mne_connectivity.spectral_connectivity_epochs

mne_connectivity.spectral_connectivity_epochs(data, names=None, method='coh', indices='lower', sfreq=None, *, mode='multitaper', fmin=None, fmax=inf, fdecim=1, faverage=False, tmin=None, tmax=None, mt_bandwidth=None, mt_adaptive=False, mt_low_bias=True, cwt_freqs=None, cwt_n_cycles=7, gc_n_lags=40, rank=None, n_components=1, block_size=1000, n_jobs=1, verbose=None)

Compute frequency- and time-frequency-domain connectivity measures.

The connectivity method(s) are specified using the “method” parameter. All methods are based on estimates of the cross- and power spectral densities (CSD/PSD) Sxy and Sxx, Syy.

Parameters:

dataarray_like, shape (n_epochs, n_signals, n_times) | Epochs | generator | EpochsSpectrum | EpochsTFR

The data from which to compute connectivity. Can be epoched time series data as an array-like or mne.Epochs object, or Fourier coefficients for each epoch as an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object. If time series data, the spectral information will be computed according to the spectral estimation mode (see the mode parameter). If an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object, existing spectral information will be used and the mode parameter will be ignored.

Note that it is also possible to combine multiple time series signals by providing a list of tuples, e.g.:

data = [(arr_0, stc_0), (arr_1, stc_1), (arr_2, stc_2)]

which corresponds to 3 epochs where arr_* is an array with the same number of time points as stc_*. Data can also be a list/generator of arrays, shape (n_signals, n_times), or a list/generator of mne.SourceEstimate or mne.VolSourceEstimate objects.

Changed in version 0.8: Fourier coefficients stored in an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object can also be passed in as data. Storing multitaper weights in mne.time_frequency.EpochsTFR objects requires mne >= 1.10.

namesarray_like | None

The names of the nodes of the dataset used to compute connectivity. If None (default), then names will be a list of integers from 0 to n_nodes. If a list of names, then it must be equal in length to n_nodes.

methodstr | list of str

Connectivity measure(s) to compute. These can be ['coh', 'cohy', 'imcoh', 'cacoh', 'mic', 'mim', 'plv', 'ciplv', 'ppc', 'pli', 'dpli', 'wpli', 'wpli2_debiased', 'gc', 'gc_tr']. These are:

• ‘coh’ : Coherence

• ‘cohy’ : Coherency

• ‘imcoh’ : Imaginary part of Coherency

• ‘cacoh’ : Canonical Coherency (CaCoh)

• ‘mic’ : Maximised Imaginary part of Coherency (MIC)

• ‘mim’ : Multivariate Interaction Measure (MIM)

• ‘plv’ : Phase-Locking Value (PLV)

• ‘ciplv’ : Corrected Imaginary PLV (ciPLV)

• ‘ppc’ : Pairwise Phase Consistency (PPC)

• ‘pli’ : Phase Lag Index (PLI)

• ‘pli2_unbiased’ : Unbiased estimator of squared PLI

• ‘dpli’ : Directed PLI (DPLI)

• ‘wpli’ : Weighted PLI (WPLI)

• ‘wpli2_debiased’ : Debiased estimator of squared WPLI

• ‘gc’ : State-space Granger Causality (GC)

• ‘gc_tr’ : State-space GC on time-reversed signals

Multivariate methods (['cacoh', 'mic', 'mim', 'gc', 'gc_tr']) cannot be called with the other methods.

indicestuple of array_like | 'lower' | 'upper' | 'all'

The indices of channels to compute connectivity between. If a tuple, then must contain two array-likes, where the first array represents the channel indices of the seeds, and the second array represents the channel indices of the targets. For multivariate methods, the seed and target indices should consist of nested arrays containing the channel indices for each multivariate connection. See the notes section for more information. If 'lower', compute the lower-triangular part of the connectivity matrix. If 'upper', compute the upper-triangular part of the connectivity matrix. If 'all', compute all connections. Default is 'lower'.

sfreqfloat | None

The sampling frequency. Required if data is an array-like.

mode'multitaper' | 'fourier' | 'cwt_morlet'

Spectrum estimation mode. Ignored if data is an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object.

