ABSTRACT
Stroke is a very frequent disorder and one of the major leading causes of death and disability worldwide. Timely detection of stroke is essential in order to select and perform the correct treatment strategy. Thus, the use of an efficient imaging method for an early diagnosis of this syndrome could result in an increased survival's rate. Nowadays, microwave imaging (MWI) for brain stroke detection and classification has attracted growing interest due to its non-invasive and non-ionising properties. In this paper, we present a feasibility study with the goal of enhancing MWI for stroke detection using metasurface (MTS) loaded antennas. In particular, three MTS-enhanced antennas integrated in different brain scanners are presented. For the first two antennas, which operate in a coupling medium, we show experimental measurements on an elliptical brain-mimicking gel phantom including cylindrical targets representing the bleeding in haemorrhagic stroke (h-stroke) and the not oxygenated tissue in ischaemic stroke (i-stroke). The reconstructed images and transmission and reflection parameter plots show that the MTS loadings improve the performance of our imaging prototype. Specifically, the signal transmitted across our head model is indeed increased by several dB's over the desired frequency range of 0.5-2.0 GHz, and an improvement in the quality of the reconstructed images is shown when the MTS is incorporated in the system. We also present a detailed simulation study on the performance of a new printed square monopole antenna (PSMA) operating in air, enhanced by a MTS superstrate loading. In particular, our previous developed brain scanner operating in an infinite lossy matching medium is compared to two tomographic systems operating in air: an 8-PSMA system and an 8-MTS-enhanced PSMA system. Our results show that our MTS superstrate enhances the antennas' return loss by around 5 dB and increases the signal difference due to the presence of a blood-mimicking target up to 25 dB, which leads to more accurate reconstructions. In conclusion, MTS structures may be a significant hardware advancement towards the development of functional and ergonomic MWI scanners for stroke detection.
ABSTRACT
This paper reports the development of a new composite material as a matching medium for medical microwave diagnostic systems, where maximizing the microwave energy that penetrates the interrogated tissue is critical for improving the quality of the diagnostic images. The proposed material has several advantages over what is commonly used in microwave diagnostic systems: it is semi-flexible and rigid, and it can maximize microwave energy coupling by matching the tissue's dielectric constant without introducing high loss. The developed matching medium is a mirocomposite of barium titanate filler in polydimethylsiloxane (PDMS) in different weight-based mixing ratios. Dielectric properties of the material are measured using a Keysight open-ended coaxial slim probe from 0.5 to 10 GHz. To avoid systematic errors, a full dielectric properties calibration is performed before measurements of sample materials. Furthermore, the repeatability of the measurements and the homogeneity of the sample of interest are considered. Finally, to evaluate the proposed matching medium, its impact on a printed monopole antenna is studied. We demonstrate that the permittivity of the investigated mixtures can be increased in a controlled manner to reach values that have been previously shown to be optimal for medical microwave imaging (MWI) such as stroke and breast cancer diagnostic applications. As a result, the material is a good candidate for supporting antenna arrays designed for portable MWI scanners in applications such as stroke detection.
ABSTRACT
We present an initial experimental validation of a microwave tomography (MWT) prototypefor brain stroke detection and classification using the distorted Born iterative method, two-stepiterative shrinkage thresholding (DBIM-TwIST) algorithm. The validation study consists of firstpreparing and characterizing gel phantoms which mimic the structure and the dielectric propertiesof a simplified brain model with a haemorrhagic or ischemic stroke target. Then, we measure theS-parameters of the phantoms in our experimental prototype and process the scattered signals from 0.5to 2.5 GHz using the DBIM-TwIST algorithm to estimate the dielectric properties of the reconstructiondomain. Our results demonstrate that we are able to detect the stroke target in scenarios where theinitial 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 theestimation of their dielectric properties.
