Facing privacy in neuroimaging: removing facial features degrades performance of image analysis methods

Background Recent studies have created awareness that facial features can be reconstructed from high-resolution MRI. Therefore, data sharing in neuroimaging requires special attention to protect participants’ privacy. Facial features removal (FFR) could alleviate these concerns. We assessed the impact of three FFR methods on subsequent automated image analysis to obtain clinically relevant outcome measurements in three clinical groups. Methods FFR was performed using QuickShear, FaceMasking, and Defacing. In 110 subjects of Alzheimer’s Disease Neuroimaging Initiative, normalized brain volumes (NBV) were measured by SIENAX. In 70 multiple sclerosis patients of the MAGNIMS Study Group, lesion volumes (WMLV) were measured by lesion prediction algorithm in lesion segmentation toolbox. In 84 glioblastoma patients of the PICTURE Study Group, tumor volumes (GBV) were measured by BraTumIA. Failed analyses on FFR-processed images were recorded. Only cases in which all image analyses completed successfully were analyzed. Differences between outcomes obtained from FFR-processed and full images were assessed, by quantifying the intra-class correlation coefficient (ICC) for absolute agreement and by testing for systematic differences using paired t tests. Results Automated analysis methods failed in 0–19% of cases in FFR-processed images versus 0–2% of cases in full images. ICC for absolute agreement ranged from 0.312 (GBV after FaceMasking) to 0.998 (WMLV after Defacing). FaceMasking yielded higher NBV ( p  = 0.003) and WMLV ( p  ≤ 0.001). GBV was lower after QuickShear and Defacing (both p  < 0.001). Conclusions All three outcome measures were affected differently by FFR, including failure of analysis methods and both “random” variation and systematic differences. Further study is warranted to ensure high-quality neuroimaging research while protecting participants’ privacy. Key Points • Protecting participants’ privacy when sharing MRI data is important . • Impact of three facial features removal methods on subsequent analysis was assessed in three clinical groups . • Removing facial features degrades performance of image analysis methods .

[1]  L. R. Dice.  Measures of the Amount of Ecologic Association Between Species , 1945 .

[2]  A. M. Andrew,et al.  Another Efficient Algorithm for Convex Hulls in Two Dimensions , 1979, Inf. Process. Lett..

[3]  J. Kurtzke.  Rating neurologic impairment in multiple sclerosis , 1983, Neurology.

[4]  D. Cicchetti.  Guidelines, Criteria, and Rules of Thumb for Evaluating Normed and Standardized Assessment Instruments in Psychology. , 1994 .

[5]  Ronald L. Rivest,et al.  Introduction to Algorithms, Second Edition , 2001 .

[6]  Michael Brady,et al.  Improved Optimization for the Robust and Accurate Linear Registration and Motion Correction of Brain Images , 2002, NeuroImage.

[7]  Stephen M Smith,et al.  Fast robust automated brain extraction , 2002, Human brain mapping.

[8]  Stephen M. Smith,et al.  Improved Optimization for the Robust and Accurate Linear Registration and Motion Correction of Brain Images , 2002, NeuroImage.

[9]  J. Weir.  Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. , 2005, Journal of strength and conditioning research.

[10]  John G. Csernansky,et al.  Open Access Series of Imaging Studies (OASIS): Cross-sectional MRI Data in Young, Middle Aged, Nondemented, and Demented Older Adults , 2007, Journal of Cognitive Neuroscience.

[11]  Gregory G. Brown,et al.  A technique for the deidentification of structural brain MR images , 2007, Human brain mapping.

[12]  Arpita Ghosh,et al.  Preventing facial recognition when rendering MR images of the head in three dimensions , 2008, Medical Image Anal..

[13]  Charles Hildebolt,et al.  Facial Recognition From Volume-Rendered Magnetic Resonance Imaging Data , 2009, IEEE Transactions on Information Technology in Biomedicine.

[14]  John Hale,et al.  Quickshear Defacing for Neuroimages , 2011, HealthSec.

[15]  Mark Jenkinson,et al.  Optimizing parameter choice for FSL-Brain Extraction Tool (BET) on 3D T1 images in multiple sclerosis , 2012, NeuroImage.

[16]  Deborah A. Nichols,et al.  Strategies for De-identification and Anonymization of Electronic Health Record Data for Use in Multicenter Research Studies , 2012, Medical care.

[17]  Daniel S. Marcus,et al.  Obscuring Surface Anatomy in Volumetric Imaging Data , 2012, Neuroinformatics.

[18]  Essa Yacoub,et al.  The WU-Minn Human Connectome Project: An overview , 2013, NeuroImage.

[19]  Owen Carmichael,et al.  Standardization of analysis sets for reporting results from ADNI MRI data , 2013, Alzheimer's & Dementia.

[20]  Ludwig Kappos,et al.  Determinants of iron accumulation in deep grey matter of multiple sclerosis patients , 2014, Multiple sclerosis.

[21]  Anlin Wang,et al.  DeID – a data sharing tool for neuroimaging studies , 2015, Front. Neurosci..

[22]  Grigorios Loukides,et al.  Medical Data Privacy Handbook , 2015, Springer International Publishing.

[23]  Bruce R. Rosen,et al.  Brain Genomics Superstruct Project initial data release with structural, functional, and behavioral measures , 2015, Scientific Data.

[24]  Connie L. Parks,et al.  Automated Facial Recognition of Computed Tomography-Derived Facial Images: Patient Privacy Implications , 2017, Journal of Digital Imaging.

[25]  David Manset,et al.  Reproducibility of hippocampal atrophy rates measured with manual, FreeSurfer, AdaBoost, FSL/FIRST and the MAPS-HBSI methods in Alzheimer's disease , 2016, Psychiatry Research: Neuroimaging.

[26]  Massimo Filippi,et al.  Performance of five research-domain automated WM lesion segmentation methods in a multi-center MS study , 2017, NeuroImage.

[27]  Paul Schmidt,et al.  Bayesian inference for structured additive regression models for large-scale problems with applications to medical imaging , 2017 .

[28]  Lei Ai,et al.  A large, open source dataset of stroke anatomical brain images and manual lesion segmentations , 2017, Scientific Data.

[29]  Bennett A. Landman,et al.  Collaborative and Reproducible Research: Goals, Challenges, and Strategies , 2018, Journal of Digital Imaging.

[30]  D. Louis Collins,et al.  A comparison of publicly available linear MRI stereotaxic registration techniques , 2018, NeuroImage.

[31]  Anders Eklund,et al.  Refacing: Reconstructing Anonymized Facial Features Using Gans , 2018, ArXiv.