Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Figure 1.

An illustration of the effects of higher order artefacts.

Contralateral 1st order artefacts are instantaneous, but there are smaller 2nd order artefacts from the true sources and the ipsilateral VC ‘source’ that are phase-lagged. In addition, there are also much weaker 3rd order artefacts between the VC ‘sources’. This is a simplified representation of the coherence patterns reported in supplementary material.

More »

Figure 1 Expand

Figure 2.

Positions of sources used in all simulations.

MNI coordinates: [−40.5 −72.5 17] and [42 −72.5 17]. Colour bar indicates normalised source strength.

More »

Figure 2 Expand

Figure 3.

Reconstructed network with phase jitter FWHM of 0 and π/4 calculated from full and NZPL sCSDs for EEG and MEG.

Colour code indicates the proportion of trials in which a voxel pair lies in the top 0.01% of the coherence matrix. Blue indicates location of the original network nodes.

More »

Figure 3 Expand

Figure 4.

Log ROC curves for performance of the DICS reconstruction for noise-normalised coherences for EEG and MEG reconstructions, plotted for full and NZPL sCSDs, for zero and non-zero (π/4) phase jitters.

FPR = false positive rate, TPR = true positive rate.

More »

Figure 4 Expand

Figure 5.

Log ROC AUCs calculated for noise-normalised coherence for EEG and MEG reconstructions, across phase jitter FWHMs.

More »

Figure 5 Expand

Table 1.

Results of all t-tests comparing AUCs with critical AUC, and comparing AUCs between sCSD types for all phase jitter FWHMs.

More »

Table 1 Expand

Figure 6.

Log ROC AUCs calculated for noise-normalised coherence for EEG and MEG reconstructions, across phase lags.

More »

Figure 6 Expand

Table 2.

Results of all t-tests comparing AUCs with critical AUC, and comparing AUCs between sCSD types for all phase lags.

More »

Table 2 Expand

Figure 7.

Estimated coherence calculated from imaginary coherency and NZPL coherence for a pair of sources with varying phase lag and SNR.

White lines indicates estimated thresholds for coherences of 0.9.

More »

Figure 7 Expand

Figure 8.

Log ROC AUCs for noise normalised coherences for EEG and MEG reconstructions, across regularisation parameters α.

More »

Figure 8 Expand

Figure 9.

Source Coherence estimates from human EEG using DICS with NZPL sCSD.

(a) shows the maximum value for each voxel value in the noise-normalised coherence matrix for faces (top) and scrambled (bottom) images at 0–400 ms, 4–8 Hz. (b) shows the maximum of the contrast between faces and scrambled connectivity matrices. (c) shows the top 0.01% of the face-scrambled contrast matrix.

More »

Figure 9 Expand