Skip to content

Latest commit

 

History

History
33 lines (17 loc) · 1.34 KB

README.md

File metadata and controls

33 lines (17 loc) · 1.34 KB

License: MIT Website shields.io

Links to model notebooks

2D CNN   1D CNN

ECG

We classify various cardiovascular conditions from Electrocardiogram (ECG) images [1]. We also study the ECG of COVID-19 patients to identify potential cardiac injury due to SARS CoV 2. We perform both image and time series classification. Our 1D CNN model using time series achieves 95 % accuracy in classifying cardiac disorders including COVID-19.

Application

https://bestecgclassifier.herokuapp.com/

Preprocessing

alt text

Best Performance (95 % Accuracy)

Time series classification using 1D CNN
alt text