MRI Quality Control & Recommended Image Inclusion Criteria

DOI: 10.15154/z563-zd24 (Release 5.1)

Published

March 14, 2025

List of Instruments

Name of Instrument Table Name
Quality Control - Recommended Image Inclusion mri_y_qc_incl
Quality Control - MRI Clinical Report/Findings mri_y_qc_clfind
Quality Control - Motion mri_y_qc_motion
Quality Control - Raw - Diffusion MRI mri_y_qc_raw_dmr
Quality Control - Raw - Event mri_y_qc_raw_event
Quality Control - Raw - Resting State fMRI mri_y_qc_raw_rsfmr
Quality Control - Raw - Structural MRI - T1 mri_y_qc_raw_smr_t1
Quality Control - Raw - Structural MRI - T2 mri_y_qc_raw_smr_t2
Quality Control - Raw - Task fMRI - All mri_y_qc_raw_tfmr_all
Quality Control - Raw - Task fMRI - MID mri_y_qc_raw_tfmr_mid
Quality Control - Raw - Task fMRI - N-Back mri_y_qc_raw_tfmr_nback
Quality Control - Raw - Task fMRI - SST mri_y_qc_raw_tfmr_sst
Quality Control - Manual - Freesurfer mri_y_qc_man_fsurf
Quality Control - Manual - Post-processing - Diffusion MRI mri_y_qc_man_post_dmr
Quality Control - Manual - Post-processing - Functional MRI mri_y_qc_man_post_fmr
Quality Control - Manual - Post-processing - Structural MRI - T2w mri_y_qc_man_post_t2w
Quality Control - Automatic - Post-processing mri_y_qc_auto_post

General Information

An overview of the ABCD Study® can be found at abcdstudy.org and detailed descriptions of the assessment protocols are available at ABCD Protocols. This page describes the contents of various instruments available for download. To understand the context of this information, refer to the release notes Start Page and Imaging Overview.

MRI Clinical Findings

T1w and T2w-weighted images, if available, were screened for incidental findings by a Board Certified Neuroradiologist. Any findings requiring clinical investigation were relayed to appropriate site personnel via the ABCD Coordinating Center (CC).

The most important measure is the Report Score (mrif_score):

  • 0 = Image artifacts prevent radiology read
  • 1 = No abnormal findings
  • 2 = Normal anatomical variant of no clinical significance
  • 3 = Consider clinical referral
  • 4 = Consider immediate clinical referral

Although not included in the recommended inclusion criteria, users may, depending on their research question and analytical design, opt to exclude participants with mrif_score != 3 OR mrif_score !=4.

MRI Raw QC

  • Protocol compliance checking
    • performed by on-site FIONA workstations to provide feedback to scan operators
    • out-of-compliance series reviewed by DAIC staff
    • criteria included whether key imaging parameters matched expected values for a given scanner, such as voxel size or repetition time
    • presence or absence of B0 distortion field map series was checked for diffusion MRI (dMRI) and functional MRI (fMRI) series
    • each imaging series checked for completeness (i.e., no missing files)
  • Automated quality control metrics
    • Structural MRI (sMRI): mean and standard deviation of brain values and spatial SNR
    • dMRI: mean motion (average framewise displacement), and the number of slices and frames affected by slice dropout caused by abrupt head motion
    • fMRI: mean motion (average framewise displacement), the number of seconds with framewise displacements less than 0.2, 0.3, or 0.4 mm (Power, et al., 2012), temporal SNR (tSNR) (Triantafyllou, et al., 2005)
  • Manual review of data quality
    • reviewers assigned binary QC score
      • 0 = reject
        • most severe artifacts or irregularities
        • rejected series excluded from subsequent processing and analysis
      • 1 = accept
    • types of images reviewed
      • T1w, T2w, dMRI, dMRI field map, fMRI, and fMRI field map
      • raw and some derived images were reviewed
      • dMRI derived images included average b=0 image, FA, MD, tensor fit residual error, and direction encoded color image
      • fMRI derived images included the average across time and the temporal standard deviation
    • inspected for signs of artifacts and poor image quality
      • presence of wrap-around field of view artifacts
      • brain cut-off due to the participant motion outside prescribed slices
      • magnetic susceptibility artifacts due to dental implants
      • T1w and T2w motion artifact (e.g. blurring and ghosting)

FreeSurfer QC

  • Manual review of FreeSurfer cortical surface reconstruction
    • reviewers assigned binary (0|1) QC score
      • 0 = reject
        • most severe artifacts or irregularities
        • results still included in shared tabulated data
        • recommended exclusion from group analyses involving cortical surface ROIs
      • 1 = accept
    • reviewers gauged the severity of five types of artifact or processing problem
      • motion
      • intensity inhomogeneity
      • white matter underestimation
      • pial overestimation
      • magnetic susceptibility artifact
    • numeric values assigned on a scale of 0-3
      • absent, mild, moderate, or severe, respectively
      • QC score of 0 assigned if severity score of 3 for any artifact type

Note: Imaging-derived results are included in shared tabulated data regardless of post-processing QC. QC variables (derived from procedures described above) are included in shared tabulated data. The overall, binary QC score described above indicates whether inclusion or exclusion is recommended based on this criterion alone (see above Recommended Imaging Inclusion). FreeSurfer QC covers ~6.6% of participant-events with imaging data.

