Abstract
Automated identification and categorization of brain anomalies like tumors is crucial to understanding and monitoring patient outcomes. In this study, we applied machine learning models to improve the effectiveness of a step-constant tapered slot antenna (STSA) for diagnosing gliomas in the brain. We have performed simulation with the CST Microwave Studio 2015 to a dataset consisting of astrocytoma (tumor radius 5 mm, skin-tumor distance 24 mm) and matching healthy controls. Phantom, an artificial human head, has four antennas placed at right angles to it. Later on, several machine learning models including the Support Vector Machine (SVM), K Nearest Neighbor (KNN), Random Forest Classifier (RFC), and Artificial Neural Network (ANN) were applied to detect and diagnose brain cancers. There was a rigorous investigation of the scattering (S), admittance (Y), and impedance (Z) parameters and we observed the best result (99.60% accuracy) from the application of ANN and the Z parameter. The proposed technology could be applied for brain tumor detection accurately and clinical applications for a more efficient healthcare system based on biomedical antennas.