[Mne_analysis] Filtering and Epoching MNE-HCP

[Winter, Matthew]

        External Email - Use Caution        

Hi all,
I am still working with the MNE-HCP tools, and have two pipelines that I seem to have hit some walls for both of them. The overall goal is to commute coherence connectivity in source space.
In the first pipeline I am simply trying to use:
raw1=hcp.read_epochs(subject, 'task_working_memory', onset='stim', run_index=0, hcp_path=hcp_path, return_fixations_motor=False)
raw2=hcp.read_epochs(subject, 'task_working_memory', onset='stim', run_index=1, hcp_path=hcp_path, return_fixations_motor=False)
But I get:
line 611, in _read_epochs
    epochs.times = times
AttributeError: can't set attribute

Since I cant seem to get that working I have also been running the raw instance through my lab’s standard preprocessing pipeline, but there are some things in the tutorial that I am not sure if I have to include that are not part of our standard (eeg) pipeline. When I plot the sensors it appears 90 degrees off, but hcp.preprocessing.map_ch_coords_to_mne(raw) seems to fix that. However, I am not sure if I need to apply hcp.preprocessing.apply_ref_correction(raw) because I am getting some unexpected things by doing it.  when I do:
ecg_evoked = mne.preprocessing.create_ecg_epochs(raw,ch_name='ECG-').average()
ecg_evoked.apply_baseline((None, None))
ecg_evoked.plot_joint()
projs, events = mne.preprocessing.compute_proj_ecg(raw, ch_name = 'ECG-', n_grad=1, n_mag=1, n_eeg=1, reject=None)
mne.viz.plot_projs_topomap(projs, info=raw.info)
raw.add_proj(projs)
without the reference correction it smears the cardiac artifact (not sure if I should be using ECG- or ECG+ or both) as a negative going component across the back left and front right of the head. With the reference corrections applied it gives me a very standard looking cardiac component but on the other side of the head near the left eye. I am use to EEG data but something odd I noticed is that there appears to be regular cardiac activity not just on the back left sensors, but also very strongly on both sides of the eyes, too regular for eyeblinks and matched up in time with the sensors that to typically pick up the cardiac artifact.

Does anyone out there know more of what is going on, I am at a loss and not sure what to do.  Thank you for any advice!

Best,
Matt
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20201001/4e2d1c78/attachment.html