An Automated Early-Stage Brain Tumor Recognition System Using Texture Feature Extraction and Support Vector Machine Classification

Abstract

Early-stage recognition of brain tumors is essential for improving patient survival rates and supporting effective treatment planning. Magnetic Resonance Imaging (MRI) is widely used for visualizing brain abnormalities; however, manual interpretation of MRI scans is time-consuming and highly dependent on radiologist expertise. This paper presents an automated early-stage brain tumor recognition system based on texture feature extraction and Support Vector Machine (SVM) classification. Gray Level Co-occurrence Matrix (GLCM)–based texture features, including contrast, energy, homogeneity, correlation, and entropy, are extracted from preprocessed MRI images to characterize the structural differences between normal and tumor-affected brain tissues. These features are used to train an SVM classifier for reliable discrimination between tumor and non-tumor cases. Experimental results demonstrate that the proposed system achieves high classification accuracy, sensitivity, and specificity, confirming its effectiveness for early-stage brain tumor recognition and its potential as a computer-aided diagnostic tool in clinical practice.