Nonlinearized electromagnetic imaging of complex biological structures - towards data fusion

Abstract : Medical imaging for detection/characterization of tumors is of great concern. Among imaging modalities, the electromagnetic (EM) one is widely investigated due to its low cost and assumed safety. Yet, due to the rather long EM wavelength, images intrinsically suffer from low resolution, thus one needs another modality involved to improve the imaging quality. Acoustic (AC) imaging is a known good choice offering high-resolution images. Thus, a joint fusion reconstruction approach is to improve the quality of the image in the both modalities. The Contrast Source Inversion (CSI) method, CSI, which regards contrast and induced current as independent quantities and update them in alternance, is used here in a 2D model case. A priori information of background and physical constraints are implemented in the inversion. Besides, a generalized version of binary CSI is proposed when the values of the EM properties of different tissues are given with some accuracy.