Dynamic inverse problem solution using a kalman filter smoother for neuronal activity estimation

  • Eduardo Giraldo-Suárez Universidad Tecnológica de Pereira, Pereira
  • Jorge I. Padilla-Buriticá Universidad Nacional de Colombia sede Manizales, Manizales
  • César G. Castellanos-Domínguez Universidad Nacional de Colombia sede Manizales, Manizales
Keywords: Inverse problem, neuronal activity, Kalman filter, physiological model.

Abstract

This article presents an estimation method of neuronal activity into the brain using a Kalman smoother approach that takes into account in the solution of the inverse problem the dynamic variability of the time series. This method is applied over a realistic head model calculated with the boundary element method. A comparative analysis for the dynamic estimation methods is made up from simulated EEG signals for several noise conditions. The solution of the inverse problem is achieved by using high performance computing techniques and an evaluation of the computational cost is performed for each method. As a result, the Kalman smoother approach presents better performance in the estimation task than the regularized static solution, and the direct Kalman filter.

Author Biographies

Eduardo Giraldo-Suárez, Universidad Tecnológica de Pereira, Pereira
Programa de Ingeniería Eléctrica, Universidad Tecnológica de Pereira, Pereira
Jorge I. Padilla-Buriticá, Universidad Nacional de Colombia sede Manizales, Manizales
Programa de Ingeniería Electrónica, Universidad Nacional de Colombia sede Manizales, Manizales
César G. Castellanos-Domínguez, Universidad Nacional de Colombia sede Manizales, Manizales
Programa de Ingeniería Electrónica, Universidad Nacional de Colombia sede Manizales, Manizales
How to Cite
[1]
E. Giraldo-Suárez, J. I. Padilla-Buriticá, and C. G. Castellanos-Domínguez, “Dynamic inverse problem solution using a kalman filter smoother for neuronal activity estimation”, TecnoL., no. 27, pp. 33-51, Dec. 2011.

Downloads

Download data is not yet available.
Published
2011-12-20
Section
Articles

More on this topic