Manual Post Processing QC

sMRI T2w Post Processing QC

  • Manual review of DTI reconstruction
    • reviewers assigned binary (0|1) QC score
      • 0 = reject
        • most severe artifacts or irregularities
        • results still included in shared tabulated data
        • recommended exclusion from group analyses involving cortical, subcortical, and tract-based ROIs
      • 1 = accept
    • reviewers gauged the severity of four types of artifact or processing problem
      • motion
      • intensity inhomogeneity
      • magnetic susceptibility artifact
      • registration with T1w image
    • numeric values assigned on a scale of 0-3
      • absent, mild, moderate, or severe, respectively
      • QC score of 0 assigned if severity score of 3 for any artifact type

Note: The T2w Post Processing QC covers ~2.4% of participant-events with imaging data.

dMRI Post Processing QC

  • Manual review of processed dMRI data
    • reviewers assigned binary (0|1) QC score
      • 0 = reject
        • most severe artifacts or irregularities
        • results still included in shared tabulated data
        • recommended exclusion from group analyses involving cortical, subcortical, and tract-based ROIs
      • 1 = accept
    • reviewers gauged the severity of five types of artifact or processing problem
      • B0 warping
      • image quality based on motion-related artifacts and magnetic susceptibility artifact
      • full head coverage
      • registration with T1w image
      • accuracy of fiber tract segmentation
    • numeric values assigned on a scale of 0-3
      • absent, mild, moderate, or severe, respectively
      • QC score of 0 assigned if severity score of 3 for any artifact type

Note: The dMRI Post Processing QC covers ~7.4% of participant-events with imaging data.

fMRI Post Processing QC

  • Manual review of processed fMRI data
    • reviewers assigned binary (0|1) QC score
      • 0 = reject
        • most severe artifacts or irregularities
        • results still included in shared tabulated data
        • recommended exclusion from group analyses involving cortical, subcortical, and tract-based ROIs
      • 1 = accept
    • reviewers gauged the severity of five types of artifact or processing problem
      • B0 warping
      • image quality based primarily on magnetic susceptibility artifact
      • full head coverage
      • registration with T1w image
    • numeric values assigned on a scale of 0-3
      • absent, mild, moderate, or severe, respectively
      • QC score of 0 assigned if severity score of 3 for any artifact type

Note: The fMRI Post Processing QC covers ~6.3% of participant-events with imaging data.

Automated Post Processing QC

Automated QC measures were defined and calculated based on processed imaging data.

  • FreeSurfer
    • number of topological defects
      • calculated from Euler number
  • dMRI
    • field of view (FOV) brain cutoff
      • quantified by % intersection of brain mask with frame borders
    • registration to T1w
      • window-based estimation of geometric registration discrepancy with respect to the T1 scan, decomposed into rigid and warp components for calculation of registration error
  • fMRI
    • field of view (FOV) brain cutoff
      • quantified by % intersection of brain mask with frame borders
    • registration to T1w
      • window-based estimation of geometric registration discrepancy with respect to the T1 scan, decomposed into rigid and warp components for calculation of registration error
  • sMRI T2w
    • registration to T1w
      • window-based estimation of geometric registration discrepancy with respect to the T1 scan, decomposed into rigid and warp components for calculation of registration error

MRI post-processing quality control

Our manual quality control process involves manual examination of brain images from a subset of participant-events for each modality, with ratings for each dataset according to pre-defined quality criteria such as brain cutoff, residual distortion, or registration to T1. Such measures are described above (see FreeSurfer QC, sMRI T2w Post Processing QC, dMRI Post Processing QC, and fMRI Post Processing QC). Because an exhaustive manual review of every scan is not practical, we deploy statistical learning/AI guided sub-sampling methods where we generate automated metrics for an array of quality control issues (see Automated Post Processing QC) and sub-select participant-events to be sent for a given modality to manual review based on their quality measure scores. The automatic selection priority scores generally indicate how likely a dataset of a given type is to contain identifiable data quality issues such as a residual distortion or brain cutoff. ABCD Release 4.0 manual review post processing sampling categories included the following four primary groups:

Failed post-processing QC from ABCD Release 3.0: All participant-events that failed QC for a given modality in Release 3.0 were manually reviewed again after Release 4.0 processing, with the expectation that they would be likely candidates for failure, unless perhaps recovered due to improvements in the current Release 4.0 processing pipeline. These made up about 25-50% of manually reviewed participant-events.

