Home // BIOTECHNO 2011, The Third International Conference on Bioinformatics, Biocomputational Systems and Biotechnologies // View article
Authors:
Lingke Fan
David Evans
Keywords: ultrasound; automated emoblus identification; TCD; MFCC application; directional analysis, auditory perceptul evaluation; artifact rejection, knowledge-based systems
Abstract:
Directional auditory perception features and energy information were extracted from transcranial Doppler (TCD) ultrasound signals and were used to automatically identify micro-embolic events (MEEs) using a novel embolic signal analysis approach. Three directional analysis methods were evaluated for their MEE identification performance using data recorded during cardiac surgery. The analysis methods were based on Mel frequency cepstrum coefficients (MFCCs), linear frequency cepstrum coefficients (LFCCs) and linear spectral components (LSCs). The results of these preliminary off-line evaluations showed that: a) the auditory perception and energy features of Doppler signals could play an important role in the identification of MMEs; b) MFCC-based analysis seems to be superior to the other two methods, achieving a sensitivity of 95.99%, a specificity of 96.43% and a positive predictive value (PPV) of 95.64%. Future studies using larger data sets and more complicated detection implementation (a rather basic rule-based system was used in the detection stage here) could further confirm or improve the identification performance and robustness of MFCC-based systems
Pages: 20 to 25
Copyright: Copyright (c) IARIA, 2011
Publication date: May 22, 2011
Published in: conference
ISSN: 2308-4383
ISBN: 978-1-61208-137-3
Location: Venice/Mestre, Italy
Dates: from May 22, 2011 to May 27, 2011