Search papers, labs, and topics across Lattice.
This paper introduces SCAR, a robust ECG-language pretraining framework that uses adversarial masking to improve zero-shot ECG diagnosis under partial observation. SCAR trains the ECG encoder to maintain semantic alignment with clinical text even when critical spatio-temporal tokens are missing, using a differentiable adversarial masker and a semantically supervised adaptive selector. Experiments across six datasets demonstrate that SCAR enhances semantic robustness and linear-probing transferability, especially when primary diagnostic evidence is absent, as measured by a new Counterfactual Missingness Resolution Score (CMRS).
Even when key ECG leads or segments are missing, SCAR enables robust zero-shot diagnosis by forcing models to compensate for lost information using secondary morphological cues.
Recent ECG--language pretraining methods enable zero-shot diagnosis by aligning cardiac signals with clinical text, but they do not explicitly model robustness to partial observation and are typically studied under fully observed ECG settings. In practice, diagnostically critical leads or temporal segments may be missing due to electrode detachment, motion artifacts, or signal corruption, causing severe degradation of cross-modal semantic alignment. In this paper, we propose \textbf{SCAR}, a robust ECG--language pretraining framework for \textbf{S}emantic \textbf{C}ompensation via \textbf{A}dversarial \textbf{R}emoval. SCAR improves robustness by explicitly training the model to remain semantically aligned with semantically critical missingness and to recover diagnostic meaning from the remaining visible evidence. Specifically, we introduce a differentiable adversarial masker to remove the most alignment-critical spatio-temporal ECG tokens during training, forcing the ECG encoder to learn representations that remain semantically aligned with clinical text even when primary diagnostic evidence is missing. Under such adversarial corruption, we equip the ECG encoder with a semantically supervised adaptive selector that learns to reweight the remaining visible tokens and compensate with secondary yet diagnostically informative morphological cues. To evaluate robustness beyond classification accuracy, we further introduce Counterfactual Missingness Resolution Score (CMRS), which quantifies how well feature preserve diagnostic semantics under missingness. Experiments on $6$ datasets show that SCAR consistently improves semantic robustness under joint lead and temporal missingness, with particularly clear advantages in harder cases where primary diagnostic evidence is unavailable, while also yielding stronger linear-probing transferability.
