An Effective Framework for Deep-Learning-Enhanced Quantitative Microwave Imaging and Its Potential for Medical Applications

Abstract

Microwave imaging is emerging as an alternative modality to conventional medical diagnostics technologies. However, its adoption is hindered by the intrinsic difficulties faced in the solution of the underlying inverse scattering problem, namely non-linearity and ill-posedness. In this paper, an innovative approach for a reliable and automated solution of the inverse scattering problem is presented, which combines a qualitative imaging technique and deep learning in a two-step framework. In the first step, the orthogonality sampling method is employed to process measurements of the scattered field into an image, which explicitly provides an estimate of the targets shapes and implicitly encodes information in their contrast values. In the second step, the images obtained in the previous step are fed into a neural network (U-Net), whose duty is retrieving the exact shape of the target and its contrast value. This task is cast as an image segmentation one, where each pixel is classified into a discrete set of permittivity values within a given range. The use of a reduced number of possible permittivities facilitates the training stage by limiting its scope. The approach was tested with synthetic data and validated with experimental data taken from the Fresnel database to allow a fair comparison with the literature. Finally, its potential for biomedical imaging is demonstrated with a numerical example related to microwave brain stroke diagnosis.

Applications of Microwaves in Medicine Leveraging Artificial Intelligence: Future Perspectives

“…Robust 3D reconstruction of microwave images using deep-learning methodologies is currently underway, offering huge promise toward clinical translation. Therefore, the future of MWI in medicine relies on developing powerful deep-learning networks to reconstruct more practical and complicated heterogeneous human structures to complement existing imaging modalities such as MRI, CT, and ultrasound [149].…”

A deep learning enhanced inverse scattering framework for microwave imaging of piece-wise homogeneous targets

“…• Different from our previous works where U-Net task consisted in classification problems (binary segmentation [18] or categorical segmentation [19]), the U-Net is herein trained within a pixel-wise regression framework, to allow retrieving a continuous set of values; • The a priori information on the piece-wise nature of the targets is encoded by representing the spatial map of the EM properties distribution to be predicted by the network in terms of the corresponding spatial gradient, which allows to explicitly enforce into the training process the implicitly sparse nature of the information to be retrieved. We refer to this map as the augmented shape, to recall that it conveys information on both the target's internal and external boundaries and the relative contrast variation with respect to the (known) background medium;…”

“…Motivated by the above considerations, the authors of this work have considered the use of the orthogonality sampling method (OSM) [17] as the domain knowledge-embedding imaging algorithm [18,19]. The OSM is a qualitative method introduced by Roland Potthast, in which an indicator function is computed to estimate the shape of the unknown targets.…”

“…Last but not least, it has been shown in [20] that OSM images encode information on the spatial behavior of the EM properties of the targets, owing to the relation between the OSM indicator and the radiating component of the induced currents. Based on these considerations, it was shown that a U-Net fed with OSM images could be trained to achieve an objective reconstruction of the targets' shape [18] or an estimate of the targets shape and EM properties, provided they belonged to a fixed and known set of values [19].…”

“…In this paper, we show how an OSM-informed U-Net can be trained to solve the more general problem of imaging piece-wise homogeneous targets, i.e. retrieve their shape and EM properties, without limiting the possible contrast to a finite set of values as in [19]. To face the increased complexity of such a problem, the following strategies are put into action:…”