Microwave sensors are gaining increasing interest in blood glucose detection, for their potential ability to perform a continuous non-invasive monitoring of the glucose concentration, by relating the change in the blood dielectric properties to a variation in the glucose level. Usually, the involved body part (phantom) is placed on the sensor to perform the reading. However, the placement modality, as well as other external factors not related to the blood glucose concentrations (BGC) (i.e. system noise, environmental temperature, human tissues variations other than blood tissue) may also have an effect on the sensor response, due to the change in the propagation path of the electromagnetic field inside the body part under test. In this work, the variation effects induced on the microwave sensor response by the changes in the thickness and the dielectric properties of skin and fat tissues are analyzed and faced. In particular, to mitigate the above drawback in terms of sensor instability, a solid ”matching layer” is interposed between the resonant sensor and the phantom under test. A specific optimization procedure is performed to design microwave sensors with a stable response not influenced by variations in tissues different from the blood. Various sensors configurations with related resolution metrics are considered to assess the proposed idea and design methodology. Numerical results confirm the possibility to achieve a good trade-off between the measurement stability against undesired phantom variations and the sensitivity toward blood glucose levels, allowing to discriminate concentrations in the range of [100-300] mg/dL.

In this paper we consider radar approaches for breast cancer detection. The aim is to give a brief review of the main features of incoherent methods, based on beam-forming and Multiple SIgnal Classification (MUSIC) algorithms, that we have recently developed, and to compare them with classical coherent beam-forming. Those methods have the remarkable advantage of not requiring antenna characterization/compensation, which can be problematic in view of the close (to the breast) proximity set-up usually employed in breast imaging. Moreover, we proceed to an experimental validation of one of the incoherent methods, i.e., the I-MUSIC, using the multimodal breast phantom we have previously developed. While in a previous paper we focused on the phantom manufacture and characterization, here we are mainly concerned with providing the detail of the reconstruction algorithm, in particular for a new multi-step clutter rejection method that was employed and only barely described. In this regard, this contribution can be considered as a completion of our previous study. The experiments against the phantom show promising results and highlight the crucial role played by the clutter rejection procedure.

In this contribution, a wearable microwave imaging system for real-time monitoring of brain stroke in the post-acute stage is described and validated. The system exploits multistatic/multifrequency (only 50 frequency samples) data collected via a low-cost and low-complexity architecture. Data are collected by an array of only 16 antennas moved by pneumatic system. Phantoms, built from ABS material and filled with appropriate Triton X-100-based mixtures to mimic the different head human tissues, are employed for the experiments. The microwave system exploits the differential scattering measures and the Incoherent MUSIC algorithm to provide a 3D image of the region under investigation. The shown results, although preliminary, confirm the potential of the proposed microwave system in providing reliable results, including for targets whose evolution is as small as 16 mL in volume.

Near-field techniques for antenna testing can require the collection of a very large number of data when the working frequency increases and/or the antenna’s is large in terms of the wavelength. To mitigate such a drawback, a greedy method, we call warping driven Maximum Noise Propagation Error (MNPE) algorithm, is introduced. Antenna under test (AUT) diagnostics is cast as the reconstruction of an equivalent surface current. Starting from a densely populated initial grid, the algorithm selects the sampling points one by one till the noise factor , a figure that controls the propagation of noise from data to the source reconstruction, does not meet a slowly varying region. Hence, the procedure is arrested since no relevant further noise factor reduction would be obtained. It is shown that MNPE allows for a significant data reduction as compared to other literature methods, especially when the starting grid is chosen according to the recently introduced warping sampling theory. Moreover, the initial grid selected according to the warping sampling in turn allows to dramatically reduce the computational cost of the selection procedure. A distinctive feature of the MNPE is that a priori information about the AUT can be easily included by projecting the unknown current onto properly selected base functions. As an example of that, we exploit the a priori information concerning the size of the source and of the measurement aperture that lead the radiation pattern being reliable within a bounded part of the visible spectrum (i.e, the so-called valid angle region). Extensive numerical analysis and a few experimental results pertaining to planar scanning confirm the effectiveness of the proposed method in dramatically reducing the number of data without incurring in performance degradation while estimating the source spatial spectrum.

In this contribution we briefly present the experimental verification of incoherent radar imaging approaches recently developed in the framework of medical imaging. In particular, a prototype radar system is described and breast and head phantoms are developed and employed to collect the measurements.