Description:The figure describes the Convolution Neural Network Architecture used in this project. This architecture of the Deep Learning Model contains four Convolution Layers, four Pooling layers and one flattening layer.

Convolution Layer: This the layer which is responsible for feature extraction with the help of linear and non-linear operations called convolution operations and activation functions.

Pooling Layer: This causes downsampling of the output from convolution layers by reducing it’s in-plane dimentionality.

Both of these layers have a few hyperparameters named as filters, strides and padding. The input, denoised, normalized signal of dimentionality 360X1 is passed to the model which is passed through convolution and pooling layers alternatively and then finally the data is flattened so that the class of output can be identified. Hence the class of the model is returned to the app which shows it to the user in their profile.
, Claims:We Claim:
1. Our model is claimed to be extremely light and pocket sized

2. Device is completely wireless and makes handling very easy.

3. Our companion app helps to store share and analyse data of the user

4. The user has the space to view the report and will get personalised recommendations.

5. The user can also preview their report and can get a shareable format.

6. The app provides a prediction of beat type based on the ECG reading along with the signal format that can be shown to the doctor easily.