BIDS validation issues: Incorrect slice timing values?

[adswa]

I am exploring whether the resulting datasets are BIDS-compliant enough to run fMRIprep on them. I will report on all problems I encounter as issues.

Invalid slice timing

Some files seem to have invalid slice timing:

I have checked it in two subjects, and they have this issue for different modalities. Subject 10027** has it for resting state fMRI:

2: [ERR] "SliceTiming" value/s contains invalid value as it is greater than RepetitionTime.  SliceTiming values should be in seconds not milliseconds (common mistake). (code: 66 - SLICETIMING_VALUES_GREATOR_THAN_REPETITION_TIME)
./sub-10027**/ses-3/func/sub-10027**_ses-3_task-rest_bold.nii.gz
Evidence: 0.805,0.895,0.9825,1.0725,1.1625,1.25,1.34,1.43,1.52,1.6075,1.6975,1.7875,1.8775,1.965,2.055,2.145,2.235,2.3225,2.4125,2.5025,2.5925,2.68,2.77,2.86,2.95,3.0375,3.1275,3.2175,3.3075,3.395,3.485,3.575,3.665,3.7525,3.8425,3.9325,4.0225,4.11,4.2,4.29,4.38,4.4675,4.5575,4.6475,4.7375,4.825,4.915,5.005,5.095,5.1825,5.2725,5.3625,5.4525,5.54,5.63

subject 10025** has is for task-hariri:

1: [ERR] "SliceTiming" value/s contains invalid value as it is greater than RepetitionTime.  SliceTiming values should be in seconds not milliseconds (common mistake). (code: 66 - SLICETIMING_VALUES_GREATOR_THAN_REPETITION_TIME)
./sub-10025**/ses-2/func/sub-10025**_ses-2_task-hariri_bold.nii.gz
Evidence: 0.805,0.8925,0.9825,1.0725,1.1625,1.25,1.34,1.43,1.52,1.6075,1.6975,1.7875,1.8775,1.965,2.055,2.145,2.235,2.3225,2.4125,2.5025,2.59,2.68,2.77,2.86,2.9475,3.0375,3.1275,3.2175,3.305,3.395,3.485,3.575,3.6625,3.7525,3.8425,3.9325,4.02,4.11,4.2,4.29,4.3775,4.4675,4.5575,4.6475,4.735,4.825,4.915,5.005,5.0925,5.1825,5.2725,5.3625,5.45,5.54,5.63

Here is the corresponding faulty (?) json sidecar for subject 10025**'s hariri task, and how it looks in subject 10027**:

cat fmriprep/sub-10025**/ses-2/func/sub-10025**_ses-2_task-hariri_bold.json | jq                                                                                       1 !
{
  "Manufacturer": "Siemens",
  "ManufacturersModelName": "Skyra",
  "ImageType": [
    "ORIGINAL",
    "PRIMARY",
    "M",
    "MB",
    "ND",
    "MOSAI"
  ],
  "AcquisitionTime": 90735.7425,
  "AcquisitionDate": 20170512,
  "MagneticFieldStrength": 3,
  "FlipAngle": 51,
  "EchoTime": 0.0424,
  "RepetitionTime": 0.735,
  "EffectiveEchoSpacing": 0.000639989,
  "SliceTiming": [
    0,
    0.09,
    0.1775,
    0.2675,
    0.3575,
    0.4475,
    0.535,
    0.625,
    0.715,
    0.805,
    0.8925,
    0.9825,
    1.0725,
    1.1625,
    1.25,
    1.34,
    1.43,
    1.52,
    1.6075,
    1.6975,
    1.7875,
    1.8775,
    1.965,
    2.055,
    2.145,
    2.235,
    2.3225,
    2.4125,
    2.5025,
    2.59,
    2.68,
    2.77,
    2.86,
    2.9475,
    3.0375,
    3.1275,
    3.2175,
    3.305,
    3.395,
    3.485,
    3.575,
    3.6625,
    3.7525,
    3.8425,
    3.9325,
    4.02,
    4.11,
    4.2,
    4.29,
    4.3775,
    4.4675,
    4.5575,
    4.6475,
    4.735,
    4.825,
    4.915,
    5.005,
    5.0925,
    5.1825,
    5.2725,
    5.3625,
    5.45,
    5.54,
    5.63
  ],
  "PhaseEncodingDirection": "j-"
}

