Repository files navigation

This project focuses on the diagnosis of brain tumors utilizing Convolutional Neural Networks (CNNs). The primary objective is to develop a model capable of accurately classifying brain tumors from MRI images.

The dataset employed in this study is theBrain Tumor MRI Dataset available on Kaggle. It comprises MRI scans categorized into:

• Glioma Tumor
• Meningioma Tumor
• Pituitary Tumor
• No Tumor

For this project, a subset of the dataset was used, consisting of 200 images for each type of tumor in v1 results and 800 images in v2 results.

The complete dataset is available for download at the following link:Brain Tumor MRI Dataset.

1. Data Preprocessing: MRI images are preprocessed to enhance quality and ensure consistency, including steps like skull stripping, histogram equalization, and noise reduction.

2. Model Architecture: A CNN model is constructed to extract features and perform classification. The architecture includes convolutional layers, pooling layers, and fully connected layers optimized for image classification tasks.

3. Training and Evaluation: The model is trained on the preprocessed subset of the dataset, and its performance is evaluated using metrics such as accuracy, precision, recall, and F1-score.

The CNN model’s performance was affected by the limited number of images used in training and testing. While the model initially showed potential, the reduced dataset size likely contributed to lower accuracy and generalization. Future improvements can be achieved by utilizing the full dataset and incorporating data augmentation techniques.

Follow the steps below to run.ipynb files locally using Jupyter and Poetry on different operating systems.

Ensure that you have Poetry installed on your system. If not, install it by following theofficial Poetry installation guide.

• Open a terminal or command prompt.
• Create a new Poetry project or navigate to your existing project directory.

poetry init

• Add Jupyter as a dependency:

poetry add notebook

• Add TensorFlow as a dependency:

poetry add tensorflow

• Activate the virtual environment created by Poetry:

poetry shell

• Start the Jupyter Notebook server:

jupyter notebook

• Open your browser and navigate to the provided URL to access Jupyter Notebook.

1. Ensure that Python and Poetry are installed and added to your PATH.
2. Follow the general setup steps above.

1. Use Homebrew to install Python if necessary:brew install python.
2. Follow the general setup steps above.

1. Use your package manager (e.g.,apt,yum, ordnf) to install Python if necessary.
2. Follow the general setup steps above.

• If you encounter issues with Jupyter Notebook not being recognized, ensure the Poetry environment is active (poetry shell).
• To stop the server, pressCtrl+C in the terminal where the Jupyter server is running.
• TensorFlow may require specific hardware or drivers for GPU support. Refer to theofficial TensorFlow installation guide if you are using a GPU.

• Brain Tumor MRI Dataset
• Human brain tumor classification and segmentation using CNN
• A hybrid deep CNN model for brain tumor image multi-classification