MEG file format "KRISS" mentions *_digitizer.txt file (and suffix) which is not defined in the spec

[sappelhoff]

See title.

https://bids-specification.readthedocs.io/en/latest/appendices/meg-file-formats.html#kriss

Each experimental run on the KRISS system produces a file with extension .kdf. Additional files can be available in the same directory: the digitized positions of the head points (_digitizer.txt), the position of the center of the MEG coils (.chn) and the event markers (.trg).

(emphasis by me)

How I read it, _digitizer.txt should be _headshape.<ext>, which is defined in the spec, and which fulfills exactly that use case (digitizing head points, and optionally electrodes).

Unfortunately, digitizer is defined in the bids-validator, see:

• https://github.com/bids-standard/bids-validator/blob/d6d2f9d7ac00e6b21f7fca6fcad3d9a5878868e0/bids-validator/bids_validator/rules/file_level_rules.json#L476
• https://github.com/bids-standard/bids-validator/blob/d6d2f9d7ac00e6b21f7fca6fcad3d9a5878868e0/bids-validator/tests/type.spec.js#L270

Was there a reason to name this _digitizer.txt for the KRISS MEG data format, and _headshape.<ext> for all other MEG data formats? Or was this simply an oversight? Either way, we need to clarify this and either:

1. properly introduce _digitizer in the spec
2. completely remove _digitizer from (i) the appendix in which it is mentioned, (ii) the bids-validator

Related to:

• file formats for digitization points in BIDS-MEG specification #20
• electrodes.tsv for simultaneous MEG/EEG data #1550

[sappelhoff]

Note that a quick search of our history yields that I was the one who introduced _digitizer.txt into both spec and validator in:

• https://github.com/bids-standard/bids-validator/pull/789
• [MISC] updating MEG doc links, manufacturer names, and adding a missing MEG example #248

My intention was to sync GitHub spec and old Google docs (see section 16.5 in the old gdoc of the spec) shortly after the move to GitHub. It's of specific note that I repeatedly tried to find someone familiar with the KRISS system, but failed.

This lets me tend towards my solution "2" here:

Was there a reason to name this _digitizer.txt for the KRISS MEG data format, and _headshape. for all other MEG data formats? Or was this simply an oversight? Either way, we need to clarify this and either:

1. properly introduce _digitizer in the spec
2. completely remove _digitizer from (i) the appendix in which it is mentioned, (ii) the bids-validator

[sappelhoff]

cc @robertoostenveld your opinion would be valuable here

[robertoostenveld]

Disclaimer: I don't know details about the KRISS system and file format.

I don't recall having discussed in the development of the MEG specification that the digitized head shape information for the KRISS format should be stored in the _digitizer.txt file. To me it would also make more sense if the file were called _headshape.txt.