Detection of Brain Tumor Stages and Treatment Suggestion Techniques Using LBP and Hybrid PSO Optimization Techniques

Abstract

The brain tumor classification procedure performs better when the best attributes are used for the brain tumor classification. The process's overall time complexity and algorithmic complexity are decreased by choosing the best features. The disorders in the tumor region photos were identified using CT scans. LBP features were used to extract features. To evaluate the effectiveness of the chosen features, five distinct classifiers were used to classify the characteristics. The process's overall effectiveness was evaluated using performance measures. The primary goal of the procedure is to choose the best characteristics from the various feature types that are extracted using various techniques. to evaluate how well various classifiers, perform in relation to the chosen features. to extract various information from the photos, such as texture- and intensity-based features. Using hybrid PSO optimization techniques, this work uses the best analysis of brain tumor detection, classification, and therapy analysis. 97% accuracy is produced in the testing datasets by this approach. Accuracy, sensitivity, and specificity were among the performance indicators used to gauge the process's effectiveness.