Context information significantly improves brain computer interface performance - A case study on text entry using a language model assisted BCI

Umut Orhan, Deniz Erdogmus, Kenneth E. Hild, Brian Roark, Barry Oken, Melanie Fried-Oken

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

We present recent results on the design of the RSVP Keyboard - a brain computer interface (BCI) for expressive language generation for functionally locked-in individuals using rapid serial visual presentation of letters or other symbols such as icons. The proposed BCI design tightly incorporates probabilistic contextual information obtained from a language model into the single or multi-trial event related potential (ERP) decision mechanism. This tight fusion of contextual information with instantaneous and independent brain activity is demonstrated to potentially improve accuracy in a dramatic manner. Specifically, a simple regularized discriminant single-trial ERP classifier's performance can be increased from a naive baseline of 75% to 98% in a 28-symbol alphabet operating at 5% false ERP detection rate.We also demonstrate results which show that trained healthy subjects can achieve real-time typing accuracies over 90% mostly relying on single-trial ERP evidence when supplemented with a rudimentary n-gram language model. Further discussion and preliminary results include our initial efforts involving a locked-in individual and our efforts to train him to improve his skill in performing the task.

Original languageEnglish (US)
Title of host publicationConference Record - Asilomar Conference on Signals, Systems and Computers
Pages132-136
Number of pages5
DOIs
StatePublished - 2011
Event45th Asilomar Conference on Signals, Systems and Computers, ASILOMAR 2011 - Pacific Grove, CA, United States
Duration: Nov 6 2011Nov 9 2011

Other

Other45th Asilomar Conference on Signals, Systems and Computers, ASILOMAR 2011
CountryUnited States
CityPacific Grove, CA
Period11/6/1111/9/11

Fingerprint

Brain computer interface
Brain
Classifiers
Fusion reactions

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Signal Processing

Cite this

Orhan, U., Erdogmus, D., Hild, K. E., Roark, B., Oken, B., & Fried-Oken, M. (2011). Context information significantly improves brain computer interface performance - A case study on text entry using a language model assisted BCI. In Conference Record - Asilomar Conference on Signals, Systems and Computers (pp. 132-136). [6189970] https://doi.org/10.1109/ACSSC.2011.6189970

Context information significantly improves brain computer interface performance - A case study on text entry using a language model assisted BCI. / Orhan, Umut; Erdogmus, Deniz; Hild, Kenneth E.; Roark, Brian; Oken, Barry; Fried-Oken, Melanie.

Conference Record - Asilomar Conference on Signals, Systems and Computers. 2011. p. 132-136 6189970.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Orhan, U, Erdogmus, D, Hild, KE, Roark, B, Oken, B & Fried-Oken, M 2011, Context information significantly improves brain computer interface performance - A case study on text entry using a language model assisted BCI. in Conference Record - Asilomar Conference on Signals, Systems and Computers., 6189970, pp. 132-136, 45th Asilomar Conference on Signals, Systems and Computers, ASILOMAR 2011, Pacific Grove, CA, United States, 11/6/11. https://doi.org/10.1109/ACSSC.2011.6189970
Orhan U, Erdogmus D, Hild KE, Roark B, Oken B, Fried-Oken M. Context information significantly improves brain computer interface performance - A case study on text entry using a language model assisted BCI. In Conference Record - Asilomar Conference on Signals, Systems and Computers. 2011. p. 132-136. 6189970 https://doi.org/10.1109/ACSSC.2011.6189970
Orhan, Umut ; Erdogmus, Deniz ; Hild, Kenneth E. ; Roark, Brian ; Oken, Barry ; Fried-Oken, Melanie. / Context information significantly improves brain computer interface performance - A case study on text entry using a language model assisted BCI. Conference Record - Asilomar Conference on Signals, Systems and Computers. 2011. pp. 132-136
@inproceedings{a5899612c7e24445b966c1d0ffbf6ef2,
title = "Context information significantly improves brain computer interface performance - A case study on text entry using a language model assisted BCI",
abstract = "We present recent results on the design of the RSVP Keyboard - a brain computer interface (BCI) for expressive language generation for functionally locked-in individuals using rapid serial visual presentation of letters or other symbols such as icons. The proposed BCI design tightly incorporates probabilistic contextual information obtained from a language model into the single or multi-trial event related potential (ERP) decision mechanism. This tight fusion of contextual information with instantaneous and independent brain activity is demonstrated to potentially improve accuracy in a dramatic manner. Specifically, a simple regularized discriminant single-trial ERP classifier's performance can be increased from a naive baseline of 75{\%} to 98{\%} in a 28-symbol alphabet operating at 5{\%} false ERP detection rate.We also demonstrate results which show that trained healthy subjects can achieve real-time typing accuracies over 90{\%} mostly relying on single-trial ERP evidence when supplemented with a rudimentary n-gram language model. Further discussion and preliminary results include our initial efforts involving a locked-in individual and our efforts to train him to improve his skill in performing the task.",
author = "Umut Orhan and Deniz Erdogmus and Hild, {Kenneth E.} and Brian Roark and Barry Oken and Melanie Fried-Oken",
year = "2011",
doi = "10.1109/ACSSC.2011.6189970",
language = "English (US)",
isbn = "9781467303231",
pages = "132--136",
booktitle = "Conference Record - Asilomar Conference on Signals, Systems and Computers",

