https://github.com/athiyaman-m/brain-tumor-classification-using-deep-learning

• This deep learning model is developed to classify and identify different types of brain tumors, as well as determine their existence, using Convolutional Neural Networks. This is achieved through the training and testing of 3264 MRI images.
https://github.com/athiyaman-m/brain-tumor-classification-using-deep-learning

keras numpy opencv pandas sklearn tensorflow

Last synced: about 1 month ago
JSON representation

• This deep learning model is developed to classify and identify different types of brain tumors, as well as determine their existence, using Convolutional Neural Networks. This is achieved through the training and testing of 3264 MRI images.

• Host: GitHub
• URL: https://github.com/athiyaman-m/brain-tumor-classification-using-deep-learning
• Owner: athiyaman-m
• License: mit
• Created: 2024-03-30T03:22:21.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2024-03-30T05:47:47.000Z (almost 2 years ago)
• Last Synced: 2024-03-30T06:27:25.711Z (almost 2 years ago)
• Topics: keras, numpy, opencv, pandas, sklearn, tensorflow
• Language: Jupyter Notebook
• Homepage:
• Size: 121 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Brain-Tumor-Classification-using-Deep-Learning

# Process Flow :

1) Dataset : 

   Taken from kaggle : https://www.kaggle.com/datasets/sartajbhuvaji/brain-tumor-classification-mri

   Consists of 3264 tumor MRI images data for training and testing.

3) Splitting data into categories

4) Convolution Neural Network (CNN)

5) Training

6) Testing

7) Prediction

# Description :

1. **Objective**:

   - The primary goal of this project is to **classify brain tumors** based on MRI (Magnetic Resonance Imaging) scans.

2. **Data Collection and Preprocessing**:

   - **Data Source**: The project uses a dataset of MRI images.

   - **Preprocessing Steps**:

     - Load the MRI images.

     - Normalize pixel values to a common range (e.g., 0 to 1).

     - Resize the images to a consistent size.

     - Split the dataset into training and validation sets.

3. **Model Architecture**:

   - The project employs a **Convolutional Neural Network (CNN)** for tumor classification.

   - **Architecture Components**:

     - Convolutional layers with filters to learn spatial features.

     - Pooling layers for downsampling.

     - Fully connected layers for classification.

4. **Model Training**:

   - The CNN model is trained using the training data.

   - **Loss Function**: Cross-entropy loss.

   - **Optimizer**: Adam optimizer.

   - **Training Process**:

     - Forward pass: Compute predictions.

     - Backward pass: Update weights using gradients.

     - Repeat for multiple epochs.

5. **Model Evaluation**:

   - The trained model is evaluated on the validation set.

   - Metrics such as **accuracy**, **precision**, **recall**, and **F1-score** are computed.

6. **Results and Interpretation**:

   - The project reports the model's performance metrics.

   - Interpretation of results:

     - High accuracy indicates good overall performance.

     - Precision and recall provide insights into class-specific performance.

7. **Comments and Further Steps**:

   - The project includes three comments:

     1. **Data Augmentation**: Discusses techniques to enhance the dataset.

     2. **Hyperparameter Tuning**: Suggests optimizing model hyperparameters.

     3. **Transfer Learning**: Explores using pre-trained models for feature extraction.

8. **Libraries Used**:

   - **NumPy**: For numerical computations and array manipulation.

   - **Pandas**: For data manipulation and analysis.

   - **Matplotlib** and **Seaborn**: For visualization.

   - **Scikit-learn**: For machine learning tools.

   - **TensorFlow (with Keras)**: For deep learning model creation.

   - **OpenCV**: For image preprocessing.