TY - JOUR
T1 - Fair patient model
T2 - Mitigating bias in the patient representation learned from the electronic health records
AU - Sivarajkumar, Sonish
AU - Huang, Yufei
AU - Wang, Yanshan
N1 - Publisher Copyright:
© 2023 Elsevier Inc.
PY - 2023/12
Y1 - 2023/12
N2 - Objective: To pre-train fair and unbiased patient representations from Electronic Health Records (EHRs) using a novel weighted loss function that reduces bias and improves fairness in deep representation learning models. Methods: We defined a new loss function, called weighted loss function, in the deep representation learning model to balance the importance of different groups of patients and features. We applied the proposed model, called Fair Patient Model (FPM), to a sample of 34,739 patients from the MIMIC-III dataset and learned patient representations for four clinical outcome prediction tasks. Results: FPM outperformed the baseline models in terms of three fairness metrics: demographic parity, equality of opportunity difference, and equalized odds ratio. FPM also achieved comparable predictive performance with the baselines, with an average accuracy of 0.7912. Feature analysis revealed that FPM captured more information from clinical features than the baselines. Conclusion: FPM is a novel method to pre-train fair and unbiased patient representations from the EHR data using a weighted loss function. The learned representations can be used for various downstream tasks in healthcare and can be extended to other domains where fairness is important.
AB - Objective: To pre-train fair and unbiased patient representations from Electronic Health Records (EHRs) using a novel weighted loss function that reduces bias and improves fairness in deep representation learning models. Methods: We defined a new loss function, called weighted loss function, in the deep representation learning model to balance the importance of different groups of patients and features. We applied the proposed model, called Fair Patient Model (FPM), to a sample of 34,739 patients from the MIMIC-III dataset and learned patient representations for four clinical outcome prediction tasks. Results: FPM outperformed the baseline models in terms of three fairness metrics: demographic parity, equality of opportunity difference, and equalized odds ratio. FPM also achieved comparable predictive performance with the baselines, with an average accuracy of 0.7912. Feature analysis revealed that FPM captured more information from clinical features than the baselines. Conclusion: FPM is a novel method to pre-train fair and unbiased patient representations from the EHR data using a weighted loss function. The learned representations can be used for various downstream tasks in healthcare and can be extended to other domains where fairness is important.
UR - http://www.scopus.com/inward/record.url?scp=85178389335&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85178389335&partnerID=8YFLogxK
U2 - 10.1016/j.jbi.2023.104544
DO - 10.1016/j.jbi.2023.104544
M3 - Article
C2 - 37995843
AN - SCOPUS:85178389335
SN - 1532-0464
VL - 148
JO - Journal of Biomedical Informatics
JF - Journal of Biomedical Informatics
M1 - 104544
ER -