README.md

---
title: SSL ECG Classification
emoji: 🌖
colorFrom: green
colorTo: indigo
sdk: gradio
sdk_version: 6.12.0
app_file: app.py
pinned: false
license: apache-2.0
---

# ECG Classification with Self-Supervised Learning

[![DOI](https://img.shields.io/badge/DOI-10.57967%2Fhf%2F8469-blue?style=flat-square)](https://doi.org/10.57967/hf/8469)

**Interactive Space** for ECG cardiovascular disease classification using SimCLR pre-trained model fine-tuned on PTB-XL.

## Quick Start

Upload an ECG file in any format (DICOM, MATLAB, HDF5, WFDB, CSV, and more) and get instant predictions across 5 diagnostic classes.

## Model Performance

- **AUROC:** 0.8717
- **Accuracy:** 0.8234
- **Training Data:** PTB-XL with 10% labeled samples (1,747 ECGs)

## Supported File Formats

- **Clinical:** DICOM (.dcm), SCP-ECG (.scp), HL7 aECG (.xml)
- **Research:** WFDB (.hea/.dat), EDF (.edf)
- **Scientific:** MATLAB (.mat), HDF5 (.h5), NumPy (.npy)
- **Generic:** CSV/TXT/TSV, Binary (.raw, .bin)

## Citation

If you use this model, please cite it using the DOI:

```
10.57967/hf/8469
```

**Bibtex:**
```bibtex
@model{ssrl_ecg_2024,
  title={SSL ECG Classification},
  doi={10.57967/hf/8469},
  url={https://huggingface.co/spaces/Tumo505/SSL-ECG-Classification},
  year={2024}
}
```

## Resources

- [Model Card](https://huggingface.co/Tumo505/SSL-ECG-Classification-model-card)
- [GitHub Repository](https://github.com/Tumo505/SSL-for-ECG-classification)
- [Model Weights](https://huggingface.co/Tumo505/SSL-ECG-Classificcation)