/xray-pneumonia-detection

xray pneumonia detection

Primary LanguageJupyter NotebookMIT LicenseMIT

xray-pneumonia-detection

xray-pneumonia-detection based on https://github.com/llSourcell/AI_Startup_Prototype

Please do not use this code in real projects,

I am not a medical doctor, this is just an image processing training

RUN

  1. open https://colab.research.google.com/
  2. copy https://github.com/DenisSouth/xray-pneumonia-detection/blob/master/chest_xray_pneumonia.ipynb
  3. change runtime to GPU
  4. make all steps in notebook one by one

p.s.

I am not sure about original layers dense, i think its overfit the network

x=Dense(1024,activation='relu')(x)

x=Dense(1024,activation='relu')(x)

x=Dense(512,activation='relu')(x)

Try to decrease it to

x=Dense(256,activation='relu')(x)

x=Dense(256,activation='relu')(x)

or some else

If you will get nice result on test data (better then 0.69% accuracy) - please make pull request or issue :-)