HRV Feature Selection for Neonatal Seizure Detection: A Wrapper Approach

QSpace/Manakin Repository

HRV Feature Selection for Neonatal Seizure Detection: A Wrapper Approach

Show simple item record


dc.contributor.author Malarvili, M.B.
dc.contributor.author Mesbah, M
dc.contributor.author Boashash, B
dc.date.accessioned 2012-03-12T05:54:43Z
dc.date.available 2012-03-12T05:54:43Z
dc.date.issued 2007-11
dc.identifier.citation IEEE International Conference on Signal Processing and Communications, 2007, page(s): 864 - 867 en_US
dc.identifier.isbn 978-1-4244-1235-8
dc.identifier.other Digital Object Identifier : 10.1109/ICSPC.2007.4728456
dc.identifier.uri http://hdl.handle.net/10576/10801
dc.description This paper presents a number of newborn HRV features from time domain and TF domain that are used to classify the HRV epochs as either seizure related or non-seizure related, using a wrapper-based feature selection process and the MBD. (Additional details can be found in the comprehensive book on Time-Frequency Signal Analysis and Processing (see http://www.elsevier.com/locate/isbn/0080443354). In addition, the most recent upgrade of the original software package that calculates Time-Frequency Distributions and Instantaneous Frequency estimators can be downloaded from the web site: www.time-frequency.net. This was the first software developed in the field, and it was first released publicly in 1987 at the 1st ISSPA conference held in Brisbane, Australia, and then continuously updated). en_US
dc.description.abstract This work addresses the feature selection problem using a wrapper approach to select a feature subset to distinguish between the classes of newborn heart rate variability (HRV) corresponding to seizure and non-seizure. The method utilizes a filter as a pre-step to remove the irrelevant and redundant features from the original set of features to provide a starting feature subset for the wrapper. This reduces the computation load and the severity of the search operations involved in a wrapper approach. The goodness of the feature subset selected is compared over 3 different classifiers, namely linear classifier, quadratic classifier and k-Nearest Neighbour (k-NN) statistical classifiers in a leave-one-out (LOO) cross validation. It was found that the 1-NN outperformed the other classifiers resulting in significant reductions in feature dimensionality and achieving 85.7% sensitivity and 84.6% specificity. en_US
dc.language.iso en en_US
dc.publisher IEEE en_US
dc.subject HRV feature selection en_US
dc.subject feature extraction en_US
dc.subject heart rate variability en_US
dc.subject neonatal seizure detection en_US
dc.subject statistical classifier en_US
dc.subject wrapper approach en_US
dc.subject feature extraction en_US
dc.subject newborn heart rate variability en_US
dc.subject seizure en_US
dc.subject time-frequency analysis en_US
dc.subject MBD en_US
dc.subject Modified B distribution en_US
dc.subject time-frequency distributions en_US
dc.subject quadratic TFD en_US
dc.subject time-frequency features en_US
dc.title HRV Feature Selection for Neonatal Seizure Detection: A Wrapper Approach en_US
dc.type Article en_US

Files in this item

Files Size Format View
Boashash-Malarv ... onatal-seizure-wrapper.pdf 2.338Mb PDF View/Open

This item appears in the following Collection(s)

Show simple item record

Search QSpace


Advanced Search

Browse

My Account