abhiksark / diabetic-retnopathy-classification-convolutionalneuralnetwork Goto Github PK

Diabetic retinopathy is the leading cause of blindness in the working-age population of the developed world. The condition is estimated to affect over 93 million people.

The need for a comprehensive and automated method of diabetic retinopathy screening has long been recognized, and previous efforts have made good progress using image classification, pattern recognition, and machine learning. With photos of eyes as input, the goal of this capstone is to create a new model, ideally resulting in realistic clinical potential.

The motivations for this project are twofold:

• Image classification has been a personal interest for years, in addition to classification on a large scale data set.

• Time is lost between patients getting their eyes scanned (shown below), having their images analyzed by doctors, and scheduling a follow-up appointment. By processing images in real-time, EyeNet would allow people to seek & schedule treatment the same day.

1. Data
2. Extraction and Preprocessing
   • Download Images to Google Colab
   • Resize Images
   • Checking Blurness of Images
   • Data Augmentation
3. CNN Architecture
4. Results
5. References
6. Authors

The data originates from a 2015 Kaggle competition. However, is an atypical Kaggle dataset. In most Kaggle competitions, the data has already been cleaned, giving the data scientist very little to preprocess. With this dataset, this isn't the case.

All images are taken of different people, using different cameras, and of different sizes. Pertaining to the preprocessing section, this data is extremely noisy, and requires multiple preprocessing steps to get all images to a useable format for training a model.

The training data is comprised of 35,126 images, which are augmented during preprocessing.

You'll need to install:

• Anaconda
• Python (Minimum 3)
• pandas
• Seaborn
• Fast.ai

The preprocessing pipeline is the following:

Google Colab was used as a platform for this dataset.

The given images were mostly very large. ~2000 by ~2000. So to keep things unifrom all images were scaled down to 512 by 512.

The method is simple. Straightforward. Has sound reasoning. And can be implemented in only a single line of code:

cv2.Laplacian(image, cv2.CV_64F).var()

We simply take a single channel of an image and convolve it with the following 3 x 3 kernel:

And then take the variance (i.e. standard deviation squared) of the response. If the variance falls below a pre-defined threshold, then the image is considered blurry; otherwise, the image is not blurry. Here is the paper with talks about it's implementation, Ariation of the Laplacian by Pech-Pacheco et al. in their 2000 ICPR paper, Diatom autofocusing in brightfield microscopy: a comparative study.

All images were rotated and mirrored.Images without retinopathy were mirrored; images that had retinopathy were mirrored, and rotated 90, 120, 180, and 270 degrees.

The first images show two pairs of eyes, along with the black borders. Notice in the cropping and rotations how the majority of noise is removed.

First Stage

Second Stage

Third Stage

Confusion Matrix

1. What is Diabetic Retinopathy?
2. Blur Detection

• Abhik Sarkar
• Fastai

• Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License