README.md

---
license: mit
task_categories:
- translation
tags:
- biology
size_categories:
- 100B<n<1T
---

# Dataset Card for Panobench

<!-- Provide a quick summary of the dataset. -->
Panobench: A dense benchmark for Electrocardio Panorama
To enable comprehensive evaluation of panoramic ECG view synthesis, we curated a new benchmark, Panobench, which for the first time extends beyond previous datasets limited to only 8 or 12 views. Panobench comprises 4,470 ten-second recordings with 48 viewpoints (6 limb and 42 precordial leads), each annotated with CT-derived spherical coordinates (θ, φ). By expanding to 48 leads (views), Panobench provides a high-resolution representation of cardiac electrical dynamics, enabling signal analysis from diverse perspectives. The 48-viewpoint signals and their corresponding angular positions are as follows:
 ![Panobench](./Figures/Figure4.png). 


## Citation
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
Please cite the paper if the codes or the dataset we provided are helpful:  

> ```
>@inproceedings{zhan2025nef,
> title={NEF-NET+: Adapting Electrocardio panorama in the wild},
> author={Zhan, Zehui and Hu, Yaojun and Zhan, Jiajing and Lian, Wanchen and Wu, Wanqing and Chen, Jintai},
> journal={arXiv preprint arXiv:2511.02880},
> year={2025}}
> ```