Microwave imaging of brain strokes

This thesis investigates feasibility of detection and continuous monitoring of hemorrhagic brain strokes in a realistic head phantom with microwave imaging. The head phantom is illuminated by 36 line sources and the scattering field data is calculated numerically with method of moments for each scenarios. Additive white Gaussian noise with minimum level 30 dB SNR is added for scenarios that are more realistic. Two different approaches are used for detection and continuous monitoring of blood region. In first approach, contrast source inversion method is used to reconstruct the blood region. Scattering fields' data that are obtained by a simulation on matlab platform at different frequencies from 500 MHz to 1.2 GHz is used for only input parameters for CSI method. No any other priory information is used. Moreover, results with either electrically different background mediums or free space are discussed in this approach. The numerical results show that it is possible to determine the square shape blooded area with size about 2.6cm x 2.6cm between the 800-1000 MHz frequency ranges in the dielectric-matching medium. In second approach, CSI method is used to reconstruction again except the input parameters scattering field data is replaced with differential data, which is the difference of two scattering fields' data. These scattering fields' data's are obtained in two different times. Hence, reconstruction contains of changes in multiple blood regions and provides information about the locations instead of all domain reconstruction. Here, aim is to determine the change of bleeding in sequential time frames. The numerical results show that it is possible to monitor the changes in the blood regions within the human brain, if optimal values of matching medium are chosen.