General principle and optimization of magneto-acousto-electrical tomography based on image distortion evaluation

As a novel non-invasive multi-physics imaging methodology, the magneto-acousto-electrical (MAE) technology is capable of detecting electric conductivity changes for biological tissues, exhibiting prosperous perspectives in medical applications. However, the acoustic beam was often simplified to a straight line or a focused one, being perpendicular to layered boundaries of tissues in previous studies. Linear-scanning measurements were carried out to reconstruct B-mode MAE images for layered models without considering the radiation pattern of transducers. Obvious image distortions in both shape and brightness were observed in experiments without any reasonable explanation.

This study aims to establish a general physical model for MAE measurements and solve the problem of B-mode image distortion, and hence provide theoretical and technical supports for the improvement of MAE imaging in practical applications.

By considering the radiation pattern of actual transducers and the inclined angle of electric conductivity boundaries, a general principal of MAE measurements applicable for objects of arbitrary shapes is proposed based on the theories of acoustic radiation, Hall Effect and electrical detection. The influences of inclined conductivity boundaries and transducer directivities are numerically analyzed with Matlab programming and also demonstrated by experimental measurements. To evaluate the degree of B-mode image distortion, the deformation length (3 dB amplitude decrease) of approximate straight lines for a circular model is defined as L = dtan(β_m /2), with d and β_m being the measurement distance and the half radiation angle of the main-lobe, respectively. The rotary-scanning-based MAE tomography (MAET) is employed to reduce the image distortion, and the rotation angle step is further optimized based on the criterion of the boundary radius fluctuation coefficient <0.01 mm.

The experimental results of MAE signals and B-mode images as well as MAETs show good agreements with simulations. It is demonstrated that, as the increase of the inclined angle, the MAE decreases rapidly with an extended time interval and reaches the 20 dB amplitude decrease when the angle exceeds 12°. Meanwhile, the deformation length of B-mode MAE imaging increases with the increase of the radiation angle for the transducer with a weaker radiation pattern, and hence results in a more serious image distortion. A smaller rotation angle step should be adopted to the MAET system with a longer deformation length, and the optimized maximum angle step of 12° is also achieved for the omnidirectional radiation of point sources with a long deformation length.

The image distortion is originated from the amplitude decrease, the time shift and the time interval expansion of MAE signals introduced by the deformation length and the incident angle. The favorable results demonstrate that the fast high-resolution imaging can be accomplished by the minimum rotations of the rotary-scanning-based MAET using an actual transducer, and also provide an optimized scheme for the rotary-based MAET without scanning using a linear array of point sources.