aiqbench analysis
This research introduces Asymmetric Supervised Contrastive Learning (Asymmetric SupCon) to address the heterogeneity of Electronic Health Records across different national healthcare systems. By pre-training temporal Transformer encoders on large-scale Taiwanese data and transferring them to U.S. datasets, the method effectively narrows performance gaps for incident disease prediction without requiring explicit alignment of negative clinical trajectories. The results demonstrate that this asymmetric objective outperforms standard approaches in few-shot scenarios, offering a viable pathway for cross-border medical AI deployment.
Bridging the EHR Divide: Asymmetric Contrastive Learning for Cross-National Medical Representation Transfer
arXiv:2610.04946v1 Announce Type: new Abstract: Cross-system transfer of longitudinal Electronic Health Record (EHR) representations is challenging because clinical coding, patient populations, and healthcare workflows differ substantially across institutions and countries. We introduce Asymmetric Supervised Contrastive Learning (Asymmetric SupCon), a task-specific pre-training objective motivated by the heterogeneity of negative clinical outcomes. The objective clusters patients sharing a target positive outcome without explicitly attracting negative trajectories toward one another. We pre-train temporal Transformer encoders on longitudinal records from 3.98 million patients in the Taiwanese National Health Insurance Research Database (NHIRD) and transfer them to two U.S. EHR datasets, MIMIC-IV and EHRSHOT. A hybrid semantic mapping pipeline combining direct mappings with embedding-based retrieval enables transfer across heterogeneous clinical vocabularies. On MIMIC-IV, NHIRD pre-training consistently improves over random initialization while substantially narrowing the performance gap to task-specific in-domain pre-training. On EHRSHOT, the transferred models show particularly strong few-shot performance for incident disease prediction. A controlled objective ablation shows that Asymmetric SupCon achieves the best AUPRC on three of four evaluated tasks and is 0.003 AUPRC below Standard SupCon on the fourth. These results support asymmetric contrastive pre-training as an effective approach for task-specific cross-national EHR representation transfer. Code is available at https://github.com/qingYzhang/Asymmetric_SupCon.
Abstract:Cross-system transfer of longitudinal Electronic Health Record (EHR) representations is challenging because clinical coding, patient populations, and healthcare workflows differ substantially across institutions and countries. We introduce Asymmetric Supervised Contrastive Learning (Asymmetric SupCon), a task-specific pre-training objective motivated by the heterogeneity of negative clinical outcomes. The objective clusters patients sharing a target positive outcome without explicitly attracting negative trajectories toward one another. We pre-train temporal Transformer encoders on longitudinal records from 3.98 million patients in the Taiwanese National Health Insurance Research Database (NHIRD) and transfer them to two U.S. EHR datasets, MIMIC-IV and EHRSHOT. A hybrid semantic mapping pipeline combining direct mappings with embedding-based retrieval enables transfer across heterogeneous clinical vocabularies. On MIMIC-IV, NHIRD pre-training consistently improves over random initialization while substantially narrowing the performance gap to task-specific in-domain pre-training. On EHRSHOT, the transferred models show particularly strong few-shot performance for incident disease prediction. A controlled objective ablation shows that Asymmetric SupCon achieves the best AUPRC on three of four evaluated tasks and is 0.003 AUPRC below Standard SupCon on the fourth. These results support asymmetric contrastive pre-training as an effective approach for task-specific cross-national EHR representation transfer. Code is available at this https URL.
Machine Learning (cs.LG)
arXiv:2610.04946 [cs.LG]
(or arXiv:2610.04946v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.04946 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 289 words
