Personalized Federated Graph Learning on Non-IID Electronic Health Records
Understanding the latent disease patterns embedded in electronic health records (EHRs) is crucial for making precise and proactive healthcare decisions. Federated graph learning-based methods are commonly employed to extract complex disease patterns from the distributed EHRs without sharing the client-side raw data. However, the intrinsic characteristics of the distributed EHRs are typically non-independent and identically distributed (Non-IID), significantly bringing challenges related to data imbalance and leading to a notable decrease in the effectiveness of making healthcare decisions derived from the global model. To address these challenges, we introduce a novel personalized federated learning framework named PEARL, which is designed for disease prediction on Non-IID EHRs. Specifically, PEARL incorporates disease diagnostic code attention and admission record attention to extract patient embeddings from all EHRs. Then, PEARL integrates self-supervised learning into a federated learning framework to train a global model for hierarchical disease prediction. To improve the performance of the client model, we further introduce a fine-tuning scheme to personalize the global model using local EHRs. During the global model updating process, a differential privacy (DP) scheme is implemented, providing a high-level privacy guarantee. Extensive experiments conducted on the real-world MIMIC-III dataset validate the effectiveness of PEARL, demonstrating competitive results when compared with baselines.
Medienart: |
E-Artikel |
---|
Erscheinungsjahr: |
2024 |
---|---|
Erschienen: |
2024 |
Enthalten in: |
Zur Gesamtaufnahme - volume:PP |
---|---|
Enthalten in: |
IEEE transactions on neural networks and learning systems - PP(2024) vom: 19. März |
Sprache: |
Englisch |
---|
Beteiligte Personen: |
Tang, Tao [VerfasserIn] |
---|
Links: |
---|
Themen: |
---|
Anmerkungen: |
Date Revised 19.03.2024 published: Print-Electronic Citation Status Publisher |
---|
doi: |
10.1109/TNNLS.2024.3370297 |
---|
funding: |
|
---|---|
Förderinstitution / Projekttitel: |
|
PPN (Katalog-ID): |
NLM369922514 |
---|
LEADER | 01000naa a22002652 4500 | ||
---|---|---|---|
001 | NLM369922514 | ||
003 | DE-627 | ||
005 | 20240320235139.0 | ||
007 | cr uuu---uuuuu | ||
008 | 240320s2024 xx |||||o 00| ||eng c | ||
024 | 7 | |a 10.1109/TNNLS.2024.3370297 |2 doi | |
028 | 5 | 2 | |a pubmed24n1337.xml |
035 | |a (DE-627)NLM369922514 | ||
035 | |a (NLM)38502617 | ||
040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
041 | |a eng | ||
100 | 1 | |a Tang, Tao |e verfasserin |4 aut | |
245 | 1 | 0 | |a Personalized Federated Graph Learning on Non-IID Electronic Health Records |
264 | 1 | |c 2024 | |
336 | |a Text |b txt |2 rdacontent | ||
337 | |a ƒaComputermedien |b c |2 rdamedia | ||
338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
500 | |a Date Revised 19.03.2024 | ||
500 | |a published: Print-Electronic | ||
500 | |a Citation Status Publisher | ||
520 | |a Understanding the latent disease patterns embedded in electronic health records (EHRs) is crucial for making precise and proactive healthcare decisions. Federated graph learning-based methods are commonly employed to extract complex disease patterns from the distributed EHRs without sharing the client-side raw data. However, the intrinsic characteristics of the distributed EHRs are typically non-independent and identically distributed (Non-IID), significantly bringing challenges related to data imbalance and leading to a notable decrease in the effectiveness of making healthcare decisions derived from the global model. To address these challenges, we introduce a novel personalized federated learning framework named PEARL, which is designed for disease prediction on Non-IID EHRs. Specifically, PEARL incorporates disease diagnostic code attention and admission record attention to extract patient embeddings from all EHRs. Then, PEARL integrates self-supervised learning into a federated learning framework to train a global model for hierarchical disease prediction. To improve the performance of the client model, we further introduce a fine-tuning scheme to personalize the global model using local EHRs. During the global model updating process, a differential privacy (DP) scheme is implemented, providing a high-level privacy guarantee. Extensive experiments conducted on the real-world MIMIC-III dataset validate the effectiveness of PEARL, demonstrating competitive results when compared with baselines | ||
650 | 4 | |a Journal Article | |
700 | 1 | |a Han, Zhuoyang |e verfasserin |4 aut | |
700 | 1 | |a Cai, Zhen |e verfasserin |4 aut | |
700 | 1 | |a Yu, Shuo |e verfasserin |4 aut | |
700 | 1 | |a Zhou, Xiaokang |e verfasserin |4 aut | |
700 | 1 | |a Oseni, Taiwo |e verfasserin |4 aut | |
700 | 1 | |a Das, Sajal K |e verfasserin |4 aut | |
773 | 0 | 8 | |i Enthalten in |t IEEE transactions on neural networks and learning systems |d 2012 |g PP(2024) vom: 19. März |w (DE-627)NLM23236897X |x 2162-2388 |7 nnns |
773 | 1 | 8 | |g volume:PP |g year:2024 |g day:19 |g month:03 |
856 | 4 | 0 | |u http://dx.doi.org/10.1109/TNNLS.2024.3370297 |3 Volltext |
912 | |a GBV_USEFLAG_A | ||
912 | |a GBV_NLM | ||
951 | |a AR | ||
952 | |d PP |j 2024 |b 19 |c 03 |