OCT Scan Image Classifier

This repository contains a Streamlit application for classifying OCT scan images using a deep learning model trained on a large dataset.

Model Details

• Model Architecture: ResNet50
• Accuracy: 99.59%
• Training Data: 84,000 OCT scan images

Watch the demo video of the application on YouTube:

Requirements

Make sure you have Python installed. You can download it from python.org.

Installation

1. Clone the repository:

   git clone https://github.com/ramchandra06032004/OCT-images-classifier.git
   cd OCT-images-classifier

2. Create a virtual environment (optional but recommended):

   python -m venv venv

3. Activate the virtual environment:

   • On Windows:

     venv\Scripts\activate

   • On macOS and Linux:

     source venv/bin/activate

4. Install the required packages individually:

   • Pillow:

     pip install Pillow

   • NumPy:

     pip install numpy

   • Pandas:

     pip install pandas

   • TensorFlow:

     pip install tensorflow

   • Streamlit:

     pip install streamlit

   • Plotly:

     pip install plotly

   • Scikit-learn:

     pip install scikit-learn

Running the Application

Before running the application, download the trained model from Google Drive (link provided below) and place it in the ModelTraining folder.

Download Trained Model

1. Navigate to the streamlit_app directory:

   cd streamlit_app

2. Run the Streamlit app:

   streamlit run app.py

3. Open your web browser and go to http://localhost:8501 to view the app.

Acknowledgements

• Streamlit
• TensorFlow
• Pandas
• Plotly
• Scikit-learn
• Pillow
• NumPy

Highlights

• High Accuracy: Our model achieves an impressive 99.59% accuracy in classifying OCT scan images.
• Robust Architecture: Utilizes the powerful ResNet50 architecture, ensuring efficient and accurate image classification.
• Extensive Training Data: Trained on a comprehensive dataset of 84,000 OCT scan images, enhancing the model's reliability and performance.