Search

Despite considerable effort to remove it, the ballistocardiogram (BCG) remains a major artifact in electroencephalographic data (EEG) acquired inside magnetic resonance imaging (MRI) scanners, particularly in continuous (as opposed to event-related) recordings. In this study, we have developed a new Direct Recording Prior Encoding (DRPE) method to extract and separate the BCG and EEG components from contaminated signals, and have demonstrated its performance by comparing it quantitatively to the popular Optimal Basis Set (OBS) method. Our modified recording configuration allows us to obtain representative bases of the BCG- and EEG-only signals. Further, we have developed an optimization-based reconstruction approach to maximally incorporate prior knowledge of the BCG/EEG subspaces, and of the signal characteristics within them. Both OBS and DRPE methods were tested with experimental data, and compared quantitatively using cross-validation. In the challenging continuous EEG studies, DRPE outperforms the OBS method by nearly sevenfold in separating the continuous BCG and EEG signals. © 2014 Xia, Ruan and Cohen.

Ballistocardiogram (BCG) artifact remains a major challenge that renders electroencephalographic (EEG) signals hard to interpret in simultaneous EEG and functional MRI (fMRI) data acquisition. Here, we propose an integrated learning and inference approach that takes advantage of a commercial high-density EEG cap, to estimate the BCG contribution in noisy EEG recordings from inside the MR scanner. To estimate reliably the full-scalp BCG artifacts, a near-optimal subset (20 out of 256) of channels first was identified using a modified recording setup. In subsequent recordings inside the MR scanner, BCG-only signal from this subset of channels was used to generate continuous estimates of the full-scalp BCG artifacts via inference, from which the intended EEG signal was recovered. The reconstruction of the EEG was performed with both a direct subtraction and an optimization scheme. We evaluated the performance on both synthetic and real contaminated recordings, and compared it to the benchmark Optimal Basis Set (OBS) method. In the challenging non-event-related-potential (non-ERP) EEG studies, our reconstruction can yield more than fourteen-fold improvement in reducing the normalized RMS error of EEG signals, compared to OBS. © 2014 Xia, Ruan and Cohen.

In simultaneous EEG/fMRI acquisition, the ballistocardiogram (BCG) artifact presents a major challenge for meaningful EEG signal interpretation and needs to be removed. This is very difficult, especially in continuous studies where BCG cannot be removed with averaging. In this study, we take advantage of a high-density EEG-cap and propose an integrated learning and inference approach to estimate the BCG contribution to the overall noisy recording. In particular, we present a special-designed experiment to enable a near-optimal subset selection scheme to identify a small set (20 out of 256 channels), and argue that in real-recording, BCG artifact signal from all channels can be estimated from this set. We call this new approach ``Direct Recording Temporal Spatial Encoding'' (DRTSE) to reflect these properties. In a preliminary evaluation, the DRTSE is combined with a direct subtraction and an optimization scheme to reconstruct the EEG signal. The performance was compared against the benchmark Optimal Basis Set (OBS) method. In the challenging non-event-related EEG studies, the DRTSE method, with the optimization-based approach, yields an EEG reconstruction that reduces the normalized RMSE by approximately 13 folds, compared to OBS.

The ballistocardiogram (BCG) is a major artifact in electroencephalographic (EEG) data acquired inside a magnetic resonance imaging (MRI) scanner, and is several times larger in magnitude than the actual EEG signals. Removing the BCG artifacts remains an unresolved challenge, especially in studies of continuous EEG recordings. In this work, we propose a Direct Recording-Joint Incoherent Basis (DRJIB) method to decompose the observed noisy EEG measurements into BCG and underlying EEG components. We compare its preliminary performance quantitatively with that of the benchmark Optimal Basis Set (OBS) method. Without assuming orthogonality or independence of the BCG and EEG subspaces, as in conventional methods, our approach learns the bases faithfully from BCG-only and EEG-only signals acquired from our new experimental setup. Specifically, to promote subspace separability, a paired set of low-dimensional and semi-orthogonal (BCG, EEG) basis representations is obtained by minimizing a cost function consisting of group sparsity penalties for automatic dimension selection and an energy term for encouraging incoherence. Reconstruction is subsequently obtained by fitting the contaminated data to a generative model using the learned bases subject to regularization. In the challenging non-event-related EEG studies, our DRJIB method outperforms the OBS method by nearly 12-fold in separating and preserving the continuous BCG and EEG signals.

Developing EEG-based computer aided diagnostic (CAD) tools would allow identification of epilepsy in individuals who have experienced possible seizures, yet such an algorithm requires efficient identification of meaningful features out of potentially more than 35,000 features of EEG activity. Mutual information can be used to identify a subset of minimally-redundant and maximally relevant (mRMR) features but requires a priori selection of two parameters: the number of features of interest and the number of quantization levels into which the continuous features are binned. Here we characterize the variance of cross-validation accuracy with respect to changes in these parameters for four classes of machine learning (ML) algorithms. This assesses the efficiency of combining mRMR with each of these algorithms by assessing when the variance of cross-validation accuracy is minimized and demonstrates how naA¯ve parameter selection may artificially depress accuracy. Our results can be used to improve the understanding of how feature selection interacts with four classes of ML algorithms and provide guidance for better a priori parameter selection in situations where an overwhelming number of redundant, noisy features are available for classification.