Comparison of Principal Component Analysis and Multiscale Principal Component Analysis Performance on Autism Electroencephalography Signals

Electroencephalography (EEG) is a non-invasive method used to record brain electrical activity and is widely applied in the analysis of neurological conditions, such as autism spectrum disorder (ASD). However, EEG signals are often contaminated by noise from both internal and external sources, such as eye movements, muscle activity, and cardiac signals, which degrade signal quality and complicate the analysis process. Therefore, EEG signal processing requires specialized methods to suppress noise while preserving essential information within the signal. This study aims to compare the performance of Principal Component Analysis (PCA) and Multiscale Principal Component Analysis (MSPCA) in noise reduction of EEG signals from individuals with ASD. An EEG dataset from King Abdul Aziz University was utilized and processed using Python programming. The performance evaluation was conducted based on three key parameters: mean squared error (MSE), signal-to-noise ratio (SNR), and percentage root mean square difference (PRD). The experimental results demonstrate that MSPCA outperforms PCA in signal reconstruction and noise suppression. These findings have practical implications for researchers and clinicians, providing a more effective method for processing EEG signals in individuals with ASD.