences challenges. Dr. Hochberg began by pointing to the changing landscape among the neurosciences, neuro-engineering, neurology/neuro-surgery/physical medicine and AAC fields. He reviewed challenges for AAC, challenges for BCI, and challenges for both AAC and BCI.
The second session began with a presentation from Ms. Theresa Vaughan, who focused on clinical issues relative to the adult population. The presentation outlined the three groups of interest: researchers, clinicians, and users, and stressed the importance of taking into consideration input from users. She then presented on the experience of the Wadsworth group to date.
Ms. Vaughan’s presentation was followed by a presentation from Dr. MaryAnn Romski on clinical issues relative to the pediatric population. She began with related research advances from the 1990s (e.g., cochlear implants) and moved to current practice (e.g., eye tracking AAC). Dr. Romski also outlined the two groups of pediatric users: those with congenital conditions and those with acquired conditions. Dr. Romski’s presentation focused on BCI adaptations and considered potential uses, such as assessment and integrating BCI use for intervention.
Discussion
Following the presentations, the group was asked to name and discuss the challenges in the path towards synergy between the AAC and BCI fields so that NIDCD-funded research can better serve individuals with SSPI. The following is a summary, as noted by Dr. Fried-Oken:
Defining the population of potential BCI users:
The BCI literature contains a wide range of descriptions of research participants. The field lacks a consensus regarding the target population for BCI technology. A clear, consistent description of this population would help researchers and developers understand the needs of their target users. It would be beneficial to describe the scope of different age groups with a wide range of disabilities, complications from motor impairments, and cognitive/linguistic impairments (both developmental and acquired conditions).
User-centered design:
The needs, abilities, limitations, and goals of people who may benefit from BCI technology are not always appreciated fully by researchers. The difficulties in expression experienced by people with SSPI naturally contribute to this. AAC clinician-researchers can assist with this. People with SSPI are seldom included in BCI research and development (R&D) teams, so they lack the opportunity to provide input on how BCI can be developed to best meet their needs. As one expert put it, this situation can lead to “researchers [who] don’t understand the needs of the target population and often provide ‘example’ applications that have no real applicability.”
The concepts of user-centered design, participatory action research, and patient-centered outcomes should be applied to ensure that BCIs will meet the needs of users. BCIs should be made available to users with a wide variety of physical, cognitive, and linguistic limitations and abilities.
Multidisciplinary collaboration:
BCI research frequently takes place in engineering labs, with investigators who have limited experience in communication sciences, AAC, or working with individuals with disabilities. Researchers and clinicians with expertise in these areas may see the value in BCI technology for people with SSPI, but lack the technical and engineering skills required to create it. These groups (and others, including neuroscientists, natural language processing experts, etc.) can all provide valuable contributions to BCI R&D, and can make the greatest impact on the field—and on the lives of BCI users—by working together. Unfortunately, multidisciplinary teams in this domain of research work are rare, and many research groups that would like to establish such teams have no easy way of connecting with experts in other fields.
Improving BCI technology:
Current BCI technology leaves considerable room for improvement in areas such as accuracy, reliability, convenience, and ease of use. BCI must be considered in the context of the larger field of assistive technology (AT); for some users, a stand-alone BCI may be appropriate, while others may use BCI to supplement or interface with other AT.
Planning for technology transfer and clinical implementation:
Numerous issues must be addressed before BCI technology becomes available to individuals with SSPI for independent home use. In addition to the technology improvements described above, BCI users will need access to well-informed prescribers and clinicians, training resources, and technical support services. Funding issues, such as health insurance reimbursement for BCI-based AAC technology, must also be considered.
Obstacles for BCI researchers:
BCI R&D groups applying for grant funding have encountered difficulty finding a review group that understands and recognizes the importance of this topic. The current funding environment and academic culture may discourage multidisciplinary collaboration and leave researchers with little time to do the work that needs to be done.
