INTEGRATING MACHINE VISION TECHNIQUES FOR ACCURATE BRAIN TUMOR IDENTIFICATION IN MAGNETIC RESONANCE IMAGES(MRI)

Abstract

Detecting brain tumors from MRI scans presents a significant challenge for medical practitioners. Early and accurate diagnosis greatly enhances a patient’s chances of recovery. In recent years, machine vision techniques have gained remarkable importance in the field of brain tumor detection. In this study, image-processing methods are applied to enhance MRI images and improve the visibility of tumor regions for more precise detection.

The Computer Vision and Image Processing (CVIP) tool plays a central role in this research. It serves as an effective platform for image pre-processing, feature extraction, and classification, providing an interactive environment for students, researchers, and educators to explore the potential of digital image processing. Using the CVIP tool, various classifiers were implemented and evaluated. Among these, the Naïve Bayes classifier achieved an accuracy of 90%, the Bag of Visual Words Model (BOVM)-based SVM achieved 97.3%, and the Convolutional Neural Network (CNN) outperformed others with an accuracy of 98.5%. These results indicate that the CNN-based approach provides superior performance in the classification of brain tumor MRI images compared to traditional classifiers.

Keywords

Deep Learning, Brain Tumour, Classification, SVP, Machine Vision