AN AUTOMATED INTELIGENCE SYSTEM FOR DETECTION AND CLASSIFICATION WITH MEDICAL IMAGE BRAIN TUMOR USING DEEP LEARNING TECHNIQUES

Abstract

Brain tumors remain one of the most challenging pathologies to diagnose and treat in clinical neurology. Magnetic Resonance Imaging (MRI) plays a pivotal role in the early detection and assessment of brain tumors, which are typically classified as low-grade (slow- growing) or high-grade (fast-growing). Accurate and prompt grading is crucial for determining appropriate therapeutic strategies and improving patient outcomes. This study proposes an automated brain tumor detection and classification system that integrates advanced deep learning techniques to enhance diagnostic accuracy and reduce dependency on manual evaluation. The system employs wavelet-based thresholding to denoise MRI images, followed by improved semantic segmentation using a U-Net architecture to delineate tumor regions with high precision. Feature extraction and classification are performed using a Deep Convolutional Neural Network (DCNN) optimized through transfer learning. Publicly available medical MRI datasets were used, ensuring wide variability in tumor types, locations, and sizes. Experimental results indicate high accuracy in differentiating normal and abnormal scans, as well as in classifying tumor grades. Evaluation through confusion matrices and performance metrics,