SafeECGMatch: Calibration-Aware Joint Frequency and Time Space Semi-Supervised Learning for Open-Set ECG Classification
arXiv:2606.08037v1 Announce Type: new Abstract: Electrocardiogram (ECG) classification models often suffer from severe label scarcity, making semi-supervised learning (SSL) an attractive strategy for reducing annotation costs. In clinical settings, however, unlabeled pools frequently contain out-of-distribution (OOD) anomalies or diagnostic groups absent from the labeled set. Standard SSL forces incorrect pseudo-labels onto these unseen classes, producing overconfident predictions. To address this, we propose SafeECGMatch, a calibration-aware safe SSL framework for single-label ECG classificat
The continuous growth in AI applications, particularly in healthcare, drives ongoing research into more robust and reliable machine learning methods like semi-supervised learning.
Improving the robustness and safety of AI in critical applications like medical diagnostics by addressing issues like out-of-distribution data is crucial for widespread adoption and trust.
This research introduces a method to make semi-supervised ECG classification more reliable in clinical settings by mitigating issues with unseen anomaly classes.
- · AI healthcare developers
- · Patients receiving AI-assisted diagnoses
- · Medical research institutions
- · Traditional, less robust AI classification models
More accurate and trustworthy AI models for medical diagnostics, particularly in cardiology.
Increased integration of AI into clinical decision-making processes for screening and early detection.
Potential for reduced diagnostic errors and improved patient outcomes in underserved regions with limited specialist access.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG