ECG Beat Classification Model for Edge Computing

A machine learning model for classifying cardiac arrhythmias from raw 1D ECG signals. It incorporates spatial and temporal feature extraction and is designed to reduce training costs and hardware needs by avoiding computationally expensive signal-to-image conversions. This makes it suitable for edge computing, wearable health, and real-time applications.

Use Cases

• Real-time arrhythmia detection based on raw 1D ECG signal processing
• Model optimization for wearable health devices based on reduced computational requirements
• Benchmarking edge-computing models against signal-to-image conversion methods

Strengths

• Model is shared under a permissive CC-BY-4.0 license
• Designed for computational efficiency, targeting reduced hardware needs
• Cross-platform presence on figshare suggests established sharing

Limitations

• No column or sample data provided to verify model inputs or outputs
• Dataset size (669 MB) is listed but row count and specific model architecture are unknown
• Future update date (2026-03-21) is inconsistent with current timelines

Provenance

Source

Phaneendra Puppala

Collection Method

Model development and sharing, method not specified

Time Range

null

Freshness

2026-03-21 04:56:21

Geography

null