}

TY - GEN

T1 - Context information significantly improves brain computer interface performance - A case study on text entry using a language model assisted BCI

AU - Orhan, Umut

AU - Erdogmus, Deniz

AU - Hild, Kenneth E.

AU - Roark, Brian

AU - Oken, Barry

AU - Fried-Oken, Melanie

PY - 2011

Y1 - 2011

N2 - We present recent results on the design of the RSVP Keyboard - a brain computer interface (BCI) for expressive language generation for functionally locked-in individuals using rapid serial visual presentation of letters or other symbols such as icons. The proposed BCI design tightly incorporates probabilistic contextual information obtained from a language model into the single or multi-trial event related potential (ERP) decision mechanism. This tight fusion of contextual information with instantaneous and independent brain activity is demonstrated to potentially improve accuracy in a dramatic manner. Specifically, a simple regularized discriminant single-trial ERP classifier's performance can be increased from a naive baseline of 75% to 98% in a 28-symbol alphabet operating at 5% false ERP detection rate.We also demonstrate results which show that trained healthy subjects can achieve real-time typing accuracies over 90% mostly relying on single-trial ERP evidence when supplemented with a rudimentary n-gram language model. Further discussion and preliminary results include our initial efforts involving a locked-in individual and our efforts to train him to improve his skill in performing the task.

AB - We present recent results on the design of the RSVP Keyboard - a brain computer interface (BCI) for expressive language generation for functionally locked-in individuals using rapid serial visual presentation of letters or other symbols such as icons. The proposed BCI design tightly incorporates probabilistic contextual information obtained from a language model into the single or multi-trial event related potential (ERP) decision mechanism. This tight fusion of contextual information with instantaneous and independent brain activity is demonstrated to potentially improve accuracy in a dramatic manner. Specifically, a simple regularized discriminant single-trial ERP classifier's performance can be increased from a naive baseline of 75% to 98% in a 28-symbol alphabet operating at 5% false ERP detection rate.We also demonstrate results which show that trained healthy subjects can achieve real-time typing accuracies over 90% mostly relying on single-trial ERP evidence when supplemented with a rudimentary n-gram language model. Further discussion and preliminary results include our initial efforts involving a locked-in individual and our efforts to train him to improve his skill in performing the task.

UR - http://www.scopus.com/inward/record.url?scp=84861325069&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84861325069&partnerID=8YFLogxK

U2 - 10.1109/ACSSC.2011.6189970

DO - 10.1109/ACSSC.2011.6189970

M3 - Conference contribution

AN - SCOPUS:84861325069

SN - 9781467303231

SP - 132

EP - 136

BT - Conference Record - Asilomar Conference on Signals, Systems and Computers

ER -