Analysis of the Convolutional Neural Network Model in Detecting Brain Tumor

Abstract

Detecting brain tumors is an active area of research in brain image processing. This paper proposes a methodology to segment and classify brain tumors using magnetic resonance images (MRI). Convolutional Neural Networks (CNN) are one of the effective detection methods and have been employed for tumor segmentation. We optimized the total number of layers and epochs in the model.  First, we run the CNN with 1000 epochs to see its best-optimized number.  Then we consider six models, increasing the number of layers from one to six.  It allows seeing the overfitting according to the number of layers.