Project Summary: Blindness Detection using Fundus Images

I developed a machine learning model to detect diabetic retinopathy (DR) using fundus images of the eye. The project aimed to assist in the early detection and prevention of blindness in diabetic patients.

Data Collection and Preprocessing:

Obtained a dataset from Kaggle containing fundus images labeled with different stages of DR. Preprocessed the images using various augmentation techniques such as CLAHE, GaussNoise, RandomBrightnessContrast, and Padding to improve model performance. Exploratory Data Analysis (EDA):

Analyzed the dataset to understand the distribution of images across different DR stages. Visualized image dimensions and conducted statistical analysis to understand the variation in image sizes. Model Development:

Utilized a MERN stack (MongoDB, Express.js, React.js, Node.js) for building the web application. Trained a convolutional neural network (CNN) using TensorFlow/Keras to classify DR stages. Achieved a validation accuracy of XX% on the test dataset. Outcome:

The model successfully classified fundus images into different stages of DR, aiding in early diagnosis and treatment. The project highlighted the importance of AI in healthcare and its potential to revolutionize medical diagnostics. Future Scope:

Plan to deploy the model as a web application for easy access by healthcare professionals and patients. Explore additional datasets and refine the model to improve its accuracy and efficiency. This project showcased my expertise in machine learning, image processing, and web development, demonstrating my ability to tackle complex problems and deliver impactful solutions.