Heudiconv error: concatenated volumes while converting DICOM to BIDS

, ,
1

Hi everyone,

I’m running heudiconv on singularity to convert DICOMs to BIDS. All MPRAGE and rest scans seems to have converted correctly, but I’m having error with only the topup scans.
Here in dicominfo.tsv file, dimension 3 should be 75 and dimension 4 should be 3 in topup scans, but seems to be concatenating volumes.

Screen Shot 2021-04-23 at 10.39.30 AM
Screen Shot 2021-04-23 at 10.39.30 AM1850×350 144 KB

Here’s the heuristic file:

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):
“”“Heuristic evaluator for determining which runs belong where
allowed template fields - follow python string module:
item: index within category
subject: participant id
seqitem: run number during scanning
subindex: sub index within group
“””
t1w = create_key(‘sub-{subject}/{session}/anat/sub-{subject}_{session}_run-00{item:01d}T1w’)
func_rest = create_key('sub-{subject}/{session}/func/sub-{subject}{session}_task-rest_run-00{item:01d}bold’)
func_2Hz = create_key('sub-{subject}/{session}/func/sub-{subject}{session}_task-stim2Hz_run-00{item:01d}bold’)
func_10Hz = create_key('sub-{subject}/{session}/func/sub-{subject}{session}_task-stim10Hz_run-00{item:01d}bold’)
func_25Hz = create_key('sub-{subject}/{session}/func/sub-{subject}{session}_task-stim25Hz_run-00{item:01d}bold’)
func_100Hz = create_key('sub-{subject}/{session}/func/sub-{subject}{session}_task-stim100Hz_run-00{item:01d}bold’)
topup_ap = create_key('sub-{subject}/{session}/fmap/sub-{subject}{session}dir-AP_epi’)
topup_pa = create_key('sub-{subject}/{session}/fmap/sub-{subject}{session}_dir-PA_epi’)

info = {t1w: [], func_rest: [], func_2Hz: [], func_10Hz: [], func_25Hz: [], func_100Hz: [], topup_ap: [], topup_pa: []}

for idx, s in enumerate(seqinfo):
    if (s.dim1 == 256) and (s.dim2 == 256) and ('MEMPRAGE_tfl_mgh_multiecho_1mm_iso' in s.protocol_name):
       info[t1w].append(s.series_id)
    if (s.dim1 == 98) and (s.dim4 == 400) and ('REST_ep2d_bold_smsX5_04_20150527' in s.protocol_name):
       info[func_rest].append(s.series_id)
    if (s.dim1 == 98) and (s.dim4 == 400) and ('2Hz_ep2d_bold_smsX5_04_20150527' in s.protocol_name):
       info[func_2Hz].append(s.series_id)
    if (s.dim1 == 98) and (s.dim4 == 400) and ('10Hz_ep2d_bold_smsX5_04_20150527' in s.protocol_name):
       info[func_10Hz].append(s.series_id)
    if (s.dim1 == 98) and (s.dim4 == 400) and ('25Hz_ep2d_bold_smsX5_04_20150527' in s.protocol_name):
       info[func_25Hz].append(s.series_id)
    if (s.dim1 == 98) and (s.dim4 == 400) and ('100Hz_ep2d_bold_smsX5_04_20150527' in s.protocol_name):
       info[func_100Hz].append(s.series_id)
    if (s.dim1 == 98) and (s.dim2 == 98) and ('topup_ap' in s.protocol_name):
       info[topup_ap].append(s.series_id)
    if (s.dim1 == 98) and (s.dim2 == 98) and ('topup_pa' in s.protocol_name):
       info[topup_pa].append(s.series_id)
return info

Could this error be based on the initial DICOM files?
Do you have any idea what is happening?

Thanks a lot!
Kyungsun

2

which scanner is it? If Philips please see if started to split 4d philips into individual files due to "Slices not stacked: trigger time varies" · Issue #395 · rordenlab/dcm2niix · GitHub applies

3

Thanks, but it was Siemens 3T skyra scanner.

4

I wonder if this reflects the fact that you are using an old version of dcm2niix. You can check this by looking at the BIDS format JSON files. If the Version is older than the current stable release (v1.0.20210317) I would suggest upgrading dcm2niix.

	"ConversionSoftware": "dcm2niix",
	"ConversionSoftwareVersion": "v1.0.20210317"

If this does not fix your issue, I would try running the latest version of dcm2niix directly on your images to remove Heudiconv from the equation:

dcm2niix /path/to/DICOMs

If this concatenates your field map volumes then you should generate a new issue on the dcm2niix Github page.

As an aside, Siemens field map DICOMs are unusual in that they reuse the image instance numbers (0020,0013). This can cause havoc with tools that assume instance numbers are unique. For example, a PACS system will assume that these are replicated scans and overwrite some slices from echo 1 with slices from echo 2. To avoid issues, users of Siemens systems VA-VE should save field maps as mosaic format (though users of Siemens XA systems should never export data as mosaics).

2 Likes
5

Hi Chris,

Thanks a lot for your help. Updating the dcm2niix version helped me fix the problem. Although dicominfo.tsv file still shows the same concatenated numbers in dim3, nifti files carry correct information!