Detecting and classifying multiple types of brain tumor with convolutional neural network from MRI images

With recent developments in science and technology, Artificial Intelligence has become one of the most inseparable parts of medical science and human health aid systems, from autonomous robots used in surgeries to diagnostic systems which can diagnose diseases faster and more accurate than most healthcare professionals. With the help of A.I systems and specifically Deep Learning algorithms, we have been able to rely more and more every day on the modern technologies to detect what is undetectable for humans in order to diagnose illnesses as fast as possible and save thousands of lives. Brain tumors are one of the most fatal conditions which can grow fast and unexpectedly, and endanger people’s health beyond any cure and treatments. Hence the development of a reliable A.I system that is capable of accurate diagnostician of the existence and the type of the tumor is vital. In this paper a convolutional neural networks (CNN) based model is introduced that has been able to detect and classify various types of brain tumors from MRI images with an accuracy ۹۸.۷۹% in the training phase and ۸۸.۲۴% in the validation phase.

کلیدواژه ها:

Brain tumor ، deep learning ، CNN Network ، Image processing ، Tumor detection ، MRI Dataset ، smart healthcare ، brain tumor classification