TY - JOUR
T1 - Fairness in Cardiac Magnetic Resonance Imaging
T2 - Assessing Sex and Racial Bias in Deep Learning-based Segmentation
AU - Puyol-Anton, Esther
AU - Ruijsink, Bram
AU - Mariscal Harana, Jorge
AU - Piechnik, Stefan K
AU - Neubauer, Stefan
AU - Petersen, Steffen E
AU - Razavi, Reza
AU - Chowienczyk, Philip
AU - King, Andrew P
N1 - Funding Information:
EP-A and AK were supported by the EPSRC (EP/R005516/1) and by core funding from the Wellcome/EPSRC Centre for Medical Engineering (WT203148/Z/16/Z). This research was funded in whole, or in part, by the Wellcome Trust WT203148/Z/16/Z. For the purpose of open access, the author has applied a CC BY public copyright license to any author accepted manuscript version arising from this submission. SEP, AK, and RR acknowledge funding from the EPSRC through the Smart Heart Programme grant (EP/P001009/1). EP-A, BR, JM, AK, and RR acknowledged support from the Wellcome/EPSRC Centre for Medical Engineering at King’s College London (WT 203148/Z/16/Z), the NIHR Cardiovascular MedTech Co-operative award to the Guy’s and St Thomas’ NHS Foundation Trust and the Department of Health National Institute for Health Research (NIHR) comprehensive Biomedical Research Centre award to Guy’s & St Thomas’ NHS Foundation Trust in partnership with King’s College London. SEP, SN, and SKP acknowledged the British Heart Foundation for funding the manual analysis to create a cardiovascular magnetic resonance imaging reference standard for the UK Biobank imaging resource in 5,000 CMR scans ( www.bhf.org.uk ; PG/14/89/31194). SEP acknowledged support from the National Institute for Health Research (NIHR) Biomedical Research Centre at Barts. SEP has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under grant agreement No 825903 (euCanSHare project). SEP also acknowledged support from the CAP-AI Programme, London’s First AI Enabling Programme focused on stimulating growth in the capital’s AI Sector. CAP-AI was led by Capital Enterprise in partnership with Barts Health NHS Trust and Digital Catapult and was funded by the European Regional Development Fund and Barts Charity. SEP acknowledged support from the Health Data Research UK, an initiative funded by UK Research and Innovation, Department of Health and Social Care (England) and the devolved administrations, and leading medical research charities. SN and SKP were supported by the Oxford NIHR Biomedical Research Centre and the Oxford British Heart Foundation Centre of Research Excellence.
Funding Information:
EP-A and AK were supported by the EPSRC (EP/R005516/1) and by core funding from the Wellcome/EPSRC Centre for Medical Engineering (WT203148/Z/16/Z). This research was funded in whole, or in part, by the Wellcome Trust WT203148/Z/16/Z. For the purpose of open access, the author has applied a CC BY public copyright license to any author accepted manuscript version arising from this submission. SEP, AK, and RR acknowledge funding from the EPSRC through the Smart Heart Programme grant (EP/P001009/1). EP-A, BR, JM, AK, and RR acknowledged support from the Wellcome/EPSRC Centre for Medical Engineering at King’s College London (WT 203148/Z/16/Z), the NIHR Cardiovascular MedTech Co-operative award to the Guy’s and St Thomas’ NHS Foundation Trust and the Department of Health National Institute for Health Research (NIHR) comprehensive Biomedical Research Centre award to Guy’s & St Thomas’ NHS Foundation Trust in partnership with King’s College London. SEP, SN, and SKP acknowledged the British Heart Foundation for funding the manual analysis to create a cardiovascular magnetic resonance imaging reference standard for the UK Biobank imaging resource in 5,000 CMR scans (www.bhf.org.uk ; PG/14/89/31194). SEP acknowledged support from the National Institute for Health Research (NIHR) Biomedical Research Centre at Barts. SEP has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under grant agreement No 825903 (euCanSHare project). SEP also acknowledged support from the CAP-AI Programme, London’s First AI Enabling Programme focused on stimulating growth in the capital’s AI Sector. CAP-AI was led by Capital Enterprise in partnership with Barts Health NHS Trust and Digital Catapult and was funded by the European Regional Development Fund and Barts Charity. SEP acknowledged support from the Health Data Research UK, an initiative funded by UK Research and Innovation, Department of Health and Social Care (England) and the devolved administrations, and leading medical research charities. SN and SKP were supported by the Oxford NIHR Biomedical Research Centre and the Oxford British Heart Foundation Centre of Research Excellence.
