This contribution is divided in two parts. In the first part, we present the work done at the Microwave Laboratory and the Laboratory of Bioengineering of the University of Pavia on the design of a millimeter-wave imaging system for breast cancer detection. In the second part, we present the analytical model we have developed for the numerical simulation of multi-static radar configurations against both simple scenarios, in which one or more targets share the same medium as the antenna array, and more complex configurations, in which the targets lie in a medium other than the antennas and skin is present at the interface. Normalized radar images, reconstructed with the Delay-And-Sum (DAS) algorithm, show the robustness of the model in all scenarios considered. The models proposed in this work, although based on illustrative assumptions, make it possible to overcome the bottleneck provided by the enormous computational and temporal demands of full-wave simulations, thus offering the possibility of generating in real time the desired number of configurations on which to test, for example, skin removal algorithms and/or train the neural networks for artificial intelligence purposes (aimed, as an example, to classify suspicious regions).
Millimeter-Waves Imaging for Breast Cancer Detection: Development of Analytical Models for Scattering Matrices Computation
Di Meo S.;Matrone G.;Magenes G.;Pasian M.
2023-01-01
Abstract
This contribution is divided in two parts. In the first part, we present the work done at the Microwave Laboratory and the Laboratory of Bioengineering of the University of Pavia on the design of a millimeter-wave imaging system for breast cancer detection. In the second part, we present the analytical model we have developed for the numerical simulation of multi-static radar configurations against both simple scenarios, in which one or more targets share the same medium as the antenna array, and more complex configurations, in which the targets lie in a medium other than the antennas and skin is present at the interface. Normalized radar images, reconstructed with the Delay-And-Sum (DAS) algorithm, show the robustness of the model in all scenarios considered. The models proposed in this work, although based on illustrative assumptions, make it possible to overcome the bottleneck provided by the enormous computational and temporal demands of full-wave simulations, thus offering the possibility of generating in real time the desired number of configurations on which to test, for example, skin removal algorithms and/or train the neural networks for artificial intelligence purposes (aimed, as an example, to classify suspicious regions).I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
