Detection of epileptic spikes in electroencephalographic signals for patients with temporal lobe epilepsy using wavelets

This work describes a method for the detection of epileptic tips in an electroencephalographic (EEG) surface record taking a single channel. A pattern was identified when using the multiresolution analysis with a biortogonal wavelet after processing and analyzing with the Matlab Wavelet Toolbox, 207 tips records and 132 records of devices previously classified by the neurophysiologist. This pattern allowed us to design an algorithm for the detection of tips in patients with refractory episodes of the temporal lobe, based on the maximum voltages at each of the six reconstruction levels using the biortogonal wavelet 3.7. The algorithm was applied to over-records of patients with epilepsy, obtaining a sensitivity of 92% and a specificity of 80% in the diagnosis of epileptic tips.

• Autores: Eduardo Castaño, Nelson; Zapata, José Fernando; Villegas Gutiérrez, Jairo Alberto
• Fecha de publicación: 2009
• Institución: Universidad EAFIT
• Repositorio: Repositorio EAFIT
• Idioma: spa
• OAI Identifier: oai:repository.eafit.edu.co:10784/14514
• Acceso en línea: http://hdl.handle.net/10784/14514
• Palabra clave: Multiresolution Analysis; Biortogonal Wavelet; Detection Of Epileptic Tips; Análisis Multirresolución; Wavelet Biortogonal; Detección De Puntas Epilépticas
• License: Copyright (c) 2008 Nelson Eduardo Castaño, José Fernando Zapata, Jairo Villegas G.