Explainable Artificial Intelligence (XAI) for the Classification of Brain Tumour Magnetic Resonance Image (Mri)

Yoon, Han Chel (2025) Explainable Artificial Intelligence (XAI) for the Classification of Brain Tumour Magnetic Resonance Image (Mri). Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text Yoon Han Chel_Full Text.pdf Restricted to Registered users only Download (12MB)

Abstract

The classification of brain tumour MRI plays a crucial role in diagnosis and treatment planning. In recent years, Convolutional Neural Networks (CNNs) have emerged to aid in brain tumour classification. Building upon previous research that optimized hyperparameters and demonstrated the effectiveness of CNN in classifying brain tumour MRI, this study focuses on implementing two CNN models, VGG16 and ResNet50 for the classification of four different categories of brain MRI. Performance evaluation will include metrics such as accuracy, precision, F1 score, and recall. Despite the remarkable classification capability of CNNs, the opacity of CNNs poses challenges in understanding their predictions. This project aims to enhance the transparency and interpretability of CNNs for brain tumour MRI classification. To address this, an Explainable Artificial Intelligence (XAI) framework is proposed, including techniques like Local Interpretable Model-agnostic Explanations (LIME) and Gradient-weighted Class Activation Mapping (Grad- CAM). These techniques use superpixels and heatmaps to provide insights into the decision making process of CNNs. It is anticipated that the XAI framework could improve trust and confidence in predictions, advancing AI applications in brain tumour diagnosis and treatment.