Dielectric Characterization and Design of a Non-Static Multi-Tissue Phantom for Microwave Brain Stroke Monitoring

Brain stroke is a significant cause of mortality and disability worldwide, with a substantial increase in the number of cases in recent years. The availability of new tools and instruments has greatly aided in the detection and treatment of strokes. Particularly crucial for effective treatment is the monitoring of the disease stage in the post-acute phase. Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) are commonly employed imaging systems for stroke monitoring. However, their lengthy acquisition times limit their utility in emergency situations, where immediate intervention is critical. Additionally, MRI is associated with high production and maintenance costs, making it less accessible and affordable. Moreover, CT employs ionizing radiation, which poses risks to patients. Real-time knowledge of the size and position of a stroke during the post-acute stage would significantly improve patient outcomes and enable personalized care. Microwave Imaging (MWI) systems offer an affordable, harmless, and accessible solution for real-time monitoring of brain stroke, even in non-specialized environments. This thesis aims to develop a non-static multi-tissue phantom for the experimental validation of a non-invasive MWI brain scanner. Special emphasis is placed on the dielectric characterization of materials used to mimic different head tissues in the microwave frequency range. Two different measurement methodologies, their capabilities, and results are investigated and compared. The first one relies on the open-ended coaxial probe technique, which proves effective in measuring the dielectric properties of liquid materials. The second methodology, employing a custom ridge waveguide, is expected to be better suited for testing solid materials. In this case, a numerical fitting method involving simulations in CST Studio Suite and an optimizer is implemented to determine the relative permittivity and conductivity from the measured scattering parameters. Data on different materials are then collected, and recipes for the phantom components are selected based on the available literature on the dielectric properties of head tissues. The design phase considers the essential requirements for the validation of the brain scanner, and, utilizing 3D modeling and printing techniques, a non-static multi-tissue head phantom is fabricated, incorporating both solid and liquid components, such as rubber-based and alcohol-based mixtures. This phantom is capable of simulating various stages of a stroke, representing different evolving scenarios in real-time. Additionally, a customized helmet is designed and manufactured to accommodate and couple the antenna array of the brain scanner to the phantom, ensuring alignment with the topography employed in the numerical model of the system. Finally, a preliminary test is conducted, demonstrating that the brain scanner is capable not only of detecting the presence of a stroke within the fabricated phantom but also of producing a 3D image describing its position and morphology. This research has led to the development of an innovative and functional phantom that successfully addresses the non-trivial challenge of combining a multi-tissue structure with dynamic components. The obtained results and implemented methodologies pave the way for the development of a new generation of phantoms, aiming to enhance the improvement of medical devices and demonstrate their reliability in clinical practice.