In this paper, the dielectric and hyperspectral properties of phantoms with varying percentages of water and oil were measured. The dielectric properties of the phantoms produced were compared with the theoretical values expected for this type of phantom, modeled using the Bruggeman formula. The hyperspectral characteristics of the phantoms, on the other hand, were used to train and subsequently test (on phantoms produced on different days) a neural network capable of classifying the analyzed tissue based on its water content. The combination of dielectric and hyperspectral information can aid decision-making processes for the classification of anomalies both during and after surgery, reducing the need for more expensive and invasive biopsies.
Oil-in-Gelatin Phantom Classification by Means of Artificial Neural Networks
Di Meo, S.;Torti, E.;Fasol, M.;Riboni, C.;Gandolfi, R.;Pasian, M.;Leporati, F.
2026-01-01
Abstract
In this paper, the dielectric and hyperspectral properties of phantoms with varying percentages of water and oil were measured. The dielectric properties of the phantoms produced were compared with the theoretical values expected for this type of phantom, modeled using the Bruggeman formula. The hyperspectral characteristics of the phantoms, on the other hand, were used to train and subsequently test (on phantoms produced on different days) a neural network capable of classifying the analyzed tissue based on its water content. The combination of dielectric and hyperspectral information can aid decision-making processes for the classification of anomalies both during and after surgery, reducing the need for more expensive and invasive biopsies.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