fminfloat | tuple of float | None

The lower frequency of interest. Multiple bands are defined using a tuple, e.g., (8., 20.) for two bands with 8 Hz and 20 Hz lower freq. If None, the frequency corresponding to 5 cycles based on the epoch length is used. For example, with an epoch length of 1 sec, the lower frequency would be 5 / 1 sec = 5 Hz.

fmaxfloat | tuple of float

The upper frequency of interest. Multiple bands are defined using a tuple, e.g., (13., 30.) for two bands with 13 Hz and 30 Hz upper freq.

fdecimint | None

Decimation factor in the frequency domain. Selects every Nth frequency bin from the (time-)frequency decomposition (where N is the value of fdecim). If 1 (default), no decimation occurs.

Added in version 0.8.

faveragebool

Average connectivity scores for each frequency band. If True, the output freqs will be a list with arrays of the frequencies that were averaged.

tminfloat | None

Time to start connectivity estimation. Note: when data is an array-like, the first sample is assumed to be at time 0. For mne.Epochs, the time information contained in the object is used to compute the time indices. Ignored if data is an mne.time_frequency.EpochsSpectrum object.

tmaxfloat | None

Time to end connectivity estimation. Note: when data is an array-like, the first sample is assumed to be at time 0. For mne.Epochs, the time information contained in the object is used to compute the time indices. Ignored if data is an mne.time_frequency.EpochsSpectrum object.

mt_bandwidthfloat | None

The bandwidth of the multitaper windowing function in Hz. Only used in 'multitaper' mode. Ignored if data is an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object.

mt_adaptivebool

Use adaptive weights to combine the tapered spectra into PSD. Only used in 'multitaper' mode. Ignored if data is an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object.

mt_low_biasbool

Only use tapers with more than 90 percent spectral concentration within bandwidth. Only used in 'multitaper' mode. Ignored if data is an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object.

cwt_freqsarray_like

Array-like of frequencies of interest. Only used in 'cwt_morlet' mode. Only the frequencies within the range specified by fmin and fmax are used. Ignored if data is an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object.

cwt_n_cyclesfloat | array_like

Number of cycles. Fixed number or one per frequency. Only used in 'cwt_morlet' mode. Ignored if data is an mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR object.

gc_n_lagsint

Number of lags to use for the vector autoregressive model when computing Granger causality. Higher values increase computational cost, but reduce the degree of spectral smoothing in the results. Only used if method contains any of ['gc', 'gc_tr'].

ranktuple of array_like | None

Two array-likes with the rank to project the seed and target data to, respectively, using singular value decomposition. If None, the rank of the data is computed and projected to. Only used if method contains any of ['cacoh', 'mic', 'mim', 'gc', 'gc_tr'].

n_componentsint | None

Number of connectivity components to extract from the data. If an int, the number of components must be <= the minimum rank of the seeds and targets. E.g., if the seed channels had a rank of 5 and the target channels had a rank of 3, n_components must be <= 3. If None, the number of components equal to the minimum rank of the seeds and targets is extracted (see the rank parameter). Only used if method contains any of ['cacoh', 'mic'].

Added in version 0.8.

block_sizeint

How many connections to compute at once (higher numbers are faster but require more memory).

n_jobsint

How many samples to process in parallel.

verbosebool | str | int | None

If not None, override default verbose level (see mne.verbose() for more info). If used, it should be passed as a keyword-argument only.

Returns:

coninstance of SpectralConnectivity or SpectroTemporalConnectivity | list

Computed connectivity measure(s). An instance of SpectralConnectivity, SpectroTemporalConnectivity, or a list of instances corresponding to connectivity measures if several connectivity measures are specified. The shape of the connectivity result will be:

• (n_cons, n_freqs) for 'multitaper' or 'fourier' modes

• (n_cons, n_freqs, n_times) for 'cwt_morlet' mode

• (n_cons, n_comps, n_freqs[, n_times]) for valid multivariate methods if n_components > 1

• n_cons = n_signals ** 2 for methods with indices='all'

• n_cons = n_signals * (n_signals - 1) / 2 for methods with indices='lower' or indices='upper'

• n_cons = len(indices[0]) when indices is a tuple of array-likes

See also

mne_connectivity.spectral_connectivity_time

mne_connectivity.SpectralConnectivity

mne_connectivity.SpectroTemporalConnectivity

Notes

Please note that the interpretation of the measures in this function depends on the data and underlying assumptions and does not necessarily reflect a causal relationship between brain regions.

