High-resolution microwave stroke detection system based on signal similarity algorithm

Microwave remote sensing is a good candidate to be used in a real-time, low-cost and portable brain stroke detection system for its non-invasive, non-ionizing and cost-effective characteristics. Due to the heterogeneous and sophisticated texture of brain tissue and low dielectric contrast of the stroke, the conventional stroke imaging methods face significant challenges to achieve satisfactory results. This research presents a schema of identification and localization the blood block in a brain based on the comparison of microwave signal similarity. A realistic head model and a numerical analysis technique based on finite-difference time-domain (FDTD) are adopted in the analysis. The simulation results show the schema has blood localization accuracy with resolution of 10-3m.