Using PhysIO with HCP Aging/Development data

[ALXXLARG]

Hello,

Greetings. I am trying to use PhysIO to regress out physiological parameters of HCP development data. Unlike physio log files in .txt format from HCP 1200 data, the physio log files in HCP Development database were provided in .csv format and there are five columns which are labeled as 'TRIG', 'ACQ','EXT','RESP' and 'PULS'. I know that the three columns of physio log files of HCP 1200 data were related to trigger pulse, respiratory and pulse oximeter info. Does it mean that I could use three out of five columns of .csv files and save it as txt files and do the same thing? If that is true, which three columns should I choose, but if that is wrong, then I am wondering if there is a way to use .csv files of HCP Development?

Thank you very much for your help! LX

[mrikasper]

Dear LX,

Thank you for your interest in the PhysIO Toolbox. It would be great to update our read-in functions in PhysIO then in order to support this new HCP format natively. Would you be willing to share one of these example logfiles?

With respect to your question, in the original HCP release, the column order was ACQ_ON RESP, PULS. I would have to see an example to comment on whether TRIG or ACQ corresponds to ACQ_ON... It's a column of 1s and 0s, with 1 whenever scan acquisition is on (i.e., a volume is acquired).

Also, if you happen to have the original SIEMENS log files for the physiology, I usually recommend to use those, because it's not clear to me what happened during the conversion into the HCP .csv format. Do they provide any documentation or code on that?

I hope that helps.

All the best, Lars

[ALXXLARG]

Dear Lars,

Greetings. I might not be able to send the files due to the policies of NIMH Data Archive. I will send an email to HCP and see if I could just share one physio csv file. But let me illustrate some features of those two columns of one file, both columns, 'TRIG' and 'ACQ', have 1s and 0s, but only 478 out of total 160422 rows of the column 'TRIG' have 1 while 109800 of 160422 rows of the column 'ACQ' have 1.

It seems that they do not include original SIEMENS log files in HCP Development. I will keep searching and see if there are other documentations or codes on that.

Thank you for your help!

Best, LX

[mrikasper]

Dear LX,

it seems that one of them could be volume onset (are there 478 volumes in the run?) and the other ACQ_ON (always on while scanning a volume...), but it's hard to say without seeing the data or some documentation of it.

Let me know what they say about sharing the file (you can also do it privately, my e-mail is in the PhysIO README) - if it's not possible, is there a straightforward way to sign up for data access for me (only for the one or a few phys log files?)

All the best, Lars

[ALXXLARG]

Dear Lars,

Thank you so much for your reply. I will try to contact them and get the reply from HCP ASAP. But you remind me that TRIG column may be related to volume onset since the nscans of fmri bold imaging is also 478 (TR/TE = 800ms/37ms), I did not connect these numbers together yesterday.

Here is the link that includes the instructions (https://www.humanconnectome.org/storage/app/media/documentation/LS2.0/LS_Release_2.0_Access_Instructions.pdf) about how to access the data in the case that I don't have permission to share since one data is corresponding with one certification.

Thanks again!

Best, LX