These measures are not to be interpreted over time. Each epoch passed into the dataset is interpreted as an independent sample of the same connectivity structure. Within each epoch, it is assumed that the spectral measure is stationary. The spectral measures implemented in this function are computed across epochs. Thus, spectral measures computed with only one epoch will result in errorful values and spectral measures computed with few Epochs will be unreliable. Please see spectral_connectivity_time() for time-resolved connectivity estimation.

The spectral densities can be estimated using a multitaper method with digital prolate spheroidal sequence (DPSS) windows, a discrete Fourier transform with Hanning windows, or a continuous wavelet transform using Morlet wavelets. The spectral estimation mode is specified using the mode parameter. Complex Welch, multitaper, or Morlet coefficients can also be passed in as data in the form of mne.time_frequency.EpochsSpectrum or mne.time_frequency.EpochsTFR objects.

If connectivity for all possible connections is desired, we can save time and memory by only computing and storing the lower- (or upper-) triangular part of the connectivity matrix. When the full set of connectivity values is needed, the missing values can be inferred based on what has been computed. See the notes section of the get_data() method in the connectivity containers for more information.

If one is only interested in the connectivity between some signals, the indices parameter can be specified as a tuple of array-likes. For example, to compute the connectivity between the signal with index 0 and the signals with indices 2, 3, and 4 (a total of 3 connections) one can use the following:

indices = (np.array([0, 0, 0]),  # seed indices
           np.array([2, 3, 4]))  # target indices

For multivariate methods, indices must be specified as a tuple, where the seed and target indices for each connection are nested array-likes. For example, to compute the connectivity between signals (0, 1) → (2, 3) and (0, 1) → (4, 5), indices should be specified as:

indices = (np.array([[0, 1], [0, 1]]),  # seeds
           np.array([[2, 3], [4, 5]]))  # targets

More information on working with multivariate indices and handling connections where the number of seeds and targets are not equal can be found in the Working with ragged indices for multivariate connectivity example.

Supported Connectivity Measures

The connectivity method(s) is specified using the method parameter. The following methods are supported (note: E[] denotes average over epochs). Multiple measures can be computed at once by using a list/tuple, e.g., ['coh', 'pli'] to compute coherence and PLI.

‘coh’ : Coherence given by:

         | E[Sxy] |
C = ---------------------
    sqrt(E[Sxx] * E[Syy])

‘cohy’ : Coherency given by:

           E[Sxy]
C = ---------------------
    sqrt(E[Sxx] * E[Syy])

‘imcoh’ : Imaginary coherence [1] given by:

          Im(E[Sxy])
C = ----------------------
    sqrt(E[Sxx] * E[Syy])

‘cacoh’ : Canonical Coherency (CaCoh) [2] given by:

\(\textrm{CaCoh}=\Large{\frac{\boldsymbol{a}^T\boldsymbol{D} (\Phi)\boldsymbol{b}}{\sqrt{\boldsymbol{a}^T\boldsymbol{a} \boldsymbol{b}^T\boldsymbol{b}}}}\)

where: \(\boldsymbol{D}(\Phi)\) is the cross-spectral density between seeds and targets transformed for a given phase angle \(\Phi\); and \(\boldsymbol{a}\) and \(\boldsymbol{b}\) are eigenvectors for the seeds and targets, such that \(\boldsymbol{a}^T\boldsymbol{D}(\Phi) \boldsymbol{b}\) maximises coherency between the seeds and targets. Taking the absolute value of the results gives maximised coherence.

