Can sources and receivers be interchanged for imaging?

Abstract

The design of ultrasound scanning systems for applications such as breast cancer detection is a challenging task, especially when the number of sources and receivers increases and they become spread over a large surface. In order to determine the number of transducers and above all their position, several approaches could be followed. A simple and straightforward approach is to compute the energy distribution in the region of interest for several configurations, and treat each receiver as if it is a source. Here the underlying assumption is that the number of A-scans, i.e. number of sources times receivers, determines the image quality. This assumption is mainly based on reciprocity; the observation that the response RAB measured by a receiver located at B and a source at A is identical to the response RBA obtained after interchanging the source and receiver locations. This is valid for linear imaging methods. However, the question arises whether this is also the case for non-linear inversion methods. In this work, we evaluate the SAFT, CG and CSI imaging and inversion methods for different configurations of sources and receivers located uniformly distributed over a circular array, surrounding an heterogeneous medium. The obtained results show that with the linear SAFT and CG methods, the sources and receivers can be interchanged, and the resulting images are the same. However, for the non-linear CSI method, the results show that it is preferred to have more receivers than sources, despite the fact the measured signals are reciprocal.