Here is how is looks like for subject 10027** (where this task does not through an error):

cat fmriprep/sub-10027**/ses-2/func/sub-10027**_ses-2_task-hariri_bold.json | jq      
{
  "Manufacturer": "Siemens",
  "ManufacturersModelName": "Skyra",
  "ImageType": [
    "ORIGINAL",
    "PRIMARY",
    "M",
    "MB",
    "ND",
    "MOSAI"
  ],
  "AcquisitionTime": 82349.7575,
  "AcquisitionDate": 20170829,
  "MagneticFieldStrength": 3,
  "FlipAngle": 51,
  "EchoTime": 0.0424,
  "RepetitionTime": 0.735,
  "EffectiveEchoSpacing": 0.000639989,
  "SliceTiming": [
    0,
    0.2675,
    0.535,
    0.0875,
    0.3575,
    0.625,
    0.1775,
    0.445,
    0,
    0.2675,
    0.535,
    0.0875,
    0.3575,
    0.625,
    0.1775,
    0.445,
    0,
    0.2675,
    0.535,
    0.0875,
    0.3575,
    0.625,
    0.1775,
    0.445,
    0,
    0.2675,
    0.535,
    0.0875,
    0.3575,
    0.625,
    0.1775,
    0.445,
    0,
    0.2675,
    0.535,
    0.0875,
    0.3575,
    0.625,
    0.1775,
    0.445,
    0,
    0.2675,
    0.535,
    0.0875,
    0.3575,
    0.625,
    0.1775,
    0.445,
    0,
    0.2675,
    0.535,
    0.0875,
    0.3575,
    0.625,
    0.1775,
    0.445,
    0,
    0.2675,
    0.535,
    0.0875,
    0.3575,
    0.625,
    0.1775,
    0.445
  ],
  "PhaseEncodingDirection": "j-"
}

Here is the corresponding faulty (?) json sidecar for subject 10027**'s hariri task, and how it looks in subject 10025**:

faulty:

 cat fmriprep/sub-10027**/ses-3/func/sub-10027**_ses-3_task-rest_bold.json | jq
{
  "Manufacturer": "Siemens",
  "ManufacturersModelName": "Skyra",
  "ImageType": [
    "ORIGINAL",
    "PRIMARY",
    "M",
    "MB",
    "ND",
    "MOSAI"
  ],
  "AcquisitionTime": 83200.67,
  "AcquisitionDate": 20190907,
  "MagneticFieldStrength": 3,
  "FlipAngle": 51,
  "EchoTime": 0.039,
  "RepetitionTime": 0.735,
  "EffectiveEchoSpacing": 0.000639989,
  "SliceTiming": [
    0,
    0.09,
    0.18,
    0.2675,
    0.3575,
    0.4475,
    0.5375,
    0.625,
    0.715,
    0.805,
    0.895,
    0.9825,
    1.0725,
    1.1625,
    1.25,
    1.34,
    1.43,
    1.52,
    1.6075,
    1.6975,
    1.7875,
    1.8775,
    1.965,
    2.055,
    2.145,
    2.235,
    2.3225,
    2.4125,
    2.5025,
    2.5925,
    2.68,
    2.77,
    2.86,
    2.95,
    3.0375,
    3.1275,
    3.2175,
    3.3075,
    3.395,
    3.485,
    3.575,
    3.665,
    3.7525,
    3.8425,
    3.9325,
    4.0225,
    4.11,
    4.2,
    4.29,
    4.38,
    4.4675,
    4.5575,
    4.6475,
    4.7375,
    4.825,
    4.915,
    5.005,
    5.095,
    5.1825,
    5.2725,
    5.3625,
    5.4525,
    5.54,
    5.63
  ],
  "PhaseEncodingDirection": "j-"
}

Here is how it looks for the other subject (where the task does not throw an error)

cat fmriprep/sub-10025**/ses-3/func/sub-10025**_ses-3_task-rest_bold.json | jq 
{
  "Manufacturer": "Siemens",
  "ManufacturersModelName": "Skyra",
  "ImageType": [
    "ORIGINAL",
    "PRIMARY",
    "M",
    "MB",
    "ND",
    "MOSAI"
  ],
  "AcquisitionTime": 93213.705,
  "AcquisitionDate": 20190806,
  "MagneticFieldStrength": 3,
  "FlipAngle": 51,
  "EchoTime": 0.0424,
  "RepetitionTime": 0.735,
  "EffectiveEchoSpacing": 0.000639989,
  "SliceTiming": [
    0,
    0.2675,
    0.535,
    0.09,
    0.3575,
    0.625,
    0.1775,
    0.4475,
    0,
    0.2675,
    0.535,
    0.09,
    0.3575,
    0.625,
    0.1775,
    0.4475,
    0,
    0.2675,
    0.535,
    0.09,
    0.3575,
    0.625,
    0.1775,
    0.4475,
    0,
    0.2675,
    0.535,
    0.09,
    0.3575,
    0.625,
    0.1775,
    0.4475,
    0,
    0.2675,
    0.535,
    0.09,
    0.3575,
    0.625,
    0.1775,
    0.4475,
    0,
    0.2675,
    0.535,
    0.09,
    0.3575,
    0.625,
    0.1775,
    0.4475,
    0,
    0.2675,
    0.535,
    0.09,
    0.3575,
    0.625,
    0.1775,
    0.4475,
    0,
    0.2675,
    0.535,
    0.09,
    0.3575,
    0.625,
    0.1775,
    0.4475
  ],
  "PhaseEncodingDirection": "j-"
}

[mih]

Thanks for the detailed report!

