AN OVERVIEW OF DEEP LEARNING IN MEDICAL IMAGING FOCUSING ON MRI-BRAIN TUMOR

Abstract

Magnetic Resonance Imaging (MRI) is widely used for brain tumor diagnosis because it provides detailed visualization of brain tissues. However, manual examination of MRI images is time-consuming and may be affected by diagnostic variability. This paper presents a deep learning-based framework for automated brain tumor detection and classification from MRI images. The proposed system performs image preprocessing, including resizing, normalization, noise reduction, and data augmentation, followed by automatic feature extraction using a Convolutional Neural Network (CNN). U-Net can be employed for tumor segmentation, while ResNet and Vision Transformer architectures provide enhanced feature learning. The system classifies MRI images into Glioma, Meningioma, Pituitary Tumor, and No Tumor categories and provides prediction confidence through a graphical interface. Performance is evaluated using Accuracy, Precision, Recall, F1-score, Dice Similarity Coefficient, Intersection over Union, and confusion matrix analysis. The framework aims to reduce manual workload and support faster, consistent, and computer-assisted brain tumor assessment. Keywords:. MRI, Brain Tumor Detection, Brain Tumor Classification, Deep Learning, Convolutional Neural Network.