‘mic’ : Maximised Imaginary part of Coherency (MIC) [3] given by:

\(\textrm{MIC}=\Large{\frac{\boldsymbol{\alpha}^T \boldsymbol{E \beta}}{\parallel\boldsymbol{\alpha}\parallel \parallel\boldsymbol{\beta}\parallel}}\)

where: \(\boldsymbol{E}\) is the imaginary part of the transformed cross-spectral density between seeds and targets; and \(\boldsymbol{\alpha}\) and \(\boldsymbol{\beta}\) are eigenvectors for the seeds and targets, such that \(\boldsymbol{\alpha}^T \boldsymbol{E \beta}\) maximises the imaginary part of coherency between the seeds and targets.

‘mim’ : Multivariate Interaction Measure (MIM) [3] given by:

\(\textrm{MIM}=tr(\boldsymbol{EE}^T)\)

where \(\boldsymbol{E}\) is the imaginary part of the transformed cross-spectral density between seeds and targets.

‘plv’ : Phase-Locking Value (PLV) [4] given by:

PLV = |E[Sxy/|Sxy|]|

‘ciplv’ : corrected imaginary PLV (ciPLV) [5] given by:

                 |E[Im(Sxy/|Sxy|)]|
ciPLV = ------------------------------------
         sqrt(1 - |E[real(Sxy/|Sxy|)]| ** 2)

‘ppc’ : Pairwise Phase Consistency (PPC), an unbiased estimator of squared PLV [6].

‘pli’ : Phase Lag Index (PLI) [7] given by:

PLI = |E[sign(Im(Sxy))]|

‘pli2_unbiased’ : Unbiased estimator of squared PLI [8].

‘dpli’ : Directed Phase Lag Index (DPLI) [9] given by (where H is the Heaviside function):

DPLI = E[H(Im(Sxy))]

‘wpli’ : Weighted Phase Lag Index (WPLI) [8] given by:

          |E[Im(Sxy)]|
WPLI = ------------------
          E[|Im(Sxy)|]

‘wpli2_debiased’ : Debiased estimator of squared WPLI [8].

‘gc’ : State-space Granger Causality (GC) [10] given by:

\(GC = ln\Large{(\frac{\lvert\boldsymbol{S}_{tt}\rvert}{\lvert \boldsymbol{S}_{tt}-\boldsymbol{H}_{ts}\boldsymbol{\Sigma}_{ss \lvert t}\boldsymbol{H}_{ts}^*\rvert}})\)

where: \(s\) and \(t\) represent the seeds and targets, respectively; \(\boldsymbol{H}\) is the spectral transfer function; \(\boldsymbol{\Sigma}\) is the residuals matrix of the autoregressive model; and \(\boldsymbol{S}\) is \(\boldsymbol{\Sigma}\) transformed by \(\boldsymbol{H}\).

‘gc_tr’ : State-space GC on time-reversed signals [10][11] given by the same equation as for 'gc', but where the autocovariance sequence from which the autoregressive model is produced is transposed to mimic the reversal of the original signal in time [12].

References

[1]

Guido Nolte, Ou Bai, Lewis Wheaton, Zoltan Mari, Sherry Vorbach, and Mark Hallett. Identifying true brain interaction from EEG data using the imaginary part of coherency. Clinical Neurophysiology, 115(10):2292–2307, 2004. doi:10.1016/j.clinph.2004.04.029.

[2]

Carmen Vidaurre, Guido Nolte, Ingmar E.J. de Vries, M. Gómez, Tjeerd W. Boonstra, K.-R. Müller, Arno Villringer, and Vadim V. Nikulin. Canonical maximization of coherence: a novel tool for investigation of neuronal interactions between two datasets. NeuroImage, 201:116009, 2019. doi:10.1016/j.neuroimage.2019.116009.

[3] (1,2)

Arne Ewald, Laura Marzetti, Filippo Zappasodi, Frank C. Meinecke, and Guido Nolte. Estimating true brain connectivity from EEG/MEG data invariant to linear and static transformations in sensor space. NeuroImage, 60(1):476–488, 2012. doi:10.1016/j.neuroimage.2011.11.084.

[4]

Jean-Philippe Lachaux, Eugenio Rodriguez, Jacques Martinerie, and Francisco J. Varela. Measuring phase synchrony in brain signals. Human Brain Mapping, 8(4):194–208, 1999. doi:10.1002/(SICI)1097-0193(1999)8:4<194::AID-HBM4>3.0.CO;2-C.

