Objective: In this study a gaze independent event related potential (ERP)-based brain computer interface (BCI) for communication purpose was combined with an asynchronous classifier endowed with dynamical stopping feature. The aim was to evaluate if and how the performance of such asynchronous system could be negatively affected in terms of communication efficiency and robustness to false positives during the intentional no-control state. Material and methods: The proposed system was validated with the participation of 9 healthy subjects. A comparison was performed between asynchronous and synchronous classification technique outputs while users were controlling the same gaze independent BC! interface. The performance of both classification techniques were assessed both off-line and on-line by means of the efficiency metric introduced by Bianchi et al. (2007). This latter metric allows to set a different misclassification cost for wrong classifications and abstentions. Robustness was evaluated as the rate of false positives occurring during voluntary no-control states. Results: The asynchronous classifier did not exhibited significantly higher accuracy or lower error rate with respect to the synchronous classifier (accuracy: 74.66% versus 87.96%, error rate: 7.11% versus 12.04% respectively). However, the on-line and off-line analysis revealed that the communication efficiency was significantly improved (p < .05) with the asynchronous classification modality as compared with the synchronous. Furthermore, the asynchronous classifier proved to be robust to false positives during intentional no-control state which occur during the ongoing visual stimulation (less than 1 false positive every 6 min). Conclusion: As such, the proposed ERP-BCI system which combines an asynchronous classifier with a gaze independent interface is a promising solution to be further explored in order to increase the general usability of ERP-based BCI systems designed for severely disabled people with an impairment of the voluntary control of eye movements. In fact, the asynchronous classifier can improve communication efficiency automatically adapting the number of stimulus repetitions to the current user's state and suspending the control if he/she does not intend to select an item. (C) 2013 Elsevier B.V. All rights reserved.

Asynchronous gaze-independent event-related potential-based brain-computer interface / Aloise, Fabio; Aricò, Pietro; Schettini, Francesca; Salinari, Serenella; Donatella, Mattia; Cincotti, Febo. - In: ARTIFICIAL INTELLIGENCE IN MEDICINE. - ISSN 0933-3657. - 59:2(2013), pp. 61-69. [10.1016/j.artmed.2013.07.006]

Asynchronous gaze-independent event-related potential-based brain-computer interface

ALOISE, FABIO
;
Aricò, Pietro;SCHETTINI, FRANCESCA;SALINARI, Serenella;CINCOTTI, FEBO
2013

Abstract

Objective: In this study a gaze independent event related potential (ERP)-based brain computer interface (BCI) for communication purpose was combined with an asynchronous classifier endowed with dynamical stopping feature. The aim was to evaluate if and how the performance of such asynchronous system could be negatively affected in terms of communication efficiency and robustness to false positives during the intentional no-control state. Material and methods: The proposed system was validated with the participation of 9 healthy subjects. A comparison was performed between asynchronous and synchronous classification technique outputs while users were controlling the same gaze independent BC! interface. The performance of both classification techniques were assessed both off-line and on-line by means of the efficiency metric introduced by Bianchi et al. (2007). This latter metric allows to set a different misclassification cost for wrong classifications and abstentions. Robustness was evaluated as the rate of false positives occurring during voluntary no-control states. Results: The asynchronous classifier did not exhibited significantly higher accuracy or lower error rate with respect to the synchronous classifier (accuracy: 74.66% versus 87.96%, error rate: 7.11% versus 12.04% respectively). However, the on-line and off-line analysis revealed that the communication efficiency was significantly improved (p < .05) with the asynchronous classification modality as compared with the synchronous. Furthermore, the asynchronous classifier proved to be robust to false positives during intentional no-control state which occur during the ongoing visual stimulation (less than 1 false positive every 6 min). Conclusion: As such, the proposed ERP-BCI system which combines an asynchronous classifier with a gaze independent interface is a promising solution to be further explored in order to increase the general usability of ERP-based BCI systems designed for severely disabled people with an impairment of the voluntary control of eye movements. In fact, the asynchronous classifier can improve communication efficiency automatically adapting the number of stimulus repetitions to the current user's state and suspending the control if he/she does not intend to select an item. (C) 2013 Elsevier B.V. All rights reserved.
File allegati a questo prodotto
File Dimensione Formato  
Aloise_Asynchronous-gaze-independent_2013.pdf

solo gestori archivio

Tipologia: Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza: Tutti i diritti riservati (All rights reserved)
Dimensione 1.95 MB
Formato Adobe PDF
1.95 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
VE_2013_11573-536210.pdf

solo gestori archivio

Tipologia: Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza: Tutti i diritti riservati (All rights reserved)
Dimensione 1.95 MB
Formato Adobe PDF
1.95 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11573/536210
Citazioni
  • ???jsp.display-item.citation.pmc??? 11
  • Scopus 30
  • ???jsp.display-item.citation.isi??? 27
social impact