Subjects with Issues noted in the HCP quality control process
As of the HCP 1200 Subjects Release, a subject data measure (QC_Issue) has been added to ConnectomeDB to flag subjects with notable issues found in the HCP Quality Control process as described for each issue code below. The issues are notable, but were not considered severe enough to exclude the subject's imaging data from release. We are sharing these findings of issues to alert users as they are beginning their own analyses and allow them to decide whether to include or exclude these subjects.
The QC_Issue measure allows users to filter the data for subjects with or without these issues in the ConnectomeDB Subject Dashboard or in the CSV spreadsheet containing individual subject data that is downloadable from the site. For some issues (e.g. scan and surface QC issues), more subject specific information on the noted issue is available on our QC Issues wiki page.
Note: Although we did perform QC on all subjects, our QC process was limited by our staff resources and we are not claiming to have found all data issues for all HCP subjects. Therefore, the absence of a notation code for this measure does not necessarily indicate that there are no issues with the data for this subject.
Issue code A: Scan QC
All HCP subject structural scans and FreeSurfer segmentation and surface outputs from the structural pipeline were subject to a standard quality control process that involved manual viewing and rating of scan quality and anatomical abnormalities by an experienced rater (see Appendix IV: Post Session 1 Quality Control SOP, pp. 17-24). Brain anomalies evident in T1 and T2 scans were noted and further reviewed by a radiologist. Subjects with major radiologic anomalies that were expected to significantly affect connectivity were removed from the study and their imaging and behavioral data are not included in the released data.
Some subjects had focal anomalies that are considered as normal variants or benign findings. We have released their data, but because of their altered anatomy, using data from these subjects may affect some analyses. These subjects are flagged with the issue code A for the QC_Issue measure.
A description and images of the location and kind of anomaly for each of these subjects are detailed in the following Spreadsheet of Anatomical and Segmentation/Surface QC Findings and in the accompanying Powerpoint slides of Anatomical and Segmentation/Surface QC Findings.
Main HCP Issue A Subjects:
Retest HCP Issue A Subjects:
Issue code B: Segmentation and Surface QC
All brain surfaces and brain structure segmentation results generated by FreeSurfer as part of the HCP structural pipeline were also inspected by a rater for any notable issues see Appendix IV: Post Session 1 Quality Control SOP, pp. 25-28). If there were major segmentation or surface errors, we reran the pipeline to attempt to fix the issues in an automated way (due to the cohort size of the HCP and limited staff resources, hand editing of segmentations and surfaces was not attempted on any subjects).
Subjects with focal segmentation and surface errors in their structural pipeline outputs are flagged with the issue code B for the QC_Issue measure.
A description and images of the location and kind of error for each of these subjects are detailed in the following Spreadsheet of Anatomical and Segmentation/Surface QC Findings and in the accompanying Powerpoint slides of Anatomical and Segmentation/Surface QC Findings.
Main HCP Issue B Subjects:
Retest HCP Issue B Subjects:
Issue code C: Magnetic Gradient Temporal Instabilities
Image QC revealed a non-subject-specific issue with scans collected during a few discrete periods in the HCP data collection. When the scans from those periods were further investigated, the problem was revealed to be temporal instability of the scanner's gradient magnetic field.
To find a way to salvage at least part of the affected subjects' data, we evaluated network matrices generated from their rfMRI data before and after ICA-FIX processing (cleaning of artifacts). This evaluation showed that the artifacts arising from the temporal instability could be adequately cleaned from the rfMRI data by the FIX process (the artifacts could correctly be automatically classified or manually reclassified as noise components and removed from the rfMRI data). Therefore, for these subjects we highly recommend only using the FIX denoised version of the rfMRI data in subsequent analyses. We have not currently evaluated and cleaned these subjects’ tfMRI and dMRI data, despite likely remaining artifacts arising from the temporal instability, but we are making these datasets available with the hope that they will serve as a useful test-bed for testing de-noising algorithms.
Subjects collected during these periods of temporal instability are flagged with the issue code C for the QC_Issue measure.
Main HCP Issue C Subjects (no Retest subjects with this issue):
Issue code D: REST Scans with Prominent Artifact in rfMRI MPP data, but fixed by FIX
An evaluation of grayplots (Power 2016, ref below) and/or network matrices generated from each subject's minimally preprocessed rfMRI scan data revealed subjects across the entire HCP study with at least one REST scan (run) that showed prominent artifacts due either to coil problems (the majority of cases), or excessive subject motion. A spreadsheet listing the REST scans exhibiting these artifacts is available on the QC Issues wiki. We believe this process identified most of the REST scans with the most severe artifacts, nonetheless, it should not be considered a complete, comprehensive list of all of the scans with severe artifacts. The list of scans with prominent artifacts will hopefully serve as a useful test-bed for testing de-noising algorithms.
We evaluated network matrices generated from these subjects rfMRI data before and after ICA-FIX processing (cleaning of artifacts). This evaluation showed that the artifacts found could be adequately cleaned from the rfMRI data by the FIX process (the artifacts could correctly be automatically classified or manually reclassified as noise components and removed from the rfMRI data). Therefore, for these subjects we highly recommend only using the FIX denoised version of the rfMRI data in subsequent analyses.
Subjects found to have minimally preprocessed REST scan(s) with prominent artifact in rfMRI MPP data fixed by FIX are flagged with the issue code D for the QC_Issue measure and a list of these subjects is also available on the QC Issues wiki. Note: Nearly all of the temporal instability QC_Issue C subjects described above are also flagged with issue code D.
Main HCP Issue D Subjects:
Retest HCP Issue D Subjects:
Issue code E: FIX Hand Reclassification of Signal Components
An evaluation of network matrices generated from each subject's rfMRI data before and after ICA-FIX processing (cleaning of artifacts) revealed some subjects across the entire HCP study with at least one rfMRI scan (run) that was not sufficiently cleaned by FIX as it was trained early on in the project (see ICA-FIX section of Ref Manual). Investigation of the signal components in the "uncleaned by FIX" scans revealed that some signal components identified by ICA-FIX were clearly noise components likely introduced by scanner coil element temporal instabilities. For these rfMRI scans, we manually reclassified one or more "signal" components as noise components and ran the ReApplyICAFIX pipeline to clean these data. Because the results are dependent on the rfMRI data, after this second round of FIX cleaning (reapplication of FIX with the hand reclassification), we also reran the RestingStateStats, MSMDeDriftAndResample, and MSMAll pipelines and replaced the outputs of these pipelines for the affected subjects.
Subjects who have had one or more rfMRI scan for which one or more FIX signal component(s) have been reclassified by hand to be noise components and had ICA-FIX reapplied are flagged with the issue code E for the QC_Issue measure. A list of these subjects and affected scans is also available on the QC Issues wiki. Note: All of the FIX hand reclassified QC_Issue E subjects are also flagged with issue code D.
Main HCP Issue E Subjects (no Retest scans with this issue):
Power, JD. A simple but useful way to assess fMRI scan qualities. Neuroimage. 2016 Aug 7. pii: S1053-8119(16)30387-1. doi: 10.1016/j.neuroimage.2016.08.009. [Epub ahead of print]