Navigation and service

Single trial data analysis

MEG source analysis is usually applied to signal averages to reveal the most prominent stereotypic activity. However, a major drawback of this is that averaging will not preserve the temporal dynamics of each individual response. Information relating to weaker brain activity is largely suppressed, especially when multiple strong and weaker sources act in a coordinated manner. In contrast, single trial analysis retains the full temporal dynamics, but suffers from a poor signal-to-noise ratio, and is therefore rarely applied. Here we propose the concept of cross trial phase statistics (CTPS) to identify both weak and strong brain responses in independent components of MEG signals (Dammers et al., 2008; Dammers and Schiek, 2011).
In this project MEG signals from auditory stimulation and voluntary finger movement experiments were decomposed into independent components (ICs) using independent component analysis (ICA). ICA belongs to the class of blind source separation, which are used to separate signals mixtures into underlying informational components (Breuer et al., 2013; Dammers et al., 2013). The term “blind” emphasises that such methods aim to remove the superposition effect of different source signals blindly; even if very little is known about the nature of the underlying source signals. After decomposition we aim to automatically identify the source signals of interest (i.e., the signals that are involved in processing the information in the living human brain). Therefore, we apply cross-trial phase statistics (CTPS) to decomposed MEG signals (Breuer et al. 2014) and, as a result, we find strong and weak responses in phase dynamics to repetitive events of decomposed MEG signals. The selection of independent components based on CTPS is a highly sensitive approach for the extraction of weak event-related fields. After back-transformation of identified ICs, the signal-to-noise ratio of the MEG signals is highly improved, thus allowing for further single trial and connectivity analysis.

Single trial data analysis

Figure 1: (a) Identification of stereotypic brain responses using CTPS on decomposed MEG signals. MEG raw signals (grey) are shown as continuous responses to auditory stimuli (blue). One independent component (IC13) was extracted (lower light red trace) using CTPS for back-transformation into the MEG signal space (red). The signal-to-noise ratio of the unaveraged reconstructed MEG signal (red) is highly improved allowing for single trial source reconstruction.
(b) Localisation of single independent components, which can be attributed to responses from primary motor (red arrow) and somatosensory cortex (white arrow). In contrast, standard averaging (middle) reveals distributed activity over both cortices, only.


Wyss, C., Tse, D.H.Y., Kometer, M., Dammers, J., Achermann, R., Shah, N.J., Kawohl, W., Neuner, I., 2017. GABA metabolism and its role in gamma-band oscillatory activity during auditory processing: An MRS and EEG study. Hum. Brain Mapp. doi:10.1002/hbm.23642

Dammers, J., Fasoula, A., George, N., Schwartz, D., 2014. Enhanced Causality Analysis in Source Space based on Cross trial Phase Statistics, in: 19th International Conference on Biomagnetism, Halifax, Canada. Halifax, p. 1.

Breuer, L., Dammers, J.J., Roberts, T.P.L., Shah, N.J., 2014. A Constrained ICA Approach for Real-Time Cardiac Artifact Rejection in Magnetoencephalography. IEEE Trans. Biomed. Eng. 61, 405–414. doi:10.1109/TBME.2013.2280143

Dammers, J., Schiek, M., 2011. Magnetoencephalography, in: Pang, Elizabeth, W. (Ed.), Detection of Artifacts and Brain Responses Using Instantaneous Phase Statistics in Independent Components. InTech, pp. 131–150. doi:10.5772/1229

Dammers, J., Schiek, M., Boers, F., Weidner, R., Chen, Y.-H., Mathiak, K., Shah, N.J., 2010. Localization of stereotypic brain responses detected by instantaneous phase statistics from independent components. Front. Neurosci. Conf. Abstr. 17th Int. Conf. Biomagn. 2010.

Dammers, J., Schiek, M., Boers, F., Silex, C., Zvyagintsev, M., Pietrzyk, U., Mathiak, K., 2008. Integration of amplitude and phase statistics for complete artifact removal in independent components of neuromagnetic recordings. IEEE Trans. Biomed. Eng. 55, 2353–62. doi:10.1109/TBME.2008.926677