Search

This work presents the design and numerical assessment of a novel microwave imaging (MWI) system, capable of providing a full 3-D image of food/beverage products content in order to disclose the possible presence of physical contaminants, such as plastic fragments. The system here presented exploits the dielectric contrast between the food content and possible intrusions at microwave frequencies; it is based on an antenna array architecture inspecting the items in motion along a conveyor belt without interrupting the production process. The inversion problem is solved by means of linearization, assuming the viability of the Born approximation thanks to the localized intrusions, and regularization, based on the singular value decomposition of the discretized scattering operator. Furthermore, an algorithm, to balance the illumination of the considered scenario due to the nonuniform radiation of the employed antennas, is presented to enhance imaging. The system is first assessed considering an ideal case and then extended to a more realistic approach, for two different kinds of food products, with completely different dielectric properties and considering the performance of existing instrumentation for the purpose. The obtained results lay the foundations for the realization of an actual prototype. [ABSTRACT FROM AUTHOR]

In this article, an innovative approach to microwave imaging, which combines a qualitative imaging technique and deep learning (DL), is presented. The goal is to develop a tool for reliable and user-independent retrieval of the shape of unknown targets from the knowledge of the scattered fields. Qualitative imaging methods are powerful inverse scattering tools, as they provide morphological information in real time. However, their outcome is a continuous map, which has to be hard-thresholded to clearly identify the targets. This thresholding unavoidably results in case-dependent, often user-biased, results. To deal with this issue, a DL approach, based on a physics-assisted deep neural network, is proposed to automatically classify image pixels, i.e., to generate binary masks, separating the targets (foreground) from the background. In particular, the proposed network binarizes the output of a qualitative imaging inversion technique known as the orthogonality sampling method. For the sake of comparison, a DL method is also exploited, which generates the binary masks directly from the scattered fields without any qualitative imaging aid. A quantitative assessment of the performances of both methods and a test on experimental data are provided. [ABSTRACT FROM AUTHOR]

Magnetic scaffolds have been investigated as promising tools for the interstitial hyperthermia treatment of bone cancers, to control local recurrence by enhancing radio- and chemotherapy effectiveness. The potential of magnetic scaffolds motivates the development of production strategies enabling tunability of the resulting magnetic properties. Within this framework, deposition and drop-casting of magnetic nanoparticles on suitable scaffolds offer advantages such as ease of production and high loading, although these approaches are often associated with a non-uniform final spatial distribution of nanoparticles in the biomaterial. The implications and the influences of nanoparticle distribution on the final therapeutic application have not yet been investigated thoroughly. In this work, poly-caprolactone scaffolds are magnetized by loading them with synthetic magnetic nanoparticles through a drop-casting deposition and tuned to obtain different distributions of magnetic nanoparticles in the biomaterial. The physicochemical properties of the magnetic scaffolds are analyzed. The microstructure and the morphological alterations due to the reworked drop-casting process are evaluated and correlated to static magnetic measurements. THz tomography is used as an innovative investigation technique to derive the spatial distribution of nanoparticles. Finally, multiphysics simulations are used to investigate the influence on the loading patterns on the interstitial bone tumor hyperthermia treatment. [ABSTRACT FROM AUTHOR]

In a virtual experiment (VE), the data available to solve an inverse scattering problem are linearly combined in order to make the inversion task simpler. This goal is achieved by conditioning contrast sources radiating the scattered fields and exploiting their properties in the inversion procedure, for instance, by devising data-driven approximations or regularization schemes. In this framework, the design of a VE amounts to determine the coefficients ruling the recombination of the available data. In this article, we present a general and simple procedure to design VEs capable of enforcing contrast sources with desired properties. The proposed procedure is computationally straightforward and stable, as it does not require an explicit inversion and regularization process, and just consists in the evaluation of the adjoint solution of an auxiliary problem. The design approach is tested in the case of VE enforcing circularly symmetric contrast sources, and the corresponding inversion procedure is assessed using both simulated and experimental data. [ABSTRACT FROM AUTHOR]

The orthogonality sampling method (OSM) is a recently introduced qualitative inverse scattering approach for the estimation of the morphological properties of unknown targets. Both the simplicity of implementation and the applicability to various measurement configurations make this method very effective. In this article, a general physical understanding of this otherwise “pure mathematical” method is given for the first time. Such an interpretation is derived from the relationship between the currents induced in the investigated scenario and the so-called reduced scattered field, which is the core of the orthogonality sampling indicator function. Interestingly, the nature of such a relationship implies that the reduced scattered field can be related to the radiating component of the induced currents. A direct consequence of such a result is that the OSM is capable of imaging discontinuities within the unknown targets and hence identify regions with different electromagnetic properties. This possibility represents a unique feature among qualitative inverse scattering methods, as these latter have been introduced as tools to image just the external shape of the unknown targets. The new interpretation as well as the above distinctive capability of the OSM are proved with examples on both simulated and experimental data. [ABSTRACT FROM AUTHOR]

In this article, a method is presented to estimate the effective electrical parameters of the scatterers for multi-view multi-static two dimensional transverse electric (2D-TE) scattering configurations. The derived method needs the shape and position information of scatterers a-priorly to model the targets as circular cylinders (i.e., an effective radius and center is determined for each target) with constant electrical parameters. Next, the method employs the virtual experiments to focus the incident magnetic field as a $J_{0}$ on a specific target. After focusing the excitation and neglecting the multiple scattering between targets, the scattered field is in $H_{1}^{(1)}$ form. Next, the error between the simulated and the virtual fields is calculated. Then, the effective electrical parameter is estimated as the minimizer of the error. Lastly, for the cases where the shape and/or position information are noisy, the introduced method links the deviation of the electrical parameters to deviations of the effective radius and center. From this relation, the method presents a way to estimate the maximum deviations in the electrical parameters, given the maximum deviations in effective radius and center. The efficiency and accuracy of the proposed formulations are tested with both numerical and experimental examples. [ABSTRACT FROM AUTHOR]