Impact of defacing on automated brain atrophy estimation

Abstract

Background: Defacing has become mandatory for anonymization of brain MRI scans; however, concerns regarding data integrity were raised. Thus, we systematically evaluated the effect of different defacing procedures on automated brain atrophy estimation.

Methods: In total, 268 Alzheimer's disease patients were included from ADNI, which included unaccelerated (n = 154), within-session unaccelerated repeat (n = 67) and accelerated 3D T1 imaging (n = 114). Atrophy maps were computed using the open-source software veganbagel for every original, unmodified scan and after defacing using afni_refacer, fsl_deface, mri_deface, mri_reface, PyDeface or spm_deface, and the root-mean-square error (RMSE) between z-scores was calculated. RMSE values derived from unaccelerated and unaccelerated repeat imaging served as a benchmark. Outliers were defined as RMSE > 75th percentile and by using Grubbs's test.

Results: Benchmark RMSE was 0.28 ± 0.1 (range 0.12-0.58, 75th percentile 0.33). Outliers were found for unaccelerated and accelerated T1 imaging using the 75th percentile cutoff: afni_refacer (unaccelerated: 18, accelerated: 16), fsl_deface (unaccelerated: 4, accelerated: 18), mri_deface (unaccelerated: 0, accelerated: 15), mri_reface (unaccelerated: 0, accelerated: 2) and spm_deface (unaccelerated: 0, accelerated: 7). PyDeface performed best with no outliers (unaccelerated mean RMSE 0.08 ± 0.05, accelerated mean RMSE 0.07 ± 0.05). The following outliers were found according to Grubbs's test: afni_refacer (unaccelerated: 16, accelerated: 13), fsl_deface (unaccelerated: 10, accelerated: 21), mri_deface (unaccelerated: 7, accelerated: 20), mri_reface (unaccelerated: 7, accelerated: 6), PyDeface (unaccelerated: 5, accelerated: 8) and spm_deface (unaccelerated: 10, accelerated: 12).

Conclusion: Most defacing approaches have an impact on atrophy estimation, especially in accelerated 3D T1 imaging. Only PyDeface showed good results with negligible impact on atrophy estimation.

Keywords: Atrophy; Brain; De-identification; Magnetic resonance imaging; Privacy.

Fig. 1

Example of successful defacing approaches on an Alzheimer’s disease patient (female, 69 years of age). Top row shows the sagittal reformations of the defaced image volume, second row shows the axial reformations and the bottom row shows volume renderings. The volume rendering of the full face image (lower left) has been omitted for privacy reasons

Fig. 2

Box-and-whisker plots of the root-mean-square error (RMSE) values after veganbagel processing of Alzheimer’s disease patients. Left column: The benchmark result obtained by calculating the RMSE between gray matter z-scores of full face unaccelerated 3D T1 imaging and the respective same-session unaccelerated repeat 3D T1 imaging. Center and right: RMSE values obtained by comparing z-scores of defaced unaccelerated (center) or accelerated (right) 3D T1 imaging series with the respective full face 3D T1 imaging. The dotted black line denotes the 75th percentile of the RMSE values of the benchmark

Fig. 3

Absolute mean differences of the z-scores plotted as a heat map onto representative axial slices of the SPM152 standard template after applying the different defacing approaches on unaccelerated imaging

Fig. 4

Absolute mean differences of the z-scores plotted as a heat map onto representative axial slices of the SPM152 standard template after applying the different defacing approaches on accelerated imaging