Skip to content

A-EYE-George-Washington-University/Model

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

36 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Introduction

Multi-class image classifer for Fundus Opthamologic Medical Imagings

Detects the following disease:

  • Normal
  • Diabetic Retinopathy
  • Cataract
  • Glaucoma
  • Age Related Macular Degeneration
  • Patholical Myopia

Accuracy = .74

Confusion Matrix

e

ROC

e

Model Accuracy and Loss

e

Using model

from keras.models import Model, load_model
model= load_model(PATH/model.h5)

Demo to come

How to Run and Train Model Locally

training_model.ipynb

e

Download Jupyter Notebooks, Anaconda, Python 3

https://jupyter.org/install

https://www.anaconda.com/products/individual

https://www.python.org/downloads/

Have the following Python Libraries Installed

  • numpy
  • pandas
  • cv2
  • matplotlib
  • Tensorflow 2
  • Keras
  • IPython
  • PIL
  • sklearn
  • itertools

The model was trained and created in AWS Sagemaker. Please note comments to make changes to run locally.

Examples of GradCam Images

Normal Image

e

Diabetic Retinopathy Image

e

Glaucoma Image

e

Website

e

Data

Source Diease
https://www.kaggle.com/jr2ngb/cataractdataset Cataracts, Glaucoma
https://www.kaggle.com/linchundan/fundusimage1000 Myopia, AMD
https://www.kaggle.com/nguyenhung1903/diaretdb1-v21 Diabetic Retinopathy
http://www5.informatik.uni-erlangen.de/research/data/fundus-images/ Diabetic Retinopathy, Glaucoma
https://www.it.lut.fi/project/imageret/diaretdb0/#DOWNLOAD Diabetic Retinopathy
https://personalpages.manchester.ac.uk/staff/niall.p.mcloughlin/ Glaucoma, Diabetic Retinopathy, AMD
https://ai.baidu.com/broad/download AMD, Glaucoma, Myopia
https://www.kaggle.com/andrewmvd/ocular-disease-recognition-odir5k Diabetic Retinopathy, Normal, Cataract, Glaucoma, AMD, Myopia
Logo Design- Jody Serani