MRI Images Based Brain Tumor Detection Using CNN for Multiclass Classification

Brain Tumor has ranked as the 10th most common cause of death in recent years. An early brain tumor diagnosis, detect the faster treatment which improves survival of patients. A biopsy is a medical procedure that helps to use brain tumor were diagnosed and treated. The Artificial Intelligence supports the radiologist in tumor diagnosis without invasive measure. Specifically, algorithms based deep learning CNN techniques are used for the detection of brain tumor. In this paper, we proposed 9 layers convolution neural network for identifying and categorizing tumor on the basis of texture, location and shape. The CNN models are trained and tested using the Kaggle (Brain Tumor Classification - MRI) dataset. The experimental result shows that improved accuracy and can be used to assist the decision making to doctor.