Brain Tumor Detection Using Machine Learning

Abstract

There are two types of brain tumors: benign and malignant. The quality of life and life expectancy of these patients are improved by early and timely disease detection and treatment plans. Utilizing deep neural networks is one of the most useful and significant techniques (DNN). In this study, brain magnetic resonance imaging (MRI) pictures were utilized to create a convolutional neural network (CNN) to identify a tumor. CNN was the first to use images. The classification accuracy of the soft max fully connected layer was 98.67%. Additionally, the decision tree (DT) classifier's accuracy is 94.24%, while the radial basis function (RBF) classifier's accuracy is 97.34%. We employ the standards of sensitivity, as well as the accuracy requirement, network performance is measured by specificity and precision. The network accuracy on the picture testing revealed that the soft max classifier has the highest accuracy in CNN, according to the data from the categorizers. This is a novel strategy for tumor detection from brain imaging that combines feature extraction methods with CNN. The method's predicted accuracy for the test data was 99.12%. The accuracy of the doctors’ assistance in diagnosing the tumor and treating the patient rose as a result of the significance of the diagnosis provided by the doctor (Vanitha in JAMA 216:109, 228, 2002 [Vanitha CN, Malathy S, Dhanaraj RK, Nayyar A (2022) Optimized pol- lard route deviation and routeselection using Bayesian machine learning techniques in wireless sensor networks. Comput Netw 216:109,228]).