[Python | Keras| CNN]
Welcome to the "Deep Learning Brain Tumour Image Classification using CNN" repository! This project was developed as part of the "Data Mining and Machine Learning II" course, focusing on deep learning tools and techniques.
Abstract:
Brain tumours, characterized by the growth of abnormal cells or masses in the brain, pose a significant medical challenge. Magnetic Resonance Imaging (MRI) is a common diagnostic method for detecting brain tumours, generating a large number of images. Manual identification of brain tumours from these images is prone to errors, whereas deep learning-based classification approaches offer promising results.
This project aims to address the challenges in medical image processing by leveraging deep learning and advanced image processing technologies. Specifically, a Convolutional Neural Network (CNN) is employed to classify brain images. The brain scans are categorized into four classes: no tumour, glioma, meningioma, or pituitary tumours. The Brain Tumour Classification dataset from Kaggle is utilized for this project.
To evaluate the performance of the constructed model, various metrics such as the confusion matrix, accuracy, F1 score, recall score, and precision are employed.
- Code Implementation: The Python code files implementing the Convolutional Neural Network for brain tumour image classification using the Keras library.
- Dataset: The Brain Tumour Classification dataset from Kaggle, which serves as the foundation for training and evaluating the model.
- Results and Evaluation: The evaluation metrics, including the confusion matrix, accuracy, F1 score, recall score, and precision, obtained from the trained model.
- Access the repository to explore the code implementation and dataset used for brain tumour image classification.
- Review the Python code files to gain insights into the Convolutional Neural Network architecture and its implementation using the Keras library.
- Examine the Brain Tumour Classification dataset to understand the characteristics and composition of the data.
- Dive into the results and evaluation section to analyze the performance metrics of the trained model.
By leveraging deep learning techniques and the power of Convolutional Neural Networks, this project aims to contribute to the accurate classification of brain tumour images. Join us on this journey of utilizing cutting-edge technology to improve medical image processing and enhance brain tumour diagnosis.