CVNN Approach for Microwave Imaging Applications in Brain Cancer: Preliminary Results

The application of Complex-Valued Neural Network (CVNN) is proposed in this work to significantly enhance the image quality derived from the solution of Inverse Scattering Problem (ISP) which are faced with Quadratic Born Iterative Method (BIM). The proposed approach takes advantage from the exploitation of the full information contained into complex-valued data inherent to the microwave imaging context, thus leading to directly reconstruct the conductivity parameter.