It seems the second file is consistent with the description given at http://biobank.ctsu.ox.ac.uk/crystal/field.cgi?id=20227

which is:

The source DICOM images were acquired using the following parameters:

Resolution: 2.4x2.4x2.4 mm
Field-of-view: 88x88x64 matrix
Duration: 6 minutes (490 timepoints)
TR: 0.735 s
TE: 39ms
GE-EPI with x8 multislice acceleration, no iPAT, flip angle 52 degrees, fat saturation

which calls for 8 packs of simultaneously acquired slices, and is what we see for the second file.

[adswa]

I've tried a few more subjects:

sub-3****** has two errors of this kind

1: [ERR] "SliceTiming" value/s contains invalid value as it is greater than RepetitionTime.  SliceTiming values should be in seconds not milliseconds (common mistake). (code: 66 - SLICETIMING_VALUES_GREATOR_THAN_REPETITION_TIME)
		./sub-3******/ses-2/func/sub-3******_ses-2_task-hariri_bold.nii.gz
			Evidence: 0.805,0.8925,0.9825,1.0725,1.1625,1.25,1.34,1.43,1.52,1.6075,1.6975,1.7875,1.8775,1.965,2.055,2.145,2.235,2.3225,2.4125,2.5025,2.5925,2.68,2.77,2.86,2.95,3.0375,3.1275,3.2175,3.3075,3.395,3.485,3.575,3.665,3.7525,3.8425,3.9325,4.0225,4.11,4.2,4.29,4.38,4.4675,4.5575,4.6475,4.7375,4.825,4.915,5.005,5.095,5.1825,5.2725,5.3625,5.4525,5.54,5.63
		./sub-3******/ses-2/func/sub-3******_ses-2_task-rest_bold.nii.gz
			Evidence: 0.805,0.8925,0.9825,1.0725,1.1625,1.25,1.34,1.43,1.52,1.6075,1.6975,1.7875,1.875,1.965,2.055,2.145,2.2325,2.3225,2.4125,2.5025,2.59,2.68,2.77,2.86,2.9475,3.0375,3.1275,3.2175,3.305,3.395,3.485,3.575,3.6625,3.7525,3.8425,3.9325,4.02,4.11,4.2,4.29,4.3775,4.4675,4.5575,4.6475,4.735,4.825,4.915,5.005,5.0925,5.1825,5.2725,5.3625,5.45,5.54,5.63

sub-4****** has one error

1: [ERR] "SliceTiming" value/s contains invalid value as it is greater than RepetitionTime.  SliceTiming values should be in seconds not milliseconds (common mistake). (code: 66 - SLICETIMING_VALUES_GREATOR_THAN_REPETITION_TIME)
		./sub-4******/ses-2/func/sub-4******_ses-2_task-rest_bold.nii.gz
			Evidence: 0.805,0.895,0.9825,1.0725,1.1625,1.2525,1.34,1.43,1.52,1.61,1.6975,1.7875,1.8775,1.9675,2.055,2.145,2.235,2.325,2.4125,2.5025,2.5925,2.6825,2.77,2.86,2.95,3.04,3.1275,3.2175,3.3075,3.3975,3.485,3.575,3.665,3.755,3.8425,3.9325,4.0225,4.1125,4.2,4.29,4.38,4.47,4.5575,4.6475,4.7375,4.8275,4.915,5.005,5.095,5.185,5.2725,5.3625,5.4525,5.5425,5.63

sub-2****** has one

1: [ERR] "SliceTiming" value/s contains invalid value as it is greater than RepetitionTime.  SliceTiming values should be in seconds not milliseconds (common mistake). (code: 66 - SLICETIMING_VALUES_GREATOR_THAN_REPETITION_TIME)
		./sub-2******/ses-2/func/sub-2******_ses-2_task-hariri_bold.nii.gz
			Evidence: 0.8025,0.8925,0.9825,1.0725,1.16,1.25,1.34,1.43,1.5175,1.6075,1.6975,1.785,1.875,1.965,2.055,2.1425,2.2325,2.3225,2.4125,2.5,2.59,2.68,2.77,2.8575,2.9475,3.0375,3.1275,3.215,3.305,3.395,3.485,3.5725,3.6625,3.7525,3.8425,3.93,4.02,4.11,4.2,4.2875,4.3775,4.4675,4.5575,4.645,4.735,4.825,4.915,5.0025,5.0925,5.1825,5.2725,5.36,5.45,5.54,5.63

[mih]

@jbpoline have you heard a report about this kind of error before?

Incorrect values will be an issue for https://fmriprep.readthedocs.io/en/stable/workflows.html#slice-time-correction

[AlexandreHutton]

Related. Confirming the error: some of the slice timing info shows sequential acquisition, others have the (expected) 8 blocks of acquisition times (123-123-123 etc.).
The error is either with (1) the acquisition, or (2) the metadata. I suspect that it's an issue with the metadata; somebody would have noticed that something was wrong if the acquisition took 8x as long as prescribed.
Since the protocol was the same for all subjects, I suggest taking one of the correct slice timings (8x acceleration) and overwriting the ones reporting sequential acquisition.

[jbpoline]

@mih @AlexandreHutton Yes, correct, there are a bunch af wrong time slicing info that need to be overwritten by the good ones. This should be a nice test :) I pinged Fidel on neurostars to check, waiting for his answer.