Fiducials not preserved by bidsification

The bidsification of EEG recordings using PyLossless does not preserve the information about the fiducials. It is not clear if the issue is in PyLossless per se or mne-bids.

Example:

import pylossless as ll

path = ('/Users/christian/Library/CloudStorage/OneDrive-UniversityofSouthCarolina/Data/'
        'ABC/PASSAGE_EEG/ABC1020.SM.passage.vhdr')

def import_fct(path_in):
    return mne.io.read_raw_brainvision(path_in, preload=True), None, None

import_args = [{"path_in": path}]
bids_path_args = [{'subject': "1020", 'session': '01', "task": "narative"}]

raw1 = mne.io.read_raw_brainvision(path, preload=True)

bids_paths = ll.bids.convert_dataset_to_bids(import_fct, import_args, bids_path_args, 
                                                 bids_root=path_eegp / "bids_dataset", overwrite=True)

raw2 = mne_bids.read_raw_bids(bids_paths[0])

Then

print(raw1.info["dig"][:4])

gives

[<DigPoint |        LPA : (-94.3, 0.0, 0.0) mm      : head frame>,
 <DigPoint |     Nasion : (0.0, 94.3, 0.0) mm       : head frame>,
 <DigPoint |        RPA : (94.3, 0.0, 0.0) mm       : head frame>,
 <DigPoint |     EEG #1 : (-29.4, 90.4, 0.0) mm     : head frame>]

whereas

print(raw2.info["dig"][:4])

gives

[<DigPoint |        LPA : (nan, nan, nan) mm        : head frame>,
 <DigPoint |     Nasion : (nan, nan, nan) mm        : head frame>,
 <DigPoint |        RPA : (nan, nan, nan) mm        : head frame>,
 <DigPoint |     EEG #1 : (-29.4, 90.4, 0.0) mm     : head frame>]

To check and to file the issue with mne-bids if the error is on their side.

lina-usc / pylossless

Fiducials not preserved by bidsification #20