http://ieeexplore.ieee.org/abstract/document/8016662/?reload=true
Abstract:
Evaluating
the effect of stroke rehabilitation based on electroencephalogram (EEG)
is still a challenging problem. This paper presents a novel nonlinear
dynamic complexity method for the evaluation of stroke rehabilitation
effect from EEG signal. Our method calculates the nonlinearly separable
degree (NLSD) of EEG signal, and then employs an indicator, called Mean
Nonlinearly Separable complexity Degree (Mean_NLSD), to efficiently and
accurately evaluate therapy effect of stroke patients. Our study under
twelve stimuli conditions on eleven patients and eleven control subjects
indicates that in general Mean_NLSD is smaller at the lesion regions
and that the Mean_NLSD of the control subjects is stochastic. Compared
with conventional spectral methods such as mean Power Spectral Density
(PSD), Mean_NLSD is more sensitive and robust. Overall Mean_NLSD may
offer a promising approach to facilitate the evaluation of stroke
rehabilitation effect.
Published in:
IEEE Transactions on Neural Systems and Rehabilitation Engineering
(
Volume: PP, Issue: 99
)
Publisher:
IEEE
Sponsored by:
IEEE Engineering in Medicine and Biology Society
No comments:
Post a Comment