Lung cancer is one of the deadliest diseases, and early detection can significantly improve survival rates. This AI-driven Lung Cancer Detection system leverages Deep Learning to analyze CT scan images and predict cancer presence efficiently. The project integrates a 3D Convolutional Neural Network (CNN) and a user-friendly Graphical User Interface (GUI) for seamless operation.
Using a dataset of thousands of high-resolution lung scans, this model accurately determines when lesions in the lungs are cancerous. This helps in reducing false positives, providing early access to life-saving interventions, and giving radiologists more time to focus on their patients.
- 📂 Easy Data Import: Load and process DICOM CT scan images effortlessly.
- 🔄 Smart Preprocessing: Automated resizing, normalization, and feature extraction.
- 🤖 Advanced CNN Model: A powerful 3D CNN architecture tailored for medical imaging.
- 📊 Real-Time Training Metrics: Monitor training progress with accuracy and loss graphs.
- 🎨 Interactive GUI: Simple yet effective interface for non-technical users.
We have taken 50 patients as a sample dataset for training and validation.
🔗 Sample Dataset Images: Click Here
- Import Data: Click "Import Data" to load CT scans.
- Preprocess Data: Click "Pre-Process Data" to prepare images.
- Train Model: Click "Train Data" to start CNN model training.
- Predictions: The trained model detects cancerous and non-cancerous cases.
- 📌 3D CNN layers designed for volumetric image analysis.
- 📌 Input shape:
(10, 10, 5, 1)(resized slices of CT scans). - 📌 Uses Adam Optimizer and Categorical Cross-Entropy Loss.
- 📌 Predicts Cancerous (1) or Non-Cancerous (0) cases.
- Efficient lung cancer detection using AI.
- Reduced diagnostic time with automated predictions.
- Enhanced medical imaging analysis with deep learning.
LungCancerDetection-Video.mp4
Let's revolutionize lung cancer detection with AI! 💙🩺