Random selection of participant-events: Roughly 5% of the manually reviewed participant-events were chosen at random to account for possible biases and insensitivities of the automated measures that guide the sub-sampling selection process.

Statistical outliers: ROI summary vectors for all individual participant-events are used to deduce a statistical ensemble profile. Participant-events with the greatest “distance” to the ensemble are tagged as outliers and selected for manual review. The formal framework we used for defining the ensemble statistics and the “distance” measure between a scan and the ensemble is based on the Mahalanobis distance (see https://en.wikipedia.org/wiki/Mahalanobis_distance). Mahalanobis distance is a scalar measure of the distance between a point P, defined in a multi-dimensional vector space, and a distribution D. In our case, participant-events are sorted based on their Mahalanobis distances, and our manual review goal targeted the top 5% of that list. These made up ~20-25% of manually reviewed participant-events.

Classifier guided selections: Using the manual scoring of the above sets, namely the random samples, Release 3.0 failures, and Mahalanobis distance outliers, we collected a subset of manually labeled scan data. For each participant-event we generated a set of automated measures described below in Automated Post Processing QC (automated review). Using binarized (pass/fail) manual quality control (QC) labels associated with these measures, we constructed a Bayesian classifier (see https://en.wikipedia.org/wiki/Bayes_classifier) that calculates the probability of QC failure associated with different types of QC issues (e.g., bad registration or brain cutoff) for any given dataset. Participant-events are sorted based on their QC failure probabilities, and roughly the top 3-5% were selected for manual review for each modality, making up ~30-50% of manually reviewed participant-events.

Methods

Image processing and analysis methods corresponding to ABCD Release 2.0.1 are described in Hagler et al., 2019, Image processing and analysis methods for the Adolescent Brain Cognitive Development Study. Neuroimage, 202:116091 (doi: 10.1016/j.neuroimage.2019.116091). Changes to image processing and analysis methods in Release 3.0 and Release 4.0 are documented below.

Changes for ABCD 3.0

Post processing QC and inclusion criteria

As in ABCD Release 2.0.1, FreeSurfer cortical surface reconstructions were reviewed for all participant-events that successfully completed FreeSurfer processing. New for Release 3.0, we used a sampling approach for dMRI and fMRI, wherein approximately 20% of the dMRI/fMRI sessions were manually reviewed for postprocessing dMRI/fMRI QC. We selected participant-events for review based on random-sampling, Bayesian classifiers based on automated QC metrics, and multivariate outlier detection. Recommended inclusion criteria were updated to include manual post-processing QC, new automated post-processing QC metrics, and additional variables (e.g., E-Prime timing match to imaging series). Modality-specific imaging inclusion flags are included in a new NDA data structure, abcd_imgincl01.

Protocol compliance

The determination of protocol compliance for scans from GE scanners (reflected for example in elements iqc_sst_1_pc_score or iqc_sst_total_passpc in NDA data structure mriqcrp202) was changed to use the “ImagesInAcquistion” DICOM header attribute rather than the number of DICOM files. Imaging series with missing files, i.e., fewer files than the ImagesInAcquisition, are marked as incomplete. For Siemens and Philips scans, the same DICOM header field recording the number of images collected is not available, and so series with fewer files than expected for the given series type are marked as incomplete.

Changes for ABCD 4.0

MRI raw QC

The MRI raw QC NDA data structures mriqcp103 and mriqcp203 were updated to reflect a remapping of the names and abbreviations used to categorize MRI raw QC issues. This was done to remove redundancy in the categories used previously. Also, QC issue variables were removed where inappropriate for a given scan type, such as the dMRI-specific “fa” (fractional anisotropy map issues) for sMRI or fMRI (e.g., iqc_mid_fa_qc).

QC issue names Previously used names
field of view (fov) dorsal cutoff (dco)
ventral cutoff (vco)
wrap around (wr)
susceptibility artifact (sus) distortion (dis)
signal dropout (sd)
signal inhomogeneity (si)
horizontal banding (hb) slice
horizontal banding (hb)
other flag
other

Post processing QC and inclusion criteria

We extended the sampling approach used in Release 3.0 to include FreeSurfer, dMRI, fMRI, and sMRI T2w post-processing QC. Approximately 2-7% of all participant-events were manually reviewed for post-processing QC for each modality. We selected participant-events for review based on post-processing QC failure in Release 3.0, multivariate outlier detection, Bayesian classifiers based on automated QC metrics, and random-sampling (~5% of total selected for review). Multivariate outlier detection was implemented using Mahalanobis distance calculated from the tabulated data (ROI averages) for a given imaging modality. Recommended inclusion criteria were updated to include manual post-processing QC for T2w. A new manual post-processing QC NDA data structures was created for sMRI T2w: abcd_t2wqc01.

