All-to-all connectivity in sensor space in 2D instead of 3D

Support & Discussions
,
1
  • MNE version: 1.6.1
  • operating system: Windows 11

Hello everyone,

I need different all-to-all connectivity plots in sensor space, which should be in 2D and can be placed into subplots. I know about the function mne_connectivity.viz.plot_sensors_connectivity, but it is for 3D visualization. Since sensor positions are important for me, the alternative mne_connectivity.viz.plot_connectivity_circle does apply here either.
I found a similar topic, but the solution from this thread does not apply, since it links back to the 3D figure…

Is it somehow possible to get result of mne_connectivity.viz.plot_sensors_connectivity in 2D version and give axes argument so that the output can be used as a subplot?

Thanks!

2

Hi @datalw,

Is it somehow possible to get result of mne_connectivity.viz.plot_sensors_connectivity in 2D version and give axes argument so that the output can be used as a subplot?

I am not aware of a way to do so this with a 3D figure. You could try to save the 3D figure as a 2D image, but without properly projecting it to a 2D surface there will be many overlapping sensor locations, especially around the edges.


However, if you have a channel montage for your data, you could use the 2D figure returned from mne.viz.plot_montage() and then plot the connectivity between sensors on top of this, e.g.:

image
image1130×1020 86 KB

I generated this with the code below. You can just substitute in whatever all-to-all connectivity data you have:

import os.path as op
import mne
import numpy as np
from mne.datasets import sample
from matplotlib import pyplot as plt
from mne_connectivity import spectral_connectivity_epochs
from matplotlib import colors


# Read data
data_path = sample.data_path()
raw_fname = op.join(data_path, "MEG", "sample", "sample_audvis_filt-0-40_raw.fif")
event_fname = op.join(data_path, "MEG", "sample", "sample_audvis_filt-0-40_raw-eve.fif")
raw = mne.io.read_raw_fif(raw_fname).pick("eeg")
events = mne.read_events(event_fname)

# Create epochs
event_id, tmin, tmax = 3, -0.2, 1.5
epochs = mne.Epochs(raw, events, event_id, tmin, tmax, baseline=(None, 0))

# Compute connectivity for the frequency band containing the evoked response
fmin, fmax = 4.0, 9.0
tmin = 0.0
con = spectral_connectivity_epochs(
    data=epochs, method="pli", fmin=fmin, fmax=fmax, faverage=True, tmin=tmin, n_jobs=1
)

########################################################################################
# Define connections to plot
min_distance = 0.05
n_con = 10

# Get the sensor locations
fig, axs = plt.subplots(1, 1, figsize=(10, 10))
fig = mne.viz.plot_montage(raw.get_montage(), axes=axs, show_names=False, show=False)
sens_loc = fig.axes[0].collections[0].get_offsets().data

# Get the strongest n_con connections
con_vals = con.get_data("dense")[:, :, 0]
threshold = np.sort(con_vals, axis=None)[-n_con]
ii, jj = np.where(con_vals >= threshold)

# Remove close connections
con_nodes = list()
con_val = list()
for i, j in zip(ii, jj):
    if np.linalg.norm(sens_loc[i] - sens_loc[j]) > min_distance:
        con_nodes.append((i, j))
        con_val.append(con_vals[i, j])
con_val = np.array(con_val)

# Plot the connections
vmin = con_val.min()
vmax = con_val.max()
cmap = plt.get_cmap("viridis")
norm = colors.Normalize(vmin=vmin, vmax=vmax)
colours = cmap(norm(con_val))
for node_idx, (i, j) in enumerate(con_nodes):
    axs.plot(
        (sens_loc[i, 0], sens_loc[j, 0]),
        (sens_loc[i, 1], sens_loc[j, 1]),
        color=colours[node_idx],
    )
fig.colorbar(
    plt.cm.ScalarMappable(norm=norm, cmap=cmap), ax=axs, shrink=0.5, label="PLI (A.U.)"
)
3

Thanks a lot for the reply and the solution! This is exactly what I need.

1 Like
4

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.