This repro contains the Python scripts and other tools to access the data in the CACHET Contextualised Arrhythmia Database (CACHET-CADB)
CACHET-CADB is 259 days long contextualised single-channel ECG arrhythmia database from 24 patients recording under a free-living ambulatory setting. Along with the ECG, it also contains contextual information such as activities, body position, movement acceleration, patients reported unusual symptoms/events diary, step counts, patient-reported sleep quality, stress level, and food intake. I contain 1602 ten-second long ECG samples of AF, NSR, noise, and other rhythm classes, which are manually annotated by two cardiologists. The ECG is sampled at 1024 Hz and a 12-bit resolution.
The CACHET-CADB is available from DTU Data in two formats:
Python libraries for reading Unisens file format:
pip install git+https://github.com/Unisens/pyunisens
Read more about Unisens file format for simultaneous and synchronous storage of multi-sensor data at different sample rates.
The UnisensViewer is an application which can be used to view and browse the raw unisense files. Note that this tool is only avalable on Windows.
We provide the CACHET-CADB-NoteBook.ipynb as an illustration on how to use the CACHET-CADB data in Python.
The cachet-cadb_analysisMain.py Python script contains funtions to read the database and all the statistics.
Please read about existing issues and file new feature requests and bug reports at the issue tracker.
This software is copyright (c) Copenhagen Center for Health Technology (CACHET) at the Technical University of Denmark (DTU). This software is available 'as-is' under a MIT license.
When using the CACHET-CADB dataset and code, please cite its data descriptor article :
@article{kumar2022cachet,
title={CACHET-CADB: A Contextualized Ambulatory Electrocardiography Arrhythmia Dataset},
author={Kumar, Devender and Puthusserypady, Sadasivan and Dominguez, Helena and Sharma, Kamal and Bardram, Jakob E},
journal={Frontiers in Cardiovascular Medicine},
pages={1702},
year={2022},
publisher={Frontiers}
}