A brain tumor is an abnormally growing cell in the brain. Replicating a tumour cell is uncontrollable. Early diagnosis is dependent on the size and location of tumor. A technique is explored in this study to extract entropy features of sample entropy by the Magnets of Resonance Imaging (MRI) sample entropy, Dual Tree Wavelet Transform (DTWT), sample entropy and NB classifier. The experimental results illustrate the efficiency of the device proposed. The DTWT, sample entropy and NB classifier produces 96% precision, sensitivity and specificity.