A microwave imaging technique based on artificial neural networks for neck tumors detection

A microwave imaging approach for the detection of neck tumors is proposed in this paper. Specifically, a preliminary artificial neural network (ANN) is adopted for the reconstruction of the properties of a cross section of the neck starting from scattered electric fields. To this end, realistic neck phantoms are developed to test the feasibility of the proposed method. Several test cases are performed to evaluate the performance of the ANN to discriminate different sizes of the tumor and different positions inside the neck. The preliminary results indicate quite good detection capabilities.