DeepBeat is a deep learning model for
- classification of Atrial Fibrillation (AF) from photoplethysmography (PPG) signals and
- signal quality estimation
We share a deidentified train, validate and test partition that were used for training DeepBeat. These deidentified signal are shared with a non-commerical data use agreement and can be found at synapase, ID:syn21985690.
We have included juypter notebooks and trained models for reproduciblility.