AI-Based Methodologies for Next-Generation Biomedical Imaging: Recent Advances and Future Trends

Recent advances and future trends in Artificial Intelligence (A1)-based methods for biomedical imaging are discussed in this work. Among them, a focus is given on the “three-steps” Learning-by-Examples (3-LBE) paradigm which allows the efficient/effective generation of robust and accurate surrogate models (SMs) for the realtime inversion of electromagnetic (EM) data. Moreover, the pillar ideas and concepts of the System-by-Design (SbD) framework are outlined when addressing the solution of complex inverse scattering problems (ISPs) arising in several biomedical microwave imaging scenarios.