Position: Evaluation of ECG Representations Must Be Fixed

TL;DR

This paper emphasizes the need to improve ECG benchmarking practices by expanding evaluation criteria, adopting best practices, and recognizing that random encoders can serve as strong baselines, thereby enhancing clinical relevance and reliability.

Contribution

It highlights the limitations of current ECG benchmarks, proposes expanded evaluation standards, and demonstrates that random encoders can outperform pre-trained models in some settings.

Findings

Current benchmarks focus mainly on arrhythmia and waveform morphology.

Applying best practices changes the perceived performance of ECG representations.

Random encoders with linear evaluation can match or outperform pre-trained models.

Abstract

This position paper argues that current benchmarking practice in 12-lead ECG representation learning must be fixed to ensure progress is reliable and aligned with clinically meaningful objectives. The field has largely converged on three public multi-label benchmarks (PTB-XL, CPSC2018, CSN) dominated by arrhythmia and waveform-morphology labels, even though the ECG is known to encode substantially broader clinical information. We argue that downstream evaluation should expand to include an assessment of structural heart disease and patient-level forecasting, in addition to other evolving ECG-related endpoints, as relevant clinical targets. Next, we outline evaluation best practices for multi-label, imbalanced settings, and show that when they are applied, the literature's current conclusion about which representations perform best is altered. Furthermore, we demonstrate the surprising result that a randomly initialized encoder with linear evaluation matches state-of-the-art pre-training on many tasks. This motivates the use of a random encoder as a reasonable baseline model. We substantiate our observations with an empirical evaluation of three representative ECG pre-training approaches across six evaluation settings: the three standard benchmarks, a structural disease dataset, hemodynamic inference, and patient forecasting.