Rahul-Samedavar/OralCancer_Predictor

Oral Cancer Prediction & AI Explanation Portal

This project is a built for for understanding and demonstrating the role of Artificial Intelligence in Oral Cancer Diagnosis. It includes:

• Educational Portal explaining AI's role in oral cancer detection.
• CNN Working Visualization for understanding how Convolutional Neural Networks process medical images.
• Deployed Oral Cancer Prediction Model hosted on Hugging Face and integrated into the app.

---

Test it now

The project is hosted in huggingface space,

• Preview here
• Repository

---

🌟 Features

1. AI in Oral Cancer Diagnosis (Frontend 1)

• Interactive content explaining AI's use in early oral cancer detection.
• Simplified explanations of AI models for healthcare professionals and students.

2. CNN Working Visualizer (Frontend 2)

• Explains convolution, pooling, and classification stages in CNNs.
• Includes interactive diagrams and animations.

3. Oral Cancer Prediction Model

• Trained deep learning model for predicting oral cancer from images.
• Uploaded to Hugging Face.
• Integrated directly into the frontend for real-time predictions.

---

🛠️ Tech Stack

Component	Technology Used
Frontend 1	Next.js (Static Export)
Frontend 2	Vite + React
Backend	Flask (serving both frontends & API)
Model Hosting	Hugging Face Spaces
ML Framework	TensorFlow / PyTorch (depending on your implementation)
Styling	Tailwind CSS

---

🚀 Deployment

The project is hosted on Hugging Face Spaces: 🔗 Live Demo

---

⚙️ Installation & Local Setup

1. Clone the repository

git clone https://huggingface.co/Rahul-Samedavar/OralCancer_Predictor
cd OralCancer_Predictor

2. Install backend dependencies

pip install -r requirements.txt

3. Build frontends

# Next.js
cd ./Frontends/main
npm install
npm run build

# Vite
cd ./Frontends/cnn
npm install
npm run build

4. Copy builds into backend

cp -r ./Frontends/main/out ./static/next
cp -r ./Frontends/cnn/dist ./static/vite

5. Run backend

python app.py

---

🧠 Model Details

• Architecture: Convolutional Neural Network (CNN)
• Training Data: Histopathologic Oral Cancer Detection using CNNs
• Hosting: Hosted in Huggingface
• Input: Oral lesion histopathlogical images
• Output: Predicted class (Cancerous / Non-cancerous) with probability score

---