Publisher Copyright:
Copyright © 2022 Puyol-Antón, Ruijsink, Mariscal Harana, Piechnik, Neubauer, Petersen, Razavi, Chowienczyk and King.
PY - 2022/4/7
Y1 - 2022/4/7
N2 - Background: Artificial intelligence (AI) techniques have been proposed for automation of cine CMR segmentation for functional quantification. However, in other applications AI models have been shown to have potential for sex and/or racial bias. The objective of this paper is to perform the first analysis of sex/racial bias in AI-based cine CMR segmentation using a large-scale database. Methods: A state-of-the-art deep learning (DL) model was used for automatic segmentation of both ventricles and the myocardium from cine short-axis CMR. The dataset consisted of end-diastole and end-systole short-axis cine CMR images of 5,903 subjects from the UK Biobank database (61.5 ± 7.1 years, 52% male, 81% white). To assess sex and racial bias, we compared Dice scores and errors in measurements of biventricular volumes and function between patients grouped by race and sex. To investigate whether segmentation bias could be explained by potential confounders, a multivariate linear regression and ANCOVA were performed. Results: Results on the overall population showed an excellent agreement between the manual and automatic segmentations. We found statistically significant differences in Dice scores between races (white ∼94% vs. minority ethnic groups 86–89%) as well as in absolute/relative errors in volumetric and functional measures, showing that the AI model was biased against minority racial groups, even after correction for possible confounders. The results of a multivariate linear regression analysis showed that no covariate could explain the Dice score bias between racial groups. However, for the Mixed and Black race groups, sex showed a weak positive association with the Dice score. The results of an ANCOVA analysis showed that race was the main factor that can explain the overall difference in Dice scores between racial groups. Conclusion: We have shown that racial bias can exist in DL-based cine CMR segmentation models when training with a database that is sex-balanced but not race-balanced such as the UK Biobank.
AB - Background: Artificial intelligence (AI) techniques have been proposed for automation of cine CMR segmentation for functional quantification. However, in other applications AI models have been shown to have potential for sex and/or racial bias. The objective of this paper is to perform the first analysis of sex/racial bias in AI-based cine CMR segmentation using a large-scale database. Methods: A state-of-the-art deep learning (DL) model was used for automatic segmentation of both ventricles and the myocardium from cine short-axis CMR. The dataset consisted of end-diastole and end-systole short-axis cine CMR images of 5,903 subjects from the UK Biobank database (61.5 ± 7.1 years, 52% male, 81% white). To assess sex and racial bias, we compared Dice scores and errors in measurements of biventricular volumes and function between patients grouped by race and sex. To investigate whether segmentation bias could be explained by potential confounders, a multivariate linear regression and ANCOVA were performed. Results: Results on the overall population showed an excellent agreement between the manual and automatic segmentations. We found statistically significant differences in Dice scores between races (white ∼94% vs. minority ethnic groups 86–89%) as well as in absolute/relative errors in volumetric and functional measures, showing that the AI model was biased against minority racial groups, even after correction for possible confounders. The results of a multivariate linear regression analysis showed that no covariate could explain the Dice score bias between racial groups. However, for the Mixed and Black race groups, sex showed a weak positive association with the Dice score. The results of an ANCOVA analysis showed that race was the main factor that can explain the overall difference in Dice scores between racial groups. Conclusion: We have shown that racial bias can exist in DL-based cine CMR segmentation models when training with a database that is sex-balanced but not race-balanced such as the UK Biobank.
KW - cardiac magnetic resonance
KW - deep learning
KW - Fair AI
KW - segmentation
KW - Inequality Fairness in deep learning-based CMR segmentation
UR - http://www.scopus.com/inward/record.url?scp=85135177312&partnerID=8YFLogxK
U2 - 10.3389/fcvm.2022.859310
DO - 10.3389/fcvm.2022.859310
M3 - Article
SN - 2297-055X
VL - 9
JO - Frontiers in Cardiovascular Medicine
JF - Frontiers in Cardiovascular Medicine
M1 - 859310
ER -