We present an initial experimental validation of a microwave tomography (MWT) prototype for brain stroke detection and classification using the distorted Born iterative method, two-step iterative shrinkage thresholding (DBIM-TwIST) algorithm. The validation study consists of first preparing and characterizing gel phantoms which mimic the structure and the dielectric properties of a simplified brain model with a haemorrhagic or ischemic stroke target. Then, we measure the S-parameters of the phantoms in our experimental prototype and process the scattered signals from 0.5 to 2.5 GHz using the DBIM-TwIST algorithm to estimate the dielectric properties of the reconstruction domain. Our results demonstrate that we are able to detect the stroke target in scenarios where the initial guess of the inverse problem is only an approximation of the true experimental phantom. Moreover, the prototype can differentiate between haemorrhagic and ischemic strokes based on the estimation of their dielectric properties.

We present a first prototype of a wideband microwave tomography system with potential application to medical imaging. The system relies on a compact and robust printed monopole antenna which can operate in the 1.0–3.0 GHz range when fully immersed in commonly used coupling liquids, such as glycerine–water solutions. By simulating the proposed imaging setup in CST Microwave Studio, we study the signal transmission levels and array sensitivity for different target and coupling liquid media. We then present the experimental prototype design and data acquisition process, and show good agreement between experimentally measured data and results from the CST simulations. We assess imaging performance by applying our previously proposed two-dimensional (2-D) DBIM TwIST-algorithm to both simulated and experimental datasets, and demonstrate that the system can reconstruct simple cylindrical targets at multiple frequencies.

This paper studies how limited information in data acquired by a wideband microwave tomography (MWT) system can affect the quality of reconstructed images. Limitations can arise from experimental errors, mismatch between the system and its model in the imaging algorithm, or losses in the immersion and coupling medium which are required to moderate this mismatch. We also present a strategy for improving reconstruction performance by discarding data that is dominated by experimental errors. The approach relies on recording transmitted signals in a wide frequency range, and then correlating the data in different frequencies. We apply this method to our wideband MWT prototype, which has been developed in our previous work. Using this system, we present results from simulated and experimental data which demonstrate the practical value of the frequency selection approach. We also propose a K-neighbour method to identify low quality data in a robust manner. The resulting enhancement in imaging quality suggests that this approach can be useful for various medical imaging scenarios, provided that data from multiple frequencies can be acquired and used in the reconstruction process.