Back

Aamir Saeed Malik

Universiti Teknologi Petronas, Malaysia

Title: Reduction of ballistocardiogram (BCG) artifact from EEG-fMRI data

Biography

Aamir Saeed Malik completed his PhD from Gwangju Institute of Science & Technology, South Korea. Currently, he is Associate Professor and the director of Biomedical Technology Group, one of the Mission Oriented Research at Universiti Teknologi PETRONAS (UTP). In addition, he is the Editor-in-Chief of RESINNEX, a UTP research publication. He is the senior member of IEEE. He has published 3 books, more than 50 papers in reputed journals & 100 papers in conferences, has more than 10 patents and served on technical committees of various conferences and publications and he has received awards in various international exhibitions.

Abstract

There are various modalities for studying the brain including Electroencephalograph (EEG), functional Magnetic Resonance Imaging (fMRI) and Positron Emission Tomography (PET). This abstract deals with two such modalities, i.e., EEG and fMRI. EEG has very good temporal resolution while having poor spatial resolution while fMRI has excellent spatial resolution with poor temporal resolution. With the advancement in technology, it is now possible to acquire both EEG and fMRI simultaneously and hence get good spatial as well as temporal resolution. However, simultaneous acquisition results in few artifacts and one of the major artifact is called Ballistocardiogram (BCG) artifact. BCG is due to the cardiac pulsation inside the fMRI scanner. Generally, Electrocardiogram (ECG) or Electrooculogram (EOG) recordings are required to reduce BCG artifact in the traditional methods. In this abstract, a method is proposed to reduce the BCG artifact using a combination of two methods, Empirical Mode Decomposition (EMD) and Principal Component Analysis (PCA), and it does not require any such recordings. The method is compared with two well-known methods, i.e. average artifact subtraction (AAS) and optimal basis set (OBS). After the reduction of BCG artifact, alpha rhythm is used for the assessment of the methods. Time and frequency evaluation parameters are incorporated in order to assess the alpha rhythm. In addition to the simulated dataset, the EEG recordings outside the fMRI scanner are also used as reference for the assessment of the methods. The results show that the proposed method performance is superior to the AAS and OBS methods.