Skip to content

Pneumonia classifier trained on x-ray data of pneumonia patients

Notifications You must be signed in to change notification settings

richieyoum/Pneumonia-diagnosis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 

Repository files navigation

Pneumonia diagnosis software

Pneumonia classifier aimed to automate the process of detecting pneumonia from patients.

demo

Getting started

To download the data, simply replace 'xxx' in your Kaggle API ID and Key. You can access this by

My Account -> Create New API Token -> Access Kaggle.json file for ID and Key

Testing your own images

If you happen to have your chest x-ray picture lying around, or just want to test the model on another chest x-ray picture, simply replace the images in ".chest-xray/val".

About the model

It uses transfer learning on top of the VGG-16 pretrained weights, with three fully-connected lyaers with dropout regularization and maxpooling. Please refer to this paper for more information on VGG-16. (https://arxiv.org/abs/1409.1556) vgg16

The trained model weights are available in "models" folder. The weights were taken from three different intervals of epochs, at around 35~38 epochs.

Results

demo1 demo2

Author

  • Richie Youm

Disclosure

Please note that this is a side project, and is not meant to replace your doctor's appointment. Seek professional medical help if you feel sick.

About

Pneumonia classifier trained on x-ray data of pneumonia patients

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published