LGNov 1, 2024
Hierarchical Transformer for Electrocardiogram Diagnosis
arXiv:2411.00755v25 citationsh-index: 3
Originality Synthesis-oriented
AI Analysis
This work addresses ECG analysis for medical diagnostics, presenting an incremental improvement in model design.
The authors tackled ECG diagnosis by proposing a hierarchical Transformer that integrates depth-wise convolutions and attention-gated modules, achieving a lightweight and interpretable model without complex attention or downsampling.
We propose a hierarchical Transformer for ECG analysis that combines depth-wise convolutions, multi-scale feature aggregation via a CLS token, and an attention-gated module to learn inter-lead relationships and enhance interpretability. The model is lightweight, flexible, and eliminates the need for complex attention or downsampling strategies.