Coefficient of Variation Masking: A Volatility-Aware Strategy for EHR Foundation Models
This addresses the need for more effective pretraining strategies in EHR foundation models to better handle volatile biomarkers that signal acute conditions, though it is incremental as it builds on existing masked autoencoder approaches.
The paper tackled the problem of uniform random masking in masked autoencoders for electronic health records, which fails to account for the varying volatility of biomarkers, by proposing Coefficient of Variation Masking (CV-Masking) to adjust masking probabilities based on feature variability. The result was systematic improvements in reconstruction, downstream predictive performance, and convergence, producing more robust and clinically meaningful representations.
Masked autoencoders (MAEs) 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.