Interfacing with Brain Computer Interfaces

Prof Maurice Mulvenna, University of Ulster

2

Interfacing with Brain Computer Interfaces

Outline

•  What is BCI? •  Advances & problems •  The BRAIN project

3

BCI Schematic

Types of BCIs

Signal acquisition: non-invasive / invasive (field-potentials / spiking activity)

Environment: exogenous / endogenous

Operation mode: synchronous (cue-paced) / asynchronous (self-paced)

Mental or Experimental strategy: focused/selective attention / specific mental tasks / operant conditioning

Signal patterns: ERD/ERS / P300 / SSEP / SCP / intracortical recordings

Selective Attention: SSVEP

Flickering light with a specific frequency evokes SSVEP pattern with the same frequency.

Herrmann et al, Exp. Brain Research 2001

Selective Attention: P300

Component of an evoked potential (e.g. VEP) elicited by selective/focused attention in an oddball experiment.

Donchin et al, IEEE Rehab 2000

Motor Imagery: ERD/ERS

Subjects produce specific signal patterns (ERD/ERS) by performing motor imagery.

Pfurtscheller et al, Proceedings IEEE, 2001

Recent advances

•  Improved sensors •  Numerous patient successes •  Home and field validation •  BCI approaches compared •  Parameters developed and assessed •  Improved signal processing •  BCI2000 software platform •  Standardized application interfaces

Present problems

•  BCI setup is slow and unpleasant •  Expert help is required for BCI: –  Setup –  Cleaning –  Customization

•  Very limited options with: –  Approach –  Parameters –  Interface –  Applications

Project

•  BCIs with Rapid Automated Interfaces for Nonexperts (BRAIN) will develop BCIs into practical assistive tools to enhance inclusion for users with impaired communication due to illness and injury (e.g., Cerebral Palsy, brain and spinal cord injury, and stroke).

•  Supported by European Commission’s ICT for Inclusion Unit, under EU FP7 grant agreement No. 224156

•  Academia, industry and service users to develop a Brain-Computer Interface system linked directly to assistive technology and services within the home environment

BRAIN consortium

•  University of Bremen •  Philips Research Europe •  University of Ulster •  The Cedar Foundation •  University of Warsaw •  Telefonica •  Twente Medical Systems Intl.

13

BRAIN solutions •  New electrodes and amplifier system eliminate:

–  Long setup times and discomfort –  Cleaning –  Expert help

•  Automated tools assess: –  Approach (ERD, SSVEP, P300) –  Parameters (spatial filters, features, stimuli) –  User preferences

•  New: –  Signal processing tools (SSVEP and other) –  Parameters (spatial filter, features, stimuli) –  Intuitive Graphical User Interface (IGUI) –  Universal Application Interface (UAI) –  Communications and Entertainment Package

BRAIN solutions II

•  Evaluation and testing: –  In field settings –  Including patients –  Strong ethical focus –  Integration within BRAIN

•  Dissemination and exploitation: –  Conferences and workshops –  Journal publications –  Product development –  Intellectual property –  Media publicity

Impact

•  Users with severe disabilities

•  Users with mild to moderate disabilities

•  Healthy users in limited situations

•  Scientific researchers and developers

•  European standards

Advances •  New electrodes and amplifier system eliminate: –  long setup times and discomfort –  Cleaning –  Expert help

•  Automated tools assess: –  approach (ERD, SSVEP, P300) –  Parameters (spatial filters, features, stimuli) –  User preferences

•  New: –  Intuitive Graphical User Interface (IGUI) –  Universal Application Interface (UAI) –  Communications and Entertainment Package

•  Reliability, flexibility, usability, accessibility 17

Discussion

•  Understanding the EEG, and EEG dysfunction associated with complex conditions

•  Calibration / re-calibration –  Variability

•  Evolution phase: Changing algorithms & approaches •  Still too much effort to operate! •  Training: practicality? •  Assistive technology of choice?

18

Engagement

19

Conclusions

•  Significant technical complexity •  Getting results outside the lab is still difficult •  Integration of the technologies is troublesome •  Consistency of results id difficult to achieve •  Recognition of need for person-centred approach •  Participants like being part of process but… •  Are we there yet?

20

Thanks and Acknowledgements

•  The BRAIN consortium gratefully acknowledge the support of the European Commission’s ICT for Inclusion Unit, under grant agreement No. 224156.

21