PCI Registered Reports

Announcements

We are recruiting recommenders (editors) from all research fields!

Your feedback matters! If you have authored or reviewed a Registered Report at Peer Community in Registered Reports, then please take 5 minutes to leave anonymous feedback about your experience, and view community ratings.

Back

346

Title

Defacing biases in manual and automated quality assessments of structural MRI with MRIQCuse asterix (*) to get italics

Authors

Céline Provins, Yasser Alemán-Gómez, Jonas Richiardi, Russell A. Poldrack, Patric Hagmann, Oscar EstebanPlease use the format "First name initials family name" as in "Marie S. Curie, Niels H. D. Bohr, Albert Einstein, John R. R. Tolkien, Donna T. Strickland"

Year

2022

Picture

Abstract

<p>A critical requirement before data-sharing of human neuroimaging is removing facial features to protect individuals’ privacy. However, not only does this process redact identifiable information about individuals, but it also removes non-identifiable information. This may introduce undesired variability into downstream analysis and interpretation. Here, we pre-register a study design to investigate the degree to which the so-called defacing alters the quality assessment of T1-weighted images of the human brain from the openly available “IXI dataset”. The effect of defacing on manual quality assessment will be investigated on a single-site subset of the dataset (N=185). By means of repeated-measures analysis of variance (rm-ANOVA) or linear mixed-effects models in case data do not meet rm-ANOVA’s assumptions, we will determine whether four trained human raters’ perception of quality is significantly influenced by defacing by comparing their ratings on the same set of images in two conditions: “non-defaced” (i.e., preserving facial features) and “defaced”. Relatedly, we will also verify that defaced images are systematically assigned higher quality ratings. In addition, we will investigate these biases on automated quality assessments by applying multivariate rm-ANOVA (rm-MANOVA) on the image quality metrics extracted with MRIQC on the full IXI dataset (N=580; three acquisition sites). The analysis code, tested on simulated data, is made openly available with this pre-registration report. This study seeks evidence of the deleterious effects of defacing on data quality assessments by humans and machine agents.</p>

Indicate the full web address (DOI or URL) giving public access to these data (if you have any problems with the deposit of your data, please contact contact@rr.peercommunityin.org).

You should fill this box only if you chose 'All or part of the results presented in this preprint are based on data'. URL must start with http:// or https://

Indicate the full web address (DOI or URL) giving public access to these scripts (if you have any problems with the deposit of your scripts, please contact contact@rr.peercommunityin.org).

You should fill this box only if you chose 'Scripts were used to obtain or analyze the results'. URL must start with http:// or https://

Indicate the full web address (DOI, SWHID or URL) giving public access to these codes (if you have any problems with the deposit of your codes, please contact contact@rr.peercommunityin.org).

You should fill this box only if you chose 'Codes have been used in this study'. URL must start with http:// or https://

Keywords

biases, defacing, quality control, quality assessment, image quality metrics, iqms, manual ratings, MRIQC, MRI, structural MRI

Methods that require specific expertise (optional)

NonePlease indicate the methods that may require specialised expertise during the peer review process (use a comma to separate various required expertises).

Thematic fields

Medical Sciences

Suggested reviewers - Suggest up to 10 reviewers (provide names and Email addresses). (Optional)

No need for them to be recommenders of PCI Registered Reports. Please do not suggest reviewers for whom there might be a conflict of interest. Reviewers are not allowed to review preprints written by close colleagues (with whom they have published in the last four years, with whom they have received joint funding in the last four years, or with whom they are currently writing a manuscript, or submitting a grant proposal), or by family members, friends, or anyone for whom bias might affect the nature of the review - see the code of conduct

Opposed reviewers - Suggest up to 5 people not to invite as reviewers. (Optional)

e.g. John Doe [john@doe.com]

Submission date

2022-11-28 10:59:32

Recommender

D. Samuel Schwarzkopf

Reviewers

Cassandra Gould van Praag, Catherine Morgan, Abiola Akinnubi

or Register
Submit a report