A Proposed Deep Learning Approach for Brain Tumor Classification from MRI Images

This study presents a novel approach to brain tumour classification from MRI images that involves merging and analysing a large dataset using deep learning algorithms. The dataset contains 4349 images of various brain tumour types, including meningitis (864), astrocytoma (575), NORMAL (522), neurocy- toma (457), schwannoma (453), and papilloma (227). The goal is to accurately classify these tumour types using cutting-edge deep learning architectures like DenseNet201, VGG19, MobileNetV2, ResNet50, and a proposed Convolutional Neural Network (CNN). The study focuses on assessing the performance of these models on the combined dataset to determine their efficacy in identifying and classifying various tumour types. Our study aims to improve clinical decision-making in neuro-oncology and contribute to neuroimaging diagnostics by utilizing sophisticated algorithms and a diverse dataset. This could result in more effective and targeted treatment strategies.