A Correspondence Score for Validating Learned Patient Representations in Clinical Tabular Data
Abstract
Identifying patients with similar clinical characteristics is important for cohort discovery, clinical decision support, and personalized medicine. However, it is unclear whether learned patient representations preserve the similarities present in the original clinical data. We propose a correspondence score to directly assess how well these relationships are preserved, independently of labeled outcomes. The score combines complementary relationship-preservation metrics, with dataset-specific weights estimated through Bayesian optimization. We evaluated the approach on four complete blood count datasets using representations generated by five tabular foundation models and six classical transformation methods. Correspondence differed substantially across methods and representation dimensionalities. Patient relationships were generally best preserved at intermediate dimensionalities, while stronger reduction or expansion introduced greater deviations from the original data structure. Foundation models preserved correspondence more consistently than classical methods, with improvements ranging from approximately 5% to over 125%. The proposed score provides a way to evaluate representation quality before downstream use and supports informed selection of representation methods and dimensionalities.
Related
- Reuse of learned subtask representations during human planning
- Dissociating neural representations of multiple learned sensorimotor mappings
- Biologically Informed Multi-Omics Integration Reveals Clinically Meaningful Patient Representations in Acute Myeloid Leukemia
- FloREN: Decoding Immune Regulatory Networks through Interpretable Graph Transformer Patient Representations
- Between patterns and predictions: Interpretable latent EEG representations for clinical insights.
BibTeX
@misc{delouee2026correspondence,
title = {{A Correspondence Score for Validating Learned Patient Representations in Clinical Tabular Data}},
author = {Majid Lotfian Delouee and Sjors G. J. G. In 't Veld and Martijn C. Schut},
howpublished = {arXiv},
year = {2026},
month = aug,
doi = {10.48550/arxiv.2608.27489},
eprint = {2608.27489},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.27489},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - A Correspondence Score for Validating Learned Patient Representations in Clinical Tabular Data
AU - Majid Lotfian Delouee
AU - Sjors G. J. G. In 't Veld
AU - Martijn C. Schut
PY - 2026
DA - 2026/08/26
PB - arXiv
DO - 10.48550/arxiv.2608.27489
UR - https://arxiv.org/abs/2608.27489
AB - Identifying patients with similar clinical characteristics is important for cohort discovery, clinical decision support, and personalized medicine. However, it is unclear whether learned patient representations preserve the similarities present in the original clinical data. We propose a correspondence score to directly assess how well these relationships are preserved, independently of labeled outcomes. The score combines complementary relationship-preservation metrics, with dataset-specific weights estimated through Bayesian optimization. We evaluated the approach on four complete blood count datasets using representations generated by five tabular foundation models and six classical transformation methods. Correspondence differed substantially across methods and representation dimensionalities. Patient relationships were generally best preserved at intermediate dimensionalities, while stronger reduction or expansion introduced greater deviations from the original data structure. Foundation models preserved correspondence more consistently than classical methods, with improvements ranging from approximately 5% to over 125%. The proposed score provides a way to evaluate representation quality before downstream use and supports informed selection of representation methods and dimensionalities.
ER -