Dynamics of the spike-wave paroxysms in Childhood Absence Epilepsy (CAE) are automatically characterized using novel approaches. Features are extracted from scalograms formed by Continuous Wavelet Transform (CWT). Detection algorithms are designed to identify an estimate of the temporal development of frequencies in the paroxysms. A database of 106 paroxysms from 26 patients was analyzed. The database is large compared to other known studies in the field of dynamics in CAE. CWT is more efficient than the widely used Fourier transform due to CWTs ability to recognize smaller discontinuities and variations. The use of scalograms and the detection algorithms result in a potentially usable clinical tool for dividing CAE patients into subsets. Differences between the grouped paroxysms may turn out to be useful from a clinical perspective as a prognostic indicator or when adjusting drug treatment.
Ieee Engineering in Medicine and Biology Society Conference Proceedings, 2013, p. 4283-4286
drugs; electroencephalography; feature extraction; Fourier transforms; medical disorders; medical signal detection; paediatrics; patient treatment; wavelet transforms; Engineered Materials, Dielectrics and Plasmas
Main Research Area:
35th Annual International Conference of the IEEE EMBS, 2013