Source code & Pretrained model for our IMWUT (UbiComp) 2022 paper: "BayesBeat: A Bayesian Deep Learning Approach for Atrial Fibrillation Detection from Noisy Photoplethysmography (PPG) Signals" [preprint]
The Pretrained pytorch model file for CPU is provided in saved_model folder
CUDA version 10.2+
Python version 3.7+
PyTorch version 1.5.1+
- First, setup a virtual environment and activate it
- Install all the requirements and their dependencies
- Then download the dataset and put that into proper folder structure
- Finally, run
python evaluate_bayesian.py - For the version that utilizes gpu to evaluate, please refer to this repo: https://github.com/Subangkar/BayesBeat
Data Folder Structure for running evaluate_bayesian.py:
data/
test/
signal.npy
qa_label.npy
rhythm.npy
distr_split_ids.npy: A dictionary that contains list of individal ids for train, validation & test set for the distribution of dataset
If you use our work, please cite:
@article{das2022bayesbeat,
title={BayesBeat: Reliable Atrial Fibrillation Detection from Noisy Photoplethysmography Data},
author={Das, Sarkar Snigdha Sarathi and Shanto, Subangkar Karmaker and Rahman, Masum and Islam, Md Saiful and Rahman, Atif Hasan and Masud, Mohammad M and Ali, Mohammed Eunus},
journal={Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies},
volume={6},
number={1},
pages={1--21},
year={2022},
publisher={ACM New York, NY, USA}
}