Diabetic Retinopathy Detection System uses, computer vision technique, deep learning model Densenet to classify the diabetic retinopathy severity on the left and right images provided. This application built on Python, Django, SQLite, and Keras deep learning model Densenet. The model has been trained on the Kaggle Diabetic Retinopathy dataset.
Diabetic Retinopath Detection System
Deep Learning has found wider application across multiple domains, and in this project applied the same concepts to solve the problem from the health care domain. This web application project built on Python, Django, and Deep learning algorithm (Convolutional Neural Network: Keras Deep Learning Model Densenet).
Diabetic Retinopathy is the leading cause of blindness in the working-age population of the developed world. It is estimated to affect over 93 million people.
DR is a complication of diabetes and a leading cause of blindness in the United States (U.S.). The retina is the membrane that covers the back of the eye. It is highly sensitive to light. It converts any light that hits the eye into signals that can be interpreted by the brain. This process produces visual images, and it is how sight functions in the human eye.
Diabetic Retinopathy damages the blood vessels within the retinal tissue, causing them to leak fluid and distort vision.
Comparison of Normal Vision Vs Diabetic Retinopathy Vision
Diabetic Retinopathy Detection using Deep Learning
Currently, detecting DR is a time-consuming and manual process that requires a trained clinician to examine and evaluate digital color fundus photographs of the retina. By the time human readers submit their reviews, often a day or two later, the delayed results lead to lost follow up, miscommunication, and delayed treatment.
Clinicians can identify DR by the presence of lesions associated with the vascular abnormalities caused by the disease. While this approach is effective, its resource demands are high. The expertise and equipment required are often lacking in areas where the rate of diabetes in local populations is high and DR detection is most needed. As the number of individuals with diabetes continues to grow, the infrastructure needed to prevent blindness due to DR will become even more insufficient.
The need for a comprehensive and automated method of DR screening has long been recognized, and previous efforts have made good progress using image classification, pattern recognition, and machine learning.
Convolutional Neural Network: Keras Densenet Model
Densely Connected Convolutional Networks or Densenet architecture explicitly differentiates between information that is added to the network and information that is preserved. DenseNet layers are narrow, therefore adding a small set of feature-maps to the “collective knowledge” of the network and keep the remaining feature maps unchanged—and the final classifier makes a decision
based on all feature-maps in the network.
Reference paper: – [Densely Connected Convolutional Networks] (https://arxiv.org/abs/1608.06993) (CVPR 2017 Best Paper Award)
Web Application Capabilities
You can perform the following on Diabetic Retinopathy Detection System:
- New User Registration Page
- Login and Logout
- Genric Pages – Home, About and Contact Us
- Upload Images – User-friendly interface to upload patient information
- Automatic Densenet inference on the left and right eyes images
- Generates user-friendly reports
- Manage reports – show result, update and delete data
Diabetic Retinopathy web application built on following technology stack
- Django – a Python Web Framework
- Convolutional Neural Network – Keras Densenet Model