🩺 Breast Cancer Classification using Deep Learning

A deep learning-based medical imaging system that uses VGG16 Transfer Learning to classify breast ultrasound images as Benign or Malignant with 98.5% accuracy.
The project includes a Gradio-based web interface and automated PDF report generation, enabling real-time clinical decision support.

---

📌 Table of Contents

• Project Overview
• Key Features
• Model Performance
• System Workflow

---

🔍 Project Overview

Breast cancer is one of the most common and life-threatening diseases among women worldwide.
Early and accurate diagnosis significantly increases survival rates.

This project leverages transfer learning to build a highly accurate, fast, and reliable breast cancer classification system using ultrasound images, making it suitable for clinical assistance and academic research.

---

✨ Key Features

🔁 Transfer Learning

• Uses VGG16 pre-trained on ImageNet
• Training completes in under 5 minutes

📊 High Accuracy

• Achieves 98.5% accuracy
• Outperforms ResNet50 and DenseNet121

🧑‍⚕️ Clinical Support Tools

• Gradio Web Interface for easy image upload
• Automated PDF Diagnostic Report using ReportLab

⚡ Fast Inference

• Predicts results in ~0.5 seconds per image

---

📈 Model Performance

Model	Accuracy	Precision	Recall	F1-Score
VGG16	98.5%	0.91	0.89	0.90
ResNet50	96.1%	0.95	0.94	0.94
DenseNet121	97.3%	0.97	0.96	0.96

📝 Note:
All models were trained for 10 epochs using an 80:20 train-test split.

---

🧠 System Workflow

Image Upload
     ↓
Image Preprocessing
     ↓
VGG16 Feature Extraction
     ↓
Custom Classification Head
     ↓
Prediction & Probability Score
     ↓
PDF Diagnostic Report Generation

---

🚀 How to Run (Quick Steps)

1. Install dependencies
2. Train model using `train.py`
3. Run app using `app.py`
4. Upload image and get result