A real-time method to reduce ballistocardiogram artifacts from EEG during fMRI based on optimal basis sets (OBS).

Comput Methods Programs Biomed. 2016 Apr;127:114-25

Authors: Wu X, Wu T, Zhan Z, Yao L, Wen X

Abstract
BACKGROUND: The simultaneous acquisition of electroencephalogram (EEG) and functional magnetic resonance imaging (fMRI) provides both high temporal and spatial resolution when measuring brain activity. A real-time analysis during a simultaneous EEG-fMRI acquisition is essential when studying neurofeedback and conducting effective brain activity monitoring. However, the ballistocardiogram (BCG) artifacts which are induced by heartbeat-related electrode movements in an MRI scanner severely contaminate the EEG signals and hinder a reliable real-time analysis.
NEW METHOD: The optimal basis sets (OBS) method is an effective candidate for removing BCG artifacts in a traditional offline EEG-fMRI analysis, but has yet to be applied to a real-time EEG-fMRI analysis. Here, a novel real-time technique based on OBS method (rtOBS) is proposed to remove BCG artifacts on a moment-to-moment basis. Real-time electrocardiogram R-peak detection procedure and sliding window OBS method were adopted.
RESULTS: A series of simulated data was constructed to verify the feasibility of the rtOBS technique. Furthermore, this method was applied to real EEG-fMRI data to remove BCG artifacts. The results of both simulated data and real EEG-fMRI data from eight healthy human subjects demonstrate the effectiveness of rtOBS in both the time and frequency domains.
COMPARISON WITH EXISTING METHODS: A comparison between rtOBS and real-time averaged artifact subtraction (rtAAS) was conducted. The results suggest the efficacy and advantage of rtOBS in the real-time removal of BCG artifacts.
CONCLUSIONS: In this study, a novel real-time OBS technique was proposed for the real-time removal of BCG artifacts. The proposed method was tested using simulated data and applied to real simultaneous EEG-fMRI data. The results suggest the effectiveness of this method.

PMID: 27000294 [PubMed - in process]