Abstract
The purpose of this study was to explore voice activity detection (VAD) in a subject with implanted electrocorticographic (ECoG) electrodes. Accurate VAD is an important preliminary step before decoding and reconstructing speech from ECoG. For this study we used ECoG signals recorded while a subject performed a picture naming task. We extracted time-domain features from the raw ECoG and spectral features from the ECoG high gamma band (70-110Hz). The RelieF algorithm was used for selecting a subset of features to use with seven machine learning algorithms for classification. With this approach we were able to detect voice activity from ECoG signals, achieving a high accuracy using the 100 best features from all electrodes (96%) or only 12 features from the two best electrodes (94%) using the support vector machines or a linear regression classifier. These findings may contribute to the development of ECoG-based brain machine interface (BMI) systems for rehabilitating individuals with communication impairments.
Original language | English (US) |
---|---|
Title of host publication | IFMBE Proceedings |
Pages | 1643-1646 |
Number of pages | 4 |
Volume | 41 |
DOIs | |
State | Published - 2014 |
Event | 13th Mediterranean Conference on Medical and Biological Engineering and Computing 2013, MEDICON 2013 - Seville, Spain Duration: Sep 25 2013 → Sep 28 2013 |
Other
Other | 13th Mediterranean Conference on Medical and Biological Engineering and Computing 2013, MEDICON 2013 |
---|---|
Country | Spain |
City | Seville |
Period | 9/25/13 → 9/28/13 |
Fingerprint
Keywords
- Brain machine interface
- Electrocorticography
- Machine learning
- Voice activity detection
ASJC Scopus subject areas
- Biomedical Engineering
- Bioengineering
Cite this
Voice activity detection from electrocorticographic signals. / Kanas, Vasileios G.; Mporas, I.; Benz, H. L.; Huang, N.; Thakor, Nitish V; Sgarbas, K.; Bezerianos, A.; Crone, Nathan E.
IFMBE Proceedings. Vol. 41 2014. p. 1643-1646.Research output: Chapter in Book/Report/Conference proceeding › Chapter
}
TY - CHAP
T1 - Voice activity detection from electrocorticographic signals
AU - Kanas, Vasileios G.
AU - Mporas, I.
AU - Benz, H. L.
AU - Huang, N.
AU - Thakor, Nitish V
AU - Sgarbas, K.
AU - Bezerianos, A.
AU - Crone, Nathan E
PY - 2014
Y1 - 2014
N2 - The purpose of this study was to explore voice activity detection (VAD) in a subject with implanted electrocorticographic (ECoG) electrodes. Accurate VAD is an important preliminary step before decoding and reconstructing speech from ECoG. For this study we used ECoG signals recorded while a subject performed a picture naming task. We extracted time-domain features from the raw ECoG and spectral features from the ECoG high gamma band (70-110Hz). The RelieF algorithm was used for selecting a subset of features to use with seven machine learning algorithms for classification. With this approach we were able to detect voice activity from ECoG signals, achieving a high accuracy using the 100 best features from all electrodes (96%) or only 12 features from the two best electrodes (94%) using the support vector machines or a linear regression classifier. These findings may contribute to the development of ECoG-based brain machine interface (BMI) systems for rehabilitating individuals with communication impairments.
AB - The purpose of this study was to explore voice activity detection (VAD) in a subject with implanted electrocorticographic (ECoG) electrodes. Accurate VAD is an important preliminary step before decoding and reconstructing speech from ECoG. For this study we used ECoG signals recorded while a subject performed a picture naming task. We extracted time-domain features from the raw ECoG and spectral features from the ECoG high gamma band (70-110Hz). The RelieF algorithm was used for selecting a subset of features to use with seven machine learning algorithms for classification. With this approach we were able to detect voice activity from ECoG signals, achieving a high accuracy using the 100 best features from all electrodes (96%) or only 12 features from the two best electrodes (94%) using the support vector machines or a linear regression classifier. These findings may contribute to the development of ECoG-based brain machine interface (BMI) systems for rehabilitating individuals with communication impairments.
KW - Brain machine interface
KW - Electrocorticography
KW - Machine learning
KW - Voice activity detection
UR - http://www.scopus.com/inward/record.url?scp=84891286469&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84891286469&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-00846-2_405
DO - 10.1007/978-3-319-00846-2_405
M3 - Chapter
AN - SCOPUS:84891286469
SN - 9783319008455
VL - 41
SP - 1643
EP - 1646
BT - IFMBE Proceedings
ER -