Abstract
Convolutional neural networks (CNNs) are increasingly being used to automate the segmentation of brain structures in magnetic resonance (MR) images for research studies. In other applications, CNN models have been shown to exhibit bias against certain demographic groups when they are under-represented in the training sets. In this work, we investigate whether CNN models for brain MR segmentation have the potential to contain sex or race bias when trained with imbalanced training sets. We train multiple instances of the FastSurferCNN model using different levels of sex imbalance in white subjects. We evaluate the performance of these models separately for white male and white female test sets to assess sex bias, and furthermore evaluate them on black male and black female test sets to assess potential racial bias. We find significant sex and race bias effects in segmentation model performance. The biases have a strong spatial component, with some brain regions exhibiting much stronger bias than others. Overall, our results suggest that race bias is more significant than sex bias. Our study demonstrates the importance of considering race and sex balance when forming training sets for CNN-based brain MR segmentation, to avoid maintaining or even exacerbating existing health inequalities through biased research study findings.
Data used in preparation of this article were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database (adni.loni.usc.edu). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report. A complete listing of ADNI investigators can be found at: http://adni.loni.usc.edu/wp-content/uploads/how_to_apply/ADNI_Acknowledgement_List.pdf.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
A protected group is a set of samples which all share the same value of the protected attribute. A protected attribute is one where fairness needs to be guaranteed, e.g. race and sex.
- 2.
- 3.
We used the segmentations available at https://doi.gin.g-node.org/10.12751/g-node.aa605a/.
- 4.
When assessing differences for multiple regions we did not apply correction for multiple tests because our aim was to be sensitive to possible bias rather than minimise Type I errors.
References
Abbasi-Sureshjani, S., Raumanns, R., Michels, B.E.J., Schouten, G., Cheplygina, V.: Risk of training diagnostic algorithms on data with demographic bias. In: Cardoso, J., et al. (eds.) IMIMIC/MIL3ID/LABELS -2020. LNCS, vol. 12446, pp. 183–192. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-61166-8_20
Banerjee, I., et al.: Reading Race: AI Recognises Patient’s Racial Identity in Medical Images (2021)
Buolamwini, J.: Gender shades: intersectional accuracy disparities in commercial gender classification*. Proc. Mach. Learn. Res. 81, 1–15 (2018)
Cosgrove, K.P., Mazure, C.M., Staley, J.K.: Evolving knowledge of sex differences in brain structure, function, and chemistry. Biol. Psychiat. 62(8), 847–855 (2007). https://doi.org/10.1016/j.biopsych.2007.03.001
Coupé, P., et al.: AssemblyNet: a large ensemble of CNNs for 3D whole brain MRI segmentation. NeuroImage 219, 117026 (2020). https://doi.org/10.1016/J.NEUROIMAGE.2020.117026
Henschel, L., Conjeti, S., Estrada, S., Diers, K., Fischl, B., Reuter, M.: FastSurfer - a fast and accurate deep learning based neuroimaging pipeline. NeuroImage 219, 117012 (2020). https://doi.org/10.1016/j.neuroimage.2020.117012
Isamah, N., et al.: Variability in frontotemporal brain structure: the importance of recruitment of African Americans in neuroscience research. PLoS ONE 5(10), e13642 (2010). https://doi.org/10.1371/journal.pone.0013642
Isensee, F., Jaeger, P.F., Kohl, S.A., Petersen, J., Maier-Hein, K.H.: nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat. Methods 18(2), 203–211 (2021). https://doi.org/10.1038/s41592-020-01008-z
Johnson, E.B., et al.: Recommendations for the use of automated gray matter segmentation tools: evidence from Huntington’s disease. Front. Neurol. 8, 519 (2017). https://doi.org/10.3389/fneur.2017.00519
Kingma, D.P., Ba, J.L.: Adam: a method for stochastic optimization. In: 3rd International Conference on Learning Representations, ICLR 2015 - Conference Track Proceedings (2015). https://doi.org/10.48550/arxiv.1412.6980
Larrazabal, A.J., Nieto, N., Peterson, V., Milone, D.H., Ferrante, E.: Gender imbalance in medical imaging datasets produces biased classifiers for computer-aided diagnosis. Proc. Natl. Acad. Sci. U.S.A. 117(23), 12592–12594 (2020). https://doi.org/10.1073/pnas.1919012117
Ledig, C., et al.: Robust whole-brain segmentation: application to traumatic brain injury. Med. Image Anal. 21(1), 40–58 (2015). https://doi.org/10.1016/j.media.2014.12.003
Ledig, C., Schuh, A., Guerrero, R., Heckemann, R.A., Rueckert, D.: Structural brain imaging in Alzheimer’s disease and mild cognitive impairment: biomarker analysis and shared morphometry database. Sci. Rep. 8(1), 11258 (2018). https://doi.org/10.1038/s41598-018-29295-9
Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., Galstyan, A.: A survey on bias and fairness in machine learning. ACM Comput. Surv. 54(6) (2021). https://doi.org/10.1145/3457607
Puyol-Antón, E., et al.: Fairness in cardiac magnetic resonance imaging: assessing sex and racial bias in deep learning-based segmentation. Front. Cardiovascul. Med. 9, 664 (2022). https://doi.org/10.3389/fcvm.2022.859310
Puyol-Antón, E., et al.: Fairness in cardiac MR image analysis: an investigation of bias due to data imbalance in deep learning based segmentation. In: de Bruijne, M., et al. (eds.) MICCAI 2021. LNCS, vol. 12903, pp. 413–423. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-87199-4_39
Roy, A.G., Conjeti, S., Sheet, D., Katouzian, A., Navab, N., Wachinger, C.: Error corrective boosting for learning fully convolutional networks with limited data. In: Descoteaux, M., Maier-Hein, L., Franz, A., Jannin, P., Collins, D.L., Duchesne, S. (eds.) MICCAI 2017. LNCS, vol. 10435, pp. 231–239. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-66179-7_27
Ruijsink, B., et al.: Fully automated, quality-controlled cardiac analysis from CMR: validation and large-scale application to characterize cardiac function. JACC: Cardiovascul. Imaging 13(3), 684–695 (2020). https://doi.org/10.1016/j.jcmg.2019.05.030
Sudre, C.H., Li, W., Vercauteren, T., Ourselin, S., Jorge Cardoso, M.: Generalised dice overlap as a deep learning loss function for highly unbalanced segmentations. In: Cardoso, M.J., et al. (eds.) DLMIA/ML-CDS -2017. LNCS, vol. 10553, pp. 240–248. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-67558-9_28
Author information
Authors and Affiliations
Consortia
Corresponding author
Editor information
Editors and Affiliations
1 Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Ioannou, S., Chockler, H., Hammers, A., King, A.P., for the Alzheimer’s Disease Neuroimaging Initiative. (2022). A Study of Demographic Bias in CNN-Based Brain MR Segmentation. In: Abdulkadir, A., et al. Machine Learning in Clinical Neuroimaging. MLCN 2022. Lecture Notes in Computer Science, vol 13596. Springer, Cham. https://doi.org/10.1007/978-3-031-17899-3_2
Download citation
DOI: https://doi.org/10.1007/978-3-031-17899-3_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-17898-6
Online ISBN: 978-3-031-17899-3
eBook Packages: Computer ScienceComputer Science (R0)