The preprocessed connectomes project repository of manually corrected skull-stripped T1-weighted anatomical MRI data

  1. Benjamin Puccio
  2. James P Pooley
  3. John S Pellman
  4. Elise C Taverna
  5. R Cameron Craddock

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Background

Skull-stripping is the procedure of removing non-brain tissue from anatomical MRI data. This procedure can be useful for calculating brain volume and for improving the quality of other image processing steps. Developing new skull-stripping algorithms and evaluating their performance requires gold standard data from a variety of different scanners and acquisition methods. We complement existing repositories with manually corrected brain masks for 125 T1-weighted anatomical scans from the Nathan Kline Institute Enhanced Rockland Sample Neurofeedback Study.

Findings

Skull-stripped images were obtained using a semi-automated procedure that involved skull-stripping the data using the brain extraction based on nonlocal segmentation technique (BEaST) software, and manually correcting the worst results. Corrected brain masks were added into the BEaST library and the procedure was repeated until acceptable brain masks were available for all images. In total, 85 of the skull-stripped images were hand-edited and 40 were deemed to not need editing. The results are brain masks for the 125 images along with a BEaST library for automatically skull-stripping other data.

Conclusion

Skull-stripped anatomical images from the Neurofeedback sample are available for download from the Preprocessed Connectomes Project. The resulting brain masks can be used by researchers to improve preprocessing of the Neurofeedback data, as training and testing data for developing new skull-stripping algorithms, and for evaluating the impact on other aspects of MRI preprocessing. We have illustrated the utility of these data as a reference for comparing various automatic methods and evaluated the performance of the newly created library on independent data.

Article activity feed

  2. GigaScience

    Background

    **Reviewer 2. Simon Eskildsen **

    This data note describes a freely available repository of skull-stripped T1 weighted MRI data from adult individuals aged 21-45. In total 125 images and masks are available and may enable researchers to improve skull-stripping and validate algorithms. Sharing data this way is truly the way forward. The paper is well-written and concise. The authors demonstrate the benefit of the data by comparing to commonly used skull-stripping methods. I have only very minor issues/questions: Out of 125 subjects 66 had a psychiatric diagnosis (past or present). This does not seem to be a sample of the general population. Perhaps the authors could explain why more than half of the subjects had a diagnosis? Also, if you scan 125 random subjects, you're bound to find some brain abnormalities. Perhaps 125 was not the initial sample size? Was there a selection process? In the validation it should be mentioned that beast-library-1.1 consists of only 10 MRIs from young individuals.

    Missing references at page 2 lines 33 and 34, page 4 line 12, and page 5 line 62. Page 4 line 17: should this be "Figure 1"? BET is "Brain Extraction Tool" (not "Technique"). Perhaps LPI orientation should be explained?

  3. GigaScience

    Abstract

    Reviewer 1. Xin Di

    In the manuscript titled "The Preprocessed Connectomes Project Repository of Manually Corrected Skull-stripped T1-weighted Anatomical MRI Data", the authors presented a manually corrected skull striped T1 MRI image repository, and demonstrated the usage of this data to test different skull stripping methods. I think the manually corrected images library is a valuable resource, because it is generally considered "gold standard" which could be used to test other automated methods. The manuscript is also well written. I have some minor comments on this manuscript:

    1. In the abstract, it was said that "This procedure is necessary for calculating brain volume and for improving the quality of other image processing steps." I don't think that skull-stripping is really a necessary step for image preprocessing. I use SPM. And I don't typically do skullstripping, unless coregistration of functional and anatomical images failed. I do agree that skullstripping could help to prevent mis-coregistrations, and is particularly helpful for preprocessing of large-scale datasets. But it is hard to say it is necessary. In addition, if non-brain tissues such as bones and fats have been modelling in the segmentation step (e.g. SPM segmentation includes six tissue types), is it necessary to perform a separate skull-striping step before segmentation?
    2. I have downloaded the The NFBS skull-stripped repository. As has been described in the manuscript, it contains single subject's data of raw T1 image, skull-stripped image, and brain mask. Are there some probability maps generated as "NFBS BEaST library" that were used for BEaST skull-stripping? If so, can the authors also make the probability maps available?
    3. In several occasions, references were missing and marked as [?]: Lines 33 and 34, page 2 of 9, "Neurofeedback Study (NFB) [?]" Lines 34 and 35, page 2 of 9, "a deep phenotypic assessment on the rst and second visits [?]" Line 12, page 4 of 9, "FreeSurfer software package [?]". Line 62, page 5 of 9, "plots using the ggplot2 package [?]".
  4. GigaScience

    Now published in GigaScience doi: 10.1186/s13742-016-0150-5

    Benjamin Puccio 1Computational Neuroimaging Lab, Center for Biomedical Imaging and Neuromodulation, Nathan Kline Institute for Psychiatric Research, 140 Old Orangeburg Rd, 10962, Orangeburg, NY, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteJames P Pooley 2Center for the Developing Brain, Child Mind Institute, 445 Park Ave, 10022, New York, NY, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteJohn S Pellman 1Computational Neuroimaging Lab, Center for Biomedical Imaging and Neuromodulation, Nathan Kline Institute for Psychiatric Research, 140 Old Orangeburg Rd, 10962, Orangeburg, NY, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteElise C Taverna 1Computational Neuroimaging Lab, Center for Biomedical Imaging and Neuromodulation, Nathan Kline Institute for Psychiatric Research, 140 Old Orangeburg Rd, 10962, Orangeburg, NY, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteR Cameron Craddock 1Computational Neuroimaging Lab, Center for Biomedical Imaging and Neuromodulation, Nathan Kline Institute for Psychiatric Research, 140 Old Orangeburg Rd, 10962, Orangeburg, NY, USA2Center for the Developing Brain, Child Mind Institute, 445 Park Ave, 10022, New York, NY, USAFind this author on Google ScholarFind this author on PubMedSearch for this author on this siteFor correspondence: ccraddock@nki.rfmh.org

    A version of this preprint has been published in the Open Access journal GigaScience (see paper https://doi.org/10.1186/s13742-016-0150-5), where the paper and peer reviews are published openly under a CC-BY 4.0 license.

  5. Version published to 10.1186/s13742-016-0150-5
  6. Version published to 10.1101/067017v2 on bioRxiv
  7. Version published to 10.1101/067017v1 on bioRxiv