Wavelet denoising to analyze electroencephalogram signal of perfect pitch and Non Perfect Pitch Subjects

Abstract

Brain responses can be measured using Electroencephalogram (EEG). In the raw EEG signal, there are plenty of information that may suggest many different activity in the brain. Time-based EEG analysis is required for quantify brain response in a very short time after the stimulus is applied. Single Trial Event-Related Potential (ERP) is commonly used to quantify effect of cognitive activity tied with time to EEG signal. The application of wavelet denoising in single trial ERP is used to analyze capability level from people who have perfect pitch ability based on P3 and N1 parameters. On experiments, EEG measurement was taken on Cz, T3, T4 point. EEG signal was recorded when subject heard pure tones with right pitch, with 10 cent shift, and 25 cent shift. From single trial processing of latency test we obtain that perfect pitch subject gave faster response on P3 component with 25 cent shift than non perfect pitch subject. Weak correlation was also found between P3 component on Cz measurement point and N1 component on T3 and T4 measurement point. From all three measurement points Cz measurement point gave the best consistency