Label-Conditioned Cross-Modal Fusion for Adult-to-Pediatric ECG Transfer via Curriculum-Gated Contrastive Alignment
arXiv:2605.00647v2 Announce Type: replace Abstract: Automated pediatric electrocardiogram (ECG) interpretation remains challenging because developmental differences in heart rate, intervals, and waveforms limit the transferability of models trained mainly on adult data, while expert-labeled pediatric ECG cohorts are scarce. We propose PEACE (Pediatric-Adult ECG Alignment via Cross-modal Enhancement), an adult-to-pediatric ECG transfer framework pretrained on MIMIC-IV ECGs and adapted to pediatric targets. PEACE integrates label-specific bidirectional contrastive learning (LSBC) to align ECG re
The scarcity of specialized pediatric medical data and the advancements in AI for transfer learning present a timely opportunity to address critical gaps in healthcare diagnostics.
This development enables more accurate and automated pediatric ECG interpretation, which can significantly improve diagnostic capabilities in a field with limited expert resources.
Current limitations in applying adult-trained AI models to pediatric ECGs are being overcome, leading to more accessible and reliable pediatric heart condition diagnosis.
- · Pediatric healthcare providers
- · Medical AI developers
- · Patients with pediatric heart conditions
- · Hospitals and clinics
- · Manual diagnostic processes
- · Companies reliant on outdated diagnostic methods
Improved early detection rates for pediatric heart abnormalities.
Reduced healthcare costs associated with misdiagnosis or delayed diagnosis in pediatric cardiology.
Potential for broader applicability of similar transfer learning techniques to other data-scarce medical specialties, creating a new paradigm for medical AI development.
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