Efficient Classification of Brain Tumors Images Using Neural Network Technique

Abstract

Using biopsy, brain tumor classification is performed, which is not normally conducted before definitive brain surgery. Technology improvement and machine learning help radiologists with the diagnostics of tumors without invasive measures. Convolutional neural network (CNN) is the machine-learning algorithm that achieved substantial results in image classification and segmentation. Some of the most notable primary brain tumors are meningiomas, gliomas, and pituitary tumors. Gliomas is a general term for a tumor that arises from brain tissues other than nerve cells and blood vessels. But, meningiomas arise from membranes that cover the brain and surround the central nervous system, whereas pituitary tumors are the lumps that sit inside the skull. The most notable important difference between these three types is that meningiomas are generally benign, and gliomas are commonly malignant. This project develops a new CNN architecture to classify brain tumor types. With i) good generalization capability and ii) good execution speed, the newly developed CNN architecture is being used as an effective decision-support tool for radiologists in diagnostics. Python is used for the development of the project. Keywords: Deep Learning, Neural Network, Brain Tumor, MRI Images.