BRAIN TUMOR IMAGE CLASSIFICATION USING CONVOLUTIONAL NEURAL NETWORKS
The brain tumor detection, and extraction is an arduous and tedious task performed by radiologists or clinical experts. Their accuracy depends on their experience alone. However, when assisted by inexperienced personnel, there is a chance that even experts can make mistakes. And also, this task becomes increasingly onerous when there is abundant data present to be assisted. Thus, the use of computer aided technology comes in hand to overcome these limitations. Here, a method is proposed to detect brain tumor in two dimensional Magnetic Resonance brain Images (MRI) using Convolutional Neural Network, implemented using Keras. The experiment was performed on a real-time dataset having images with varied tumor factors such as image intensity, shape and size. The aim of this paper is to classify the MRI images into benign and malignant brain tissues. The experimental results achieved 95% accuracy, demonstrating the effectiveness of the proposed technique.