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

TL;DR

This paper introduces CV-Masking, a volatility-aware masking strategy for EHR models that adaptively emphasizes volatile features, leading to better representations and predictive performance.

Contribution

The paper proposes a novel volatility-aware masking method for EHR pretraining that accounts for feature heterogeneity, improving model robustness and clinical relevance.

Findings

CV-Masking improves reconstruction accuracy.

It enhances downstream predictive performance.

It accelerates model convergence.

Abstract

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.