[5]

Ernesto Pereda Ricardo Bruña, Fernando Maestú. Phase locking value revisited: teaching new tricks to an old dog. Journal of Neural Engineering, 15(5):056011, 2018. doi:10.1088/1741-2552/aacfe4.

[6]

Martin Vinck, Marijn van Wingerden, Thilo Womelsdorf, Pascal Fries, and Cyriel M.A. Pennartz. The pairwise phase consistency: a bias-free measure of rhythmic neuronal synchronization. NeuroImage, 51(1):112–122, 2010. doi:10.1016/j.neuroimage.2010.01.073.

[7]

Cornelis J. Stam, Guido Nolte, and Andreas Daffertshofer. Phase lag index: assessment of functional connectivity from multi channel EEG and MEG with diminished bias from common sources. Human Brain Mapping, 28(11):1178–1193, 2007. doi:10.1002/hbm.20346.

[8] (1,2,3)

Martin Vinck, Robert Oostenveld, Marijn van Wingerden, Franscesco Battaglia, and Cyriel M.A. Pennartz. An improved index of phase-synchronization for electrophysiological data in the presence of volume-conduction, noise and sample-size bias. NeuroImage, 55(4):1548–1565, 2011. doi:10.1016/j.neuroimage.2011.01.055.

[9]

C. J. Stam and E. C. W. van Straaten. Go with the flow: use of a directed phase lag index (dpli) to characterize patterns of phase relations in a large-scale model of brain dynamics. NeuroImage, 62(3):1415–1428, Sep 2012. doi:10.1016/j.neuroimage.2012.05.050.

[10] (1,2)

Lionel Barnett and Anil K. Seth. Granger causality for state-space models. Physical Review E, 91(4):040101, 2015. doi:10.1103/PhysRevE.91.040101.

[11]

Irene Winkler, Danny Panknin, Daniel Bartz, Klaus-Robert Müller, and Stefan Haufe. Validity of time reversal for testing granger causality. IEEE Transactions on Signal Processing, 64(11):2746–2760, 2016. doi:10.1109/TSP.2016.2531628.

[12]

Stefan Haufe, Vadim V Nikulin, and Guido Nolte. Alleviating the influence of weak data asymmetries on granger-causal analyses. In Latent Variable Analysis and Signal Separation: 10th International Conference, LVA/ICA 2012, Tel Aviv, Israel, March 12-15, 2012. Proceedings 10, 25–33. Springer, 2012. doi:10.1007/978-3-642-28551-6_4.

Examples using mne_connectivity.spectral_connectivity_epochs

Compute multivariate coherency/coherence

Compute multivariate coherency/coherence

Comparison of coherency-based methods

Comparison of coherency-based methods

Comparing spectral connectivity computed over time or over trials

Comparing spectral connectivity computed over time or over trials

Using the connectivity classes

Using the connectivity classes

Compute seed-based time-frequency connectivity in sensor space

Compute seed-based time-frequency connectivity in sensor space

Multivariate decomposition for efficient connectivity analysis

Multivariate decomposition for efficient connectivity analysis

Comparing PLI, wPLI, and dPLI

Comparing PLI, wPLI, and dPLI

Compute directionality of connectivity with multivariate Granger causality

Compute directionality of connectivity with multivariate Granger causality

Working with ragged indices for multivariate connectivity

Working with ragged indices for multivariate connectivity

Compute multivariate measures of the imaginary part of coherency

Compute multivariate measures of the imaginary part of coherency

Compute mixed source space connectivity and visualize it using a circular graph

Compute mixed source space connectivity and visualize it using a circular graph

Compute coherence in source space using a MNE inverse solution

Compute coherence in source space using a MNE inverse solution

Compute full spectrum source space connectivity between labels

Compute full spectrum source space connectivity between labels

Compute source space connectivity and visualize it using a circular graph

Compute source space connectivity and visualize it using a circular graph

Compute all-to-all connectivity in sensor space

Compute all-to-all connectivity in sensor space

Determine the significance of connectivity estimates against baseline connectivity

Determine the significance of connectivity estimates against baseline connectivity