Coefficient of Variation Masking: A Volatility-Aware Strategy for EHR Foundation Models

Rajna Fani, Rafi Al Attrach, David Restrepo, Yugang Jia, Leo Anthony Celi, Peter Schüffler
Proceedings of the Fifth Machine Learning for Health Symposium, PMLR 297:1376-1391, 2026.

Abstract

Masked autoencoders ({MAE}s) are increasingly applied to electronic health records ({EHR}) for learning general-purpose representations that support diverse clinical tasks. However, existing approaches typically rely on uniform random masking, implicitly assuming all features are equally predictable. In reality, laboratory tests exhibit substantial heterogeneity in volatility: some biomarkers (e.g., sodium) remain stable, while others (e.g., lactate) fluctuate considerably and are more difficult to model. Clinically, volatile biomarkers often signal acute pathophysiology and require more sophisticated modeling to capture their complex temporal patterns. We propose a volatility-aware pretraining strategy, Coefficient of Variation Masking ({CV}-Masking), that adaptively adjusts masking probabilities according to the intrinsic variability of each feature. Combined with a value-only masking objective aligned with clinical workflows, {CV}-Masking yields systematic improvements over random and variance-based strategies. Experiments on a large panel of laboratory tests show that {CV}-Masking enhances reconstruction, improves downstream predictive performance, and accelerates convergence, producing more robust and clinically meaningful {EHR} representations.

Cite this Paper

BibTeX
@InProceedings{pmlr-v297-fani26a, title = {Coefficient of Variation Masking: A Volatility-Aware Strategy for {EHR} Foundation Models}, author = {Fani, Rajna and Al Attrach, Rafi and Restrepo, David and Jia, Yugang and Celi, Leo Anthony and Sch{\"u}ffler, Peter}, booktitle = {Proceedings of the Fifth Machine Learning for Health Symposium}, pages = {1376--1391}, year = {2026}, editor = {Argaw, Peniel and Zhang, Haoran and Jabbour, Sarah and Chandak, Payal and Ji, Jerry and Mukherjee, Sumit and Salaudeen, Olawale and Chang, Trenton and Healey, Elizabeth and Gröger, Fabian and Adibi, Amin and Hegselmann, Stefan and Wild, Benjamin and Noori, Ayush}, volume = {297}, series = {Proceedings of Machine Learning Research}, month = {13--14 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v297/main/assets/fani26a/fani26a.pdf}, url = {https://proceedings.mlr.press/v297/fani26a.html}, abstract = {Masked autoencoders ({MAE}s) are increasingly applied to electronic health records ({EHR}) for learning general-purpose representations that support diverse clinical tasks. However, existing approaches typically rely on uniform random masking, implicitly assuming all features are equally predictable. In reality, laboratory tests exhibit substantial heterogeneity in volatility: some biomarkers (e.g., sodium) remain stable, while others (e.g., lactate) fluctuate considerably and are more difficult to model. Clinically, volatile biomarkers often signal acute pathophysiology and require more sophisticated modeling to capture their complex temporal patterns. We propose a volatility-aware pretraining strategy, Coefficient of Variation Masking ({CV}-Masking), that adaptively adjusts masking probabilities according to the intrinsic variability of each feature. Combined with a value-only masking objective aligned with clinical workflows, {CV}-Masking yields systematic improvements over random and variance-based strategies. Experiments on a large panel of laboratory tests show that {CV}-Masking enhances reconstruction, improves downstream predictive performance, and accelerates convergence, producing more robust and clinically meaningful {EHR} representations.} }
Endnote
%0 Conference Paper %T Coefficient of Variation Masking: A Volatility-Aware Strategy for EHR Foundation Models %A Rajna Fani %A Rafi Al Attrach %A David Restrepo %A Yugang Jia %A Leo Anthony Celi %A Peter Schüffler %B Proceedings of the Fifth Machine Learning for Health Symposium %C Proceedings of Machine Learning Research %D 2026 %E Peniel Argaw %E Haoran Zhang %E Sarah Jabbour %E Payal Chandak %E Jerry Ji %E Sumit Mukherjee %E Olawale Salaudeen %E Trenton Chang %E Elizabeth Healey %E Fabian Gröger %E Amin Adibi %E Stefan Hegselmann %E Benjamin Wild %E Ayush Noori %F pmlr-v297-fani26a %I PMLR %P 1376--1391 %U https://proceedings.mlr.press/v297/fani26a.html %V 297 %X Masked autoencoders ({MAE}s) are increasingly applied to electronic health records ({EHR}) for learning general-purpose representations that support diverse clinical tasks. However, existing approaches typically rely on uniform random masking, implicitly assuming all features are equally predictable. In reality, laboratory tests exhibit substantial heterogeneity in volatility: some biomarkers (e.g., sodium) remain stable, while others (e.g., lactate) fluctuate considerably and are more difficult to model. Clinically, volatile biomarkers often signal acute pathophysiology and require more sophisticated modeling to capture their complex temporal patterns. We propose a volatility-aware pretraining strategy, Coefficient of Variation Masking ({CV}-Masking), that adaptively adjusts masking probabilities according to the intrinsic variability of each feature. Combined with a value-only masking objective aligned with clinical workflows, {CV}-Masking yields systematic improvements over random and variance-based strategies. Experiments on a large panel of laboratory tests show that {CV}-Masking enhances reconstruction, improves downstream predictive performance, and accelerates convergence, producing more robust and clinically meaningful {EHR} representations.
APA
Fani, R., Al Attrach, R., Restrepo, D., Jia, Y., Celi, L.A. & Schüffler, P.. (2026). Coefficient of Variation Masking: A Volatility-Aware Strategy for EHR Foundation Models. Proceedings of the Fifth Machine Learning for Health Symposium, in Proceedings of Machine Learning Research 297:1376-1391 Available from https://proceedings.mlr.press/v297/fani26a.html.

Related Material