AlphaandThetaWaveLocalisation UsingHilbert-Huang Transform: Empirical StudyoftheAccuracy
Hassan Sharabaty, Jordi Martin, Daniel Esteve · 2006
Thispaperproposes toapplytheHilbert-Huang Transform (HHT)tolocate alpha andtheta wavesinan Electroencephalogram (EEG), inorder toestimate the vigilance level ofanoperator. TheHilbert-Huang Transform (HHT)isa recent technique fornon-stationary signal processing. Using theEmpirical ModeDecomposition (EMD),theHHT allows decomposing amulti-modal signal into asumof mono-contribution functions called Intrinsic Mode Functions (IMFs), thentocompute theinstantaneous frequency ofeachIMFbyusing Hilbert Transform (HT). Webriefly review thebasic principle ofthis method, then propose anempirical study oftheaccuracy ofthe Hilbert transform anda methodtoincrease this accuracy whentheamplitude oftheanalysed signal is small. Thelast section ofthis paper presents theresults obtained with ouralgorithm.