# CPSC 2021: Paroxysmal Atrial Fibrillation Events Detection from Dynamic ECG Recordings

## Facts

- Provider: Xingyao Wang, Caiyun Ma, Xiangyu Zhang, Hongxiang Gao, Gari D. Clifford, Chengyu Liu (China Physiological Signal Challenge)
- License: CC-BY-4.0 (https://physionet.org/content/cpsc2021/1.0.0/)
- Stored at: Amazon S3, s3://physionet-open/cpsc2021 (us-east-1)
- Access: public, readable without an account
- DOI: https://doi.org/10.13026/ksya-qw89
- Category: Physiology
- Source: https://physionet.org/content/cpsc2021/1.0.0/
- Page: https://scigantic.com/archive/e70ef41e-4b6c-5b57-a1d7-5c40deb49e64
- Keywords: Physiology, ECG, AWS Open Data

## About

The 4th China Physiological Signal Challenge 2021 (CPSC 2021) asked for algorithms that locate paroxysmal atrial fibrillation (PAF) episodes, giving onset and offset points, in long-term dynamic ECG records containing AF episodes, partial AF or non-AF rhythms. Data were recorded with 12-lead Holter or 3-lead wearable ECG monitors; the challenge provides variable-length records extracted from leads I and II, each sampled at 200 Hz. The first-stage training set consists of 730 records from 12 AF patients (5 PAF) and 42 non-AF patients, and the second-stage set of 706 records from 37 AF patients (18 PAF) and 14 non-AF patients. All data are in WFDB format with annotations following PhysioBank conventions. Measured 2026-11-13: Training_set_I/ lists 719 records, each as .hea, .dat and .atr; Training_set_II/ holds the second stage, and python_entry/, matlab_entry/, RECORDS and README.md sit at the top level. Read signals and rhythm annotations with the wfdb Python package (wfdb.rdrecord, wfdb.rdann). Useful for AF episode detection, Holter ECG rhythm analysis and wearable ECG arrhythmia research. Released under CC BY 4.0; cite the challenge paper and the standard PhysioNet reference (Goldberger et al., 2000) when using the data.

## Load the data

```python
import boto3
from botocore import UNSIGNED
from botocore.client import Config
s3 = boto3.client('s3', region_name='us-east-1', config=Config(signature_version=UNSIGNED))
s3.list_objects_v2(Bucket='physionet-open', Prefix='cpsc2021/', Delimiter='/')
```

Lists the files with anonymous access. Runs on any machine, with no cloud account.

## Data structure

Example file: `cpsc2021/1.0.0/Training_set_I/data_0_1.atr`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/e70ef41e-4b6c-5b57-a1d7-5c40deb49e64
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/e70ef41e-4b6c-5b57-a1d7-5c40deb49e64
- Over MCP: `get_schema_card` with id `e70ef41e-4b6c-5b57-a1d7-5c40deb49e64` at https://api.scigantic.com/mcp

## Related topics

- [Physiology datasets](https://scigantic.com/topics/physiology)
- [ECG datasets](https://scigantic.com/topics/ecg)
