Skip to content

Predicting Cardiac and Respiratory Events in the PSG-Audio Dataset

License

Notifications You must be signed in to change notification settings

imics-lab/multi-output-psg

Repository files navigation

Multi-Output PSG (Polysomnography)

This is code for our work to build a multi-output model that predicts both Cardiac and Respiratory abnormal events in the PSG-Audio Dataset

We are using a processed version of the dataset with shared google drive links. Please see our load data time-series repository for code and details on the raw data conversion.

A paper outlining some of the challenges with two differently and highly imbalanced classes is under review.

Currently, we are working to integrate and tune sample_weights individually for the cardiac and respiratory events within the model.

About

Predicting Cardiac and Respiratory Events in the PSG-Audio Dataset

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published