Dynamical Label Augmentation and Calibration for Noisy Electronic Health Records

TL;DR

This paper introduces ACTLL, an attention-based framework that dynamically calibrates and augments labels in noisy EHR time series data, significantly improving patient outcome predictions despite label errors.

Contribution

The paper presents a novel ACTLL framework combining a Beta mixture model with attention mechanisms to handle noisy labels in EHR data, achieving state-of-the-art results.

Findings

Achieves superior performance on large-scale EHR datasets.

Effectively handles high noise levels in label data.

Outperforms existing methods in noisy time series classification.

Abstract

Medical research, particularly in predicting patient outcomes, heavily relies on medical time series data extracted from Electronic Health Records (EHR), which provide extensive information on patient histories. Despite rigorous examination, labeling errors are inevitable and can significantly impede accurate predictions of patient outcome. To address this challenge, we propose an \textbf{A}ttention-based Learning Framework with Dynamic \textbf{C}alibration and Augmentation for \textbf{T}ime series Noisy \textbf{L}abel \textbf{L}earning (ACTLL). This framework leverages a two-component Beta mixture model to identify the certain and uncertain sets of instances based on the fitness distribution of each class, and it captures global temporal dynamics while dynamically calibrating labels from the uncertain set or augmenting confident instances from the certain set. Experimental results on large-scale EHR datasets eICU and MIMIC-IV-ED, and several benchmark datasets from the UCR and UEA repositories, demonstrate that our model ACTLL has achieved state-of-the-art performance, especially under high noise levels.