Brain Tumor Detection Using CNN and ViT

Abstract: The early detection of brain tumors plays a vital role in improving patient survival rates and treatment planning. This project presents a deep learning-based system for Brain Tumor Detection and Classification using Convolutional Neural Networks (CNN) and Vision Transformers (ViT). The system analyzes MRI images to identify and classify tumors automatically. The CNN model effectively extracts local spatial features, while the ViT captures global contextual information, resulting in improved classification accuracy. The proposed approach was trained and tested on MRI datasets, achieving high accuracy and reliability. A Gradio web interface was also developed to provide an interactive platform for real-time image upload and tumor prediction. The experimental results demonstrate that the ViT model outperforms CNN in accuracy and robustness, confirming the potential of transformer-based architectures in medical image diagnosis. This project contributes to the development of an efficient, accurate, and user-friendly system for assisting radiologists in brain tumor detection.

Keywords: Brain Tumor Detection, Convolutional Neural Networks (CNN), Vision Transformer (ViT), Medical Image Analysis, Deep Learning, MRI Classification, Hybrid Architecture, Feature Extraction, Computer-Aided Diagnosis, Transfer Learning, Tumor Classification, Medical Imaging, Neural Networks, Artificial Intelligence in Healthcare, Diagnostic Support System