The following manual post-processing QC NDA data structures were renamed:

  • freesqc01 to abcd_fsurfqc01
  • dmriqc01 to abcd_dmriqc01
  • fmriqc01 to abcd_fmriqc01

Changes to data dictionaries

  • Raw MRI QC
    • new versions of NDA data structures mriqcrp103 and mriqcrp203 based on mriqcrp102 and mriqcrp202
      • added new elements related to types of QC issues
      • removed other elements related to unused or deprecated QC issues
      • changes to DataType and Size for some elements
  • Manual post-processing QC
    • new NDA data structure abcd_t2wqc01
    • new NDA data structure abcd_fmriqc01 based on fmriqc01
      • added deap alias fmri_man_postqc_notes
      • removed alias fmri_postqc_visitid
    • new NDA data structure abcd_dmriqc01 based on dmriqc01
      • added deap aliases (with dmri_manu_postqc_...)
    • new NDA data structure abcd_fsurfqc01 based on freesqc01
      • removed alias fsqc_visit_id
      • added DEAP aliases (with fsurf_manu_postqc_...)

References

Hagler, D.J., Jr., Hatton, S., Cornejo, M.D., Makowski, C., Fair, D.A., Dick, A.S., Sutherland, M.T., Casey, B.J., Barch, D.M., Harms, M.P., Watts, R., Bjork, J.M., Garavan, H.P., Hilmer, L., Pung, C.J., Sicat, C.S., Kuperman, J., Bartsch, H., Xue, F., Heitzeg, M.M., Laird, A.R., Trinh, T.T., Gonzalez, R., Tapert, S.F., Riedel, M.C., Squeglia, L.M., Hyde, L.W., Rosenberg, M.D., Earl, E.A., Howlett, K.D., Baker, F.C., Soules, M., Diaz, J., de Leon, O.R., Thompson, W.K., Neale, M.C., Herting, M., Sowell, E.R., Alvarez, R.P., Hawes, S.W., Sanchez, M., Bodurka, J., Breslin, F.J., Morris, A.S., Paulus, M.P., Simmons, W.K., Polimeni, J.R., van der Kouwe, A., Nencka, A.S., Gray, K.M., Pierpaoli, C., Matochik, J.A., Noronha, A., Aklin, W.M., Conway, K., Glantz, M., Hoffman, E., Little, R., Lopez, M., Pariyadath, V., Weiss, S.R., Wolff-Hughes, D.L., DelCarmen-Wiggins, R., Feldstein Ewing, S.W., Miranda-Dominguez, O., Nagel, B.J., Perrone, A.J., Sturgeon, D.T., Goldstone, A., Pfefferbaum, A., Pohl, K.M., Prouty, D., Uban, K., Bookheimer, S.Y., Dapretto, M., Galvan, A., Bagot, K., Giedd, J., Infante, M.A., Jacobus, J., Patrick, K., Shilling, P.D., Desikan, R., Li, Y., Sugrue, L., Banich, M.T., Friedman, N., Hewitt, J.K., Hopfer, C., Sakai, J., Tanabe, J., Cottler, L.B., Nixon, S.J., Chang, L., Cloak, C., Ernst, T., Reeves, G., Kennedy, D.N., Heeringa, S., Peltier, S., Schulenberg, J., Sripada, C., Zucker, R.A., Iacono, W.G., Luciana, M., Calabro, F.J., Clark, D.B., Lewis, D.A., Luna, B., Schirda, C., Brima, T., Foxe, J.J., Freedman, E.G., Mruzek, D.W., Mason, M.J., Huber, R., McGlade, E., Prescot, A., Renshaw, P.F., Yurgelun-Todd, D.A., Allgaier, N.A., Dumas, J.A., Ivanova, M., Potter, A., Florsheim, P., Larson, C., Lisdahl, K., Charness, M.E., Fuemmeler, B., Hettema, J.M., Maes, H.H., Steinberg, J., Anokhin, A.P., Glaser, P., Heath, A.C., Madden, P.A., Baskin-Sommers, A., Constable, R.T., Grant, S.J., Dowling, G.J., Brown, S.A., Jernigan, T.L., Dale, A.M. (2019) Image processing and analysis methods for the Adolescent Brain Cognitive Development Study. Neuroimage, 202:116091.

Power, J.D., Barnes, K.A., Snyder, A.Z., Schlaggar, B.L., Petersen, S.E. (2012) Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage, 59:2142-54.

Triantafyllou, C., Hoge, R.D., Krueger, G., Wiggins, C.J., Potthast, A., Wiggins, G.C., Wald, L.L. (2005) Comparison of physiological noise at 1.5 T, 3 T and 7 T and optimization of fMRI acquisition parameters. Neuroimage, 26:243-50.