期刊名称：IAENG International Journal of Computer Science
印刷版ISSN：1819-656X
电子版ISSN：1819-9224
出版年度：2019
卷号：46
期号：2
页码：228-236
出版社：IAENG - International Association of Engineers
摘要：In this work, a novel identification methodof relevant Intrinsic Mode Functions, obtained fromElectroencephalographic signals, by using an entropy criteria isproposed. The idea is to reduce the number of Intrinsic ModeFunctions that are necessary for the electroencephalographicsource reconstruction. An entropy cost function is appliedon the Intrinsic Mode Functions generated by the EmpiricalMode Decomposition for automatic IMF selection. The resultingEnhanced Empirical Mode Decomposition is evaluated insimulated and real data bases containing normal and epilepticactivity by means of a relative error measure. The proposedapproach shows to improve the electroencephalographic sourcereconstruction specifically for epileptic seizure detection.
