Brain-computer interface algorithm based on wavelet-phase stability analysis in motor imagery experiment

Abstract

Listen

Translate article

Severe movement or motor disability diseases such as amyotrophic lateral sclerosis (ALS), cerebral palsy (CB), and muscular dystrophy (MD) are types of diseases which lead to the total of function loss of body parts, usually limbs. Patient with an extreme motor impairment might suffers a locked-in state, resulting in the difficulty to perform any physical movements. These diseases are commonly being treated by a specific rehabilitation procedure with prescribed medication. However, the recovery process is time-consuming through such treatments. To overcome these issues, Brain-Computer Interface system is introduced in which one of its modalities is to translate thought via electroencephalography (EEG) signals by the user and generating desired output directly to an external artificial control device or human augmentation. Here, phase synchronization is implemented to complement the BCI system by analyzing the phase stability between two input signals. The motor imagery-based experiment involved ten healthy subjects aged from 24 to 30 years old with balanced numbers between male and female. Two aforementioned input signals are the respective reference data and the real time data were measured by using phase stability technique by indicating values range from 0 (least stable) to 1 (most stable). Prior to that, feature extraction was utilized by applying continuous wavelet transform (CWT) to quantify significant features on the basis of motor imagery experiment which are right and left imaginations. The technique was able to segregate different classes of motor imagery task based on classification accuracy. This study affirmed the approach’s ability to achieve high accuracy output measurements.