EEG Brain Symmetry Index Using Hilbert Huang Transform

Electroencephalography (EEG) monitoring is known to be technically feasible and possibly clinically relevant to determine patients with acute ischemic hemispheric stroke. The EEG is very useful tool in understanding neurological dysfunction of stroke plausible improving the treatment and rehabilitation. Most of the existing techniques to diagnose stroke from the EEG signal is mainly based on Fourier Transform (FT). For instance, the Brain Symmetry Index (BSI) employed Fast Fourier Transform (FFT) as coefficients to measure symmetrical of blood flow between left and right brain hemisphere. The symmetrical index ranges between zero and one where one indicates the highest asymmetrical of blood flow. It is known that the conventional FFT has limitation in analyzing non-linear and non-stationary signal. Therefore, the existing BSI and its variations may also suffer from this transformation properties. In this study, we propose BSI based on Hilbert Huang Transform (HHT) which defined as BSI-HHT. HHT is a way to decompose a signal into so-called intrinsic mode functions (IMF) along with a trend, and obtain instantaneous frequency data. The HHT will be used as coefficients instead off FFT in calculating the BSI index. An experiment to validate the performance of BSI-HHT is conducted in this study as to compare with the existing BSI technique. The EEG signal of Middle Cerebral Artery (MCA) subjects and healthy subjects are used for this investigation. The proposed BSI-HHT has offered better interpretation as it correlates to the stimulation procedure on the gathered data especially at specific frequency band. Also, through the analysis, the HHT coefficient is able to capture the non-stationary and non-linear of the interest electrode.