Speech Recognition Features for EEG Signal Description in Detection of Neonatal Seizures

Typeset version

 

TY  - CONF
  - Temko A., Boylan G., Marnane W.P., Lightbody G.
  - 32nd Annual International IEEE EMBS Conference
  - Speech Recognition Features for EEG Signal Description in Detection of Neonatal Seizures
  - 2010
  - September
  - Published
  - 1
  - ()
  - 3281
  - 3284
  - Buenos Aires, Argentina
  - 31-AUG-10
  - 04-SEP-10
  - In this work, features which are usually employed in automatic speech recognition (ASR) are used for the detection of neonatal seizures in newborn EEG. Threeconventional ASR feature sets are compared to the feature set which has been previously developed for this task. The results indicate that the thoroughly-studied spectral envelope based ASR features perform reasonably well on their own. Additionally, the SVM Recursive Feature Elimination routine is applied to all extracted features pooled together. It is shown that ASR features consistently appear among the top-rank features.
  - 10.1109/IEMBS.2010.5627260
  - Science Foundation Ireland
  - Welcome Trust (085249/Z/08/Z) and Science Foundation Ireland (07/SRC/I1169)
DA  - 2010/09
ER  - 
@inproceedings{V119277851,
   = {Temko A.,  Boylan G. and  Marnane W.P.,  Lightbody G. },
   = {32nd Annual International IEEE EMBS Conference},
   = {{Speech Recognition Features for EEG Signal Description in Detection of Neonatal Seizures}},
   = {2010},
   = {September},
   = {Published},
   = {1},
   = {()},
  pages = {3281--3284},
   = {Buenos Aires, Argentina},
  month = {Aug},
   = {04-SEP-10},
   = {{In this work, features which are usually employed in automatic speech recognition (ASR) are used for the detection of neonatal seizures in newborn EEG. Threeconventional ASR feature sets are compared to the feature set which has been previously developed for this task. The results indicate that the thoroughly-studied spectral envelope based ASR features perform reasonably well on their own. Additionally, the SVM Recursive Feature Elimination routine is applied to all extracted features pooled together. It is shown that ASR features consistently appear among the top-rank features.}},
   = {10.1109/IEMBS.2010.5627260},
   = {Science Foundation Ireland},
   = {Welcome Trust (085249/Z/08/Z) and Science Foundation Ireland (07/SRC/I1169)},
  source = {IRIS}
}
AUTHORSTemko A., Boylan G., Marnane W.P., Lightbody G.
TITLE32nd Annual International IEEE EMBS Conference
PUBLICATION_NAMESpeech Recognition Features for EEG Signal Description in Detection of Neonatal Seizures
YEAR2010
MONTHSeptember
STATUSPublished
PEER_REVIEW1
TIMES_CITED()
SEARCH_KEYWORD
EDITORS
START_PAGE3281
END_PAGE3284
LOCATIONBuenos Aires, Argentina
START_DATE31-AUG-10
END_DATE04-SEP-10
ABSTRACTIn this work, features which are usually employed in automatic speech recognition (ASR) are used for the detection of neonatal seizures in newborn EEG. Threeconventional ASR feature sets are compared to the feature set which has been previously developed for this task. The results indicate that the thoroughly-studied spectral envelope based ASR features perform reasonably well on their own. Additionally, the SVM Recursive Feature Elimination routine is applied to all extracted features pooled together. It is shown that ASR features consistently appear among the top-rank features.
FUNDED_BY
URL
DOI_LINK10.1109/IEMBS.2010.5627260
FUNDING_BODYScience Foundation Ireland
GRANT_DETAILSWelcome Trust (085249/Z/08/Z) and Science Foundation Ireland (07/SRC/I1169)