Seismic imaging method for medical ultrasound systems

Abstract

Conventional medical ultrasound systems normally implement ray-based imaging algorithms, such as delay-and-sum beamforming, whose severest limitation derives from the implicit assumption of constant-velocity media. As a result, in the case of two or more tissues with different velocities, the image of the underlying targets appears strongly degraded both in placement and in resolution. The proposed ultrasound-imaging strategy, a value-added application of concepts developed in the context of seismic prospecting, avoids this restriction by relying on the undulatory description of the physical process and not on the geometric one. Echoes are sensed in the synthetic aperture configuration by the transducer, whose elements sequentially emit a nearly spherical wave front that covers the whole region of interest. Given a macrovelocity model, the recorded echoes become the initial condition for the downward propagator of the ultrasound wavefield. The processing is in three steps: Time-reverse propagation of the sensed echoes, forward simulation of the wavefield emitted by the source element, and partial imaging by computing the zero-lag temporal correlation of the two wavefields to detect the scattering structures. The final reconstruction is obtained by stacking all the partial results on a single image. Laboratory tests, performed on experimental data acquired on a physical phantom, with and without an aberrant layer, prove the effectiveness of the proposed method even in the case of vertical and lateral velocity variations, with images of impressive spatial resolution and highly accurate target placement.