https://github.com/nishithat-28/glaucoma-detection-using-deep-learning
Glaucoma Detection using Deep Learning | A ResNet50-based deep learning model for automated glaucoma diagnosis from fundus images.
https://github.com/nishithat-28/glaucoma-detection-using-deep-learning
computer-vision deep-learning glaucoma-detection medical-imaging optic-cup resnet50 streamlit tensorflow
Last synced: about 2 months ago
JSON representation
Glaucoma Detection using Deep Learning | A ResNet50-based deep learning model for automated glaucoma diagnosis from fundus images.
- Host: GitHub
- URL: https://github.com/nishithat-28/glaucoma-detection-using-deep-learning
- Owner: nishithat-28
- Created: 2025-02-14T16:08:20.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-02-15T17:54:06.000Z (over 1 year ago)
- Last Synced: 2025-02-15T18:34:59.457Z (over 1 year ago)
- Topics: computer-vision, deep-learning, glaucoma-detection, medical-imaging, optic-cup, resnet50, streamlit, tensorflow
- Language: Jupyter Notebook
- Homepage:
- Size: 1.6 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# **🩺 Glaucoma Detection using Deep Learning**
🔬 **Early diagnosis of glaucoma is crucial to prevent vision loss!** This project leverages **Deep Learning (ResNet50)** for automated glaucoma detection using fundus images.
Glaucoma is a **chronic eye disease** that damages the **optic nerve**, leading to irreversible blindness if untreated. Our project uses a **Convolutional Neural Network (CNN)** based on **ResNet50** to classify images into:
- ✅ **Normal**
- ⚠️ **Glaucoma-Affected**
## **🛠️ Tech Stack**
- **Programming Language:** Python
- **Deep Learning Framework:** TensorFlow & Keras
- **Model Architecture:** ResNet50
- **Preprocessing:** OpenCV (CLAHE, Image Augmentation)
- **Deployment:** Streamlit
## **📂 Dataset**
### 📌 **Source**
This dataset contains **fundus images** from **3 combined sources** (ACRIMA, DRISHTI-GS, ORIGA).
- Kaggle: [Glaucoma Classification Datasets](https://www.kaggle.com/datasets/ayush02102001/glaucoma-classification-datasets)
It includes:
- **"Normal"** 🟢 (Healthy eyes)
- **"Glaucoma"** 🔴 (Affected eyes)
🔄 **Preprocessing Techniques:**
- **CLAHE** - Enhances contrast
- **Image Augmentation** - Random rotations, flips, and zooms to improve generalization
- **Resizing** - Standardizing images to **256x256**
## **🔢 Training Results**
- **Final Training Accuracy:** **98.86%**
- **Final Validation Accuracy:** **88.89%**
- **Test Accuracy:** **77.64%**
## **📌 How to Run the Project**
### **🔧 Install Dependencies**
```sh
pip install -r requirements.txt
```
### **🏃 Run the Streamlit App**
```sh
python -m streamlit run app.py
```
## **🎛️ Streamlit UI**
We developed an **interactive UI** using **Streamlit**, allowing users to:
- 🔍 **Upload a fundus image**
- 🖼️ **Preview the image**
- 🧠 **Get instant glaucoma detection results with confidence score**
## **🖥️ UI Preview**
The landing screen of the Glaucoma Detection web app. It provides a brief introduction and guides the user to upload a retinal image.
---
## 🧾 License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.