A Machine Learning Approach for The Brain Tumor Classification Using MR Imaging

This research presents a novel approach for denoising, extracting, and detecting tumors on MRI images. Images obtained from an MRI scanner are helpful to medical professionals in the research and diagnosis of brain disorders and malignancies. This activity aims to assist the radiologist and the physician in obtaining a second opinion on the diagnosis. The ambiguity that existed in the characteristics of magnetic resonance (MR) images has been resolved more straightforwardly. In the paper, the magnetic resonance imaging (MRI) image that was obtained from the machine is analyzed. The study takes advantage of the data collected in real-time. A variety of noise-reduction filters are used throughout the execution of the fundamental preprocessing steps. After the image has been de-noised, it is segmented, and then feature extraction is carried out. The wavelet transform is used in order to extract the features. The wavelet transform is superior to other techniques in terms of its applicability to the MRI image feature extraction process. The characteristics are then sent to the classifier, which conducts classification via Random Forest. A comparison is made between the categorization procedure and more traditional approaches.