Classification of Glioma and Meningioma Brain Tumour Disease Using MRI Image Based on Texture Feature with Random Forest Method

Abstract

This research aims to develop an automatic classification system that distinguishes between glioma and meningioma brain tumours using texture-based MRI images and the Random Forest method. Gliomas, which originate from glial cells, and meningiomas, which originate from the meninges, are the two most common types of brain tumours and have different characteristics. The accuracy of diagnosis is crucial for determining suitable treatment options. MRI image analysis has become a significant method in brain tumour diagnosis, although visual interpretation is often subjective and prone to errors. Therefore, the Random Forest algorithm is applied to overcome these limitations by identifying complex patterns from image data. This study used 300 glioma and 306 meningioma images from the Kaggle database, with texture features extracted using histograms, the Gray Level Co-occurrence Matrix (GLCM), and the Gray Level Run Length Matrix (GLRLM). The Random Forest algorithm in this research achieved an accuracy of 78.88%, a precision of 77.39%, and a recall of 81.00%. These findings demonstrate the potential of Random Forest as an effective tool in brain tumour diagnosis.