# PhysioNet/CinC Challenge 2016: Classification of Heart Sound Recordings

## Facts

- Provider: Chengyu Liu, David Springer, Benjamin Moody, Ikaro Silva, Alistair Johnson, Maryam Samieinasab, Reza Sameni, Roger Mark, Gari D. Clifford
- License: ODC-By-1.0 (https://physionet.org/content/challenge-2016/1.0.0/)
- Format: CSV (in sampled files)
- Stored at: Amazon S3, s3://physionet-open/challenge-2016 (us-east-1)
- Access: public, readable without an account
- Category: Physiology
- Source: https://physionet.org/content/challenge-2016/1.0.0/
- Page: https://scigantic.com/archive/62eeb15a-3598-5c48-aa91-9705f831b391
- Keywords: Physiology, image segmentation, CSV, AWS Open Data

## About

The 2016 PhysioNet/Computing in Cardiology Challenge asked for algorithms that classify phonocardiogram (PCG) heart sound recordings as normal or abnormal, using data collected in clinical and nonclinical settings (such as in-home visits) from healthy subjects and patients with confirmed cardiac diagnoses, children and adults. The training set combines databases contributed from several groups worldwide, with 3,126 heart sound recordings lasting from 5 seconds to just over 120 seconds, all resampled to 2,000 Hz and provided as .wav files. Measured 2026-11-13: training-a/ to training-f/ and validation/ folders; training-a/ alone holds 409 .wav recordings with matching .hea headers, 405 .dat files and RECORDS-normal and RECORDS-abnormal lists. an annotations/ folder is present, and training.zip, validation.zip, annotations.zip and sample entries are at the top level with sources/ and the papers. Read audio with scipy.io.wavfile or soundfile and the WFDB headers with the wfdb Python package. The largest public PCG collection at release, used for heart sound classification, murmur detection, segmentation and cardiac auscultation research. Released under the Open Data Commons Attribution License v1.0; 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='challenge-2016/', Delimiter='/')
```

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

## Data structure

File types in sampled files: CSV

Example file: `challenge-2016/1.0.0/annotations/Online_Appendix_Diagnosis_meanings.csv`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/62eeb15a-3598-5c48-aa91-9705f831b391
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/62eeb15a-3598-5c48-aa91-9705f831b391
- Over MCP: `get_schema_card` with id `62eeb15a-3598-5c48-aa91-9705f831b391` at https://api.scigantic.com/mcp

## Related topics

- [CSV datasets](https://scigantic.com/topics/csv)
- [Image segmentation datasets](https://scigantic.com/topics/image-segmentation)
- [Physiology datasets](https://scigantic.com/topics/physiology)
