Semi-supervised-DL-method-for-Retinopathy-Automatic-Detection

Table of Contents

• About the Project
• Getting Started
  • Prerequisites
  • Package used
• Usage
• License

About The Project

The paper: Automatic Detection of Retinopathy with Optical Coherence Tomography Images via a Semi-supervised Deep Learning Method

A semi-supervised deep learning method for retinopathy detection with OCT images is proposed.

Getting Started

To get a local copy and runn following simple steps.

Prerequisites

Download the dataset from below links and unzip the data files into folder dataset.

• BOE
• CELL

Package used

 torch              1.6.0+cu101
 scikit-learn       0.21.3
 numpy              1.18.1

Usage

1. Split the data into four subsets: train, validation, test and unlabel. The proposed method will use samples from train and unlabel folders for model training

python BOE_dataset_split.py

python CELL_dataset_split.py

2. Run the model training with the selected dataset

python VAT_semiDL_train.py -d 'BOE'

License

Distributed under the MIT License. See LICENSE for more information.