Enhanced Brain Tumor Detection from Microwave Imaging with Hybrid Inception-CNN and UWB Circular Monopole Patch Antenna

Brain tumors are among the most prevalent tumor types worldwide.When tissues grow abnormally, they form a tumor.It is classified according to its anatomical location and dimensions.Positron Emission Tomography (PET) scans, Magnetic Resonance Imaging (MRI), biopsies, Computerized Tomography (CT) scans, lumbar punctures, myelograms, Electroencephalograms (EEG), and other methods are available for the diagnosis of brain tumors.The existing approaches are non-invasive but constrained by cost and reliability concerns.The diagnostic procedures are inadequate for detecting tumors beyond a depth of 20% and raise significant concerns about the consequences of ionizing radiation.The radiative effect of the microwave imaging system could serve as a beneficial way to find brain tumors early on because it makes up for the problems with current diagnostic methods for ionizing effects.The proposed work's main goal is to design an effective deep learning approach for identifying brain tumors from microwave brain images.The proposed approach mainly consists of two phases.During the initial phase, an Ultra-Wide Band (UWB) circular monopole patch antenna has been designed for microwave imaging of brain tumors.The generated microwave brain images are fed into the proposed hybrid inception-CNN based brain tumor detection system in the second phase.These extracted features using the inception-CNN model is passed through the fully connected layer for efficient brain tumor classification.The simulation results indicates that the proposed brain tumor detection model using microwave brain images attained better detection performance, with an accuracy of 99.48 %.It also emphasizes how effectively medical imaging techniques can improve detection performance by incorporating artificial intelligence.