Guillaume Dumas

Even if this measure does not exist, we think that it should be possible to adapt wPLI to cross-frequency like in nm-PLV.

The current implementation of connectivity calculations is fast but require a lot of memory. We can provide a choice to users as a parameter

Cross-Wavelet Transform (CWT)

16

[Python implementation](https://pycwt.readthedocs.io/en/latest/)

Using [STRF](https://mne.tools/dev/auto_tutorials/machine-learning/30_strf.html#sphx-glr-auto-tutorials-machine-learning-30-strf-py) and [MVPA](https://mne.tools/dev/auto_tutorials/machine-learning/50_decoding.html).

Provide the ability to load physiological recording e.g. ECG using standard i/o methods. Note: the first trial will be with E4 Empatica files.

Group-hyperscanning

1

Extend from dual-recording to group-recording.

Sources visualization

3

Integrate different parcelations to plot inter-brain connectivity. Example: 

Mutual Information? Transfer Entropy? Granger Emergence? Information Integration?

Pool sensors by ROI (e.g., lobes) Heatmap visualization (with statistical links indicated with a little star) => also cool for sensors!

Implementation in [R](https://www.frontiersin.org/articles/10.3389/fpsyg.2014.00510/full) and [Python](https://pypi.org/project/PyRQA/).