[Neuroimaging] Fwd: [bids-discussion] Make DelayTime compulsory for all sparse acquisitions

[JB Poline]

Hey

Forwarding this or the relevance with nibabel and others - I remember
a similar-ish discussion

Cheers
JB


---------- Forwarded message ----------
From: Xiangrui Li <li.2327 at osu.edu>
Date: 25 July 2017 at 22:35
Subject: Re: [bids-discussion] Make DelayTime compulsory for all
sparse acquisitions
To: "bids-discussion at googlegroups.com" <bids-discussion at googlegroups.com>


Talking about sparse acquisition, I would like to mention my recent
experience related to multiband sparse-like sequence. If this has not
been brought up before, I would like to propose an optional field to
encode the timing of NIfTI's 4th dimension.

We use the CMRR multiband sequence, which allows to set several TRs as
"quiet" period without image acquisition, so one can play sound during
this period.

For example, we use TR of 1 second, and acquire 10 volumes
continuously, then have 4-TR quiet period. This results in 14-second
cycles with only 10-second of image data in each cycle. The dicom
files have discontinuity in InstanceNumber, like 1 to 10, then jump to
15 to 24 etc.

Most NIfTI converters quietly convert these files. The result NIfTI is
actually not strictly NIfTI compliant, since its 4th dim (timing) has
variable intervals. My matlab-based converter (dicm2nii.m) would have
failed to convert, complaining missing files. I just updated it,
allowing missing files with regular spacing (missing 4 in every 14
files in above example), while it gives a warning for the
InstanceNumber discontinuity, and stores the InstanceNumber in a field
in NIfTI extension.

I was thinking to pad zeros for those missing volumes to make it NIfTI
compliant, but it seems dumb to increase the file size. After asking
opinion from some auditory people, I think it is better to make it
consistent with other NIfTI converters, so everyone is violating the
NIfTI rule :)

The field I am proposing can be called something like "TimeIndices".
It has the same length as the 4th dim of NIfTI image. For the above
example, it can be those InstanceNumber minus 1 (0-based), or we call
it a different name and store the time in seconds: (InstanceNumber-1)
* TR.

Any thoughts, comments? Thanks.
_____________

Xiangrui Li, Ph.D.

Center for Cognitive and Behavioral Brain Imaging

The Ohio State University

Phone: 614-292-1847

From: Satrajit Ghosh
Date: 2017-07-25 18:15
To: bids-discussion at googlegroups.com
Subject: [bids-discussion] Make DelayTime compulsory for all sparse acquisitions
hi folks,

chris and i were chatting about a dataset with a sparse acquisition
protocol: collect n-slices wait m seconds, collect again. on siemens
scanners this is implemented using TR and delay, with TA (time of
acquisition) = TR - delay.

the current bids specification says if slice timing is provided,
DelayTime is not required. i would like to propose that DelayTime is
provided for all sparse acquisitions independent of slice timing.

here is the reasoning. as it stands i have to write code that kind of
looks like this:

if multiband_sparse:
     good_luck_figuring_out_slice_timing is correct or not.
     (on our scanner siemens generated incorrect slice timing numbers).
else:
     TA = max(slice_times) + np.average(np.diff(sort(slice_times)))

with DelayTime, it would always be: RepetitionTime - DelayTime.

let any discussions ensue.

cheers,

satra

--
You received this message because you are subscribed to the Google
Groups "bids-discussion" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to bids-discussion+unsubscribe at googlegroups.com.
To post to this group, send email to bids-discussion at googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/bids-discussion/CA%2BA4wOnk%3DJvddeWSmZL1puexaCMA-Y_YyQbKjYTtPJ1kS_YcZg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google
Groups "bids-discussion" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to bids-discussion+unsubscribe at googlegroups.com.
To post to this group, send email to bids-discussion at googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/bids-discussion/2017072521355953127853%40osu.edu.

For more options, visit https://groups.google.com/d/optout.