P3G-3 The "N-Shot" Maximum Brightness Algorithm: An Efficient and Robust Delay-Aberration Estimator

Abstract

The iterative "maximum brightness" delay aberration correction algorithm searches on each array element for the beamformer delay that maximizes the spatial-average speckle brightness. The array elements are optimized in turn until the entire aperture is corrected. A minimum of two transmits per line in the ROI (two "shots") are required for each element but the actual number is typically much greater, especially when small delay steps are employed. Furthermore, thermal or motioninduced noise can introduce false, transient maxima. These often cause the iterative search to halt prematurely at incorrect delay values, a problem which is also exacerbated as the step size is reduced. This paper describes a new family of algorithms, termed "N-shot", in which the number of shots is constant and small (as low as N = 2). All are based on the empirical observation that the average speckle brightness exhibits a nearly sinusoidal variation with respect