Heudiconv adding unspecified suffix

,
1

Hey Neurostars,
I’ve used heudiconv before with little problem but I am running into one odd problem for one subject. There are 6 acquisitions of BOLD data.
All of the acquisitions seem to be correctly identified when I look at the .auto.txt file in the .heudiconv directory

image
image1191×387 123 KB

However, it adds an incorrect suffix to the mb3me4 acquisition and fails to convert the mb6me4 altogether.
Screenshot from 2022-02-04 10-47-48
Screenshot from 2022-02-04 10-47-481159×420 172 KB

It only seems to fail on the acquisitions that are both multiband and multi-echo however this isn’t a problem with another subject.

Here’s what the current heuristics file and dicominfo.tsv for the subject looks like. Any help would be appreciated

image
image802×427 119 KB

import os

def create_key(template, outtype=('nii.gz',), annotation_classes=None):
    if template is None or not template:
        raise ValueError('Template must be a valid format string')
    return template, outtype, annotation_classes

def infotodict(seqinfo):
    t1w = create_key('sub-{subject}/anat/sub-{subject}_T1w')
    mag = create_key('sub-{subject}/fmap/sub-{subject}_run-{item:01d}_magnitude')
    phase = create_key('sub-{subject}/fmap/sub-{subject}_run-{item:01d}_phasediff')


    #me1
    mb1me1 = create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb1me1_bold')
    mb3me1 = create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb3me1_bold')
    mb3me1_sbref = create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb3me1_sbref')
    mb6me1 =create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb6me1_bold')
    mb6me1_sbref = create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb6me1_sbref')

    #me4
    mb1me4 =create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb1me4_bold')
    mb3me4 =create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb3me4_bold')
    mb3me4_sbref =create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb3me4_sbref')
    mb6me4 =  create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb6me4_bold')
    mb6me4_sbref = create_key('sub-{subject}/func/sub-{subject}_task-sharedreward_acq-mb6me4_sbref')



        # mag: [],
        # phase: [],

    info = {t1w: [],mag: [],phase: [],

            mb1me1: [],
            mb3me1: [], mb3me1_sbref: [],
            mb6me1: [], mb6me1_sbref: [],

            mb1me4: [],
            mb3me4: [], mb3me4_sbref: [],
            mb6me4: [], mb6me4_sbref: [],}

    list_of_ids = [s.series_id for s in seqinfo]
    for s in seqinfo:
        if ('T1w-anat_mpg_07sag_iso' in s.protocol_name) and ('NORM' in s.image_type):
            info[t1w] = [s.series_id]
        if ('gre_field' in s.protocol_name) and ('NORM' in s.image_type):
            info[mag].append(s.series_id)
        if ('gre_field' in s.protocol_name) and ('P' in s.image_type):
            info[phase].append(s.series_id)

        # no multi-echo
        if (s.dim4 >= 150) and ('MB1_' in s.protocol_name) and ('_ME1' in s.protocol_name):
            info[mb1me1].append(s.series_id)
            idx = list_of_ids.index(s.series_id)
        elif (s.dim4 >= 150) and ('MB3_' in s.protocol_name) and ('_ME1' in s.protocol_name):
            info[mb3me1].append(s.series_id)
            idx = list_of_ids.index(s.series_id)
            info[mb3me1_sbref].append(list_of_ids[idx -1])
        elif (s.dim4 >= 150) and ('MB6_' in s.protocol_name) and ('_ME1' in s.protocol_name):
            info[mb6me1].append(s.series_id)
            idx = list_of_ids.index(s.series_id)
            info[mb6me1_sbref].append(list_of_ids[idx -1])

        # multi-echo standard
        if (s.dim4 >= 150) and ('MB1_' in s.protocol_name) and ('_ME4' in s.protocol_name):
            info[mb1me4].append(s.series_id)
            idx = list_of_ids.index(s.series_id)
        elif (s.dim4 >= 150) and ('MB3_' in s.protocol_name) and ('_ME4' in s.protocol_name) :
            info[mb3me4].append(s.series_id)
            idx = list_of_ids.index(s.series_id)
            info[mb3me4_sbref].append(list_of_ids[idx -1])
        elif (s.dim4 >= 150) and ('MB6_' in s.protocol_name) and ('_ME4' in s.protocol_name):
            info[mb6me4].append(s.series_id)
            idx = list_of_ids.index(s.series_id)
            info[mb6me4_sbref].append(list_of_ids[idx -1])


    return info
2

I also realized I get this error right after it finishes converting the mb3me4 acquisition.

image
image773×344 62 KB

but this doesn’t appear for the single band with 4 echos, or for my other participant.
Is there something to look for in the dicominfo or header?

3

Your error appears to be related to this issue: https://github.com/nipy/heudiconv/issues/460. I’m not sure why it’s only appearing for one scan from one subject, but we do have an open PR to fix it: https://github.com/nipy/heudiconv/pull/461.

EDIT: Just to clarify the weird suffixes, the _e* suffixes are, I believe, dcm2niix’s default naming for multi-echo data. Heudiconv would normally automatically rename those files based on your heuristic file, but that step is failing (hence the error you brought up), so the original converted files are still around.

1 Like
4

Hey Taylor,
Thanks that’s a big help to have those links. I’m not entirely sure but maybe there was a corruption in the data download it looks like the rest of my images have a “echo number” not just an “echo time:” which is where it looks like the set vs list issue comes in.
I kinda made the assumption about the “_e” myself but wasn’t sure if the “heudiconv491” aspect had some meaning where I shouldn’t just rename them. If it’s just this subject I can probably do by hand or just quickly edit the update_multiecho_name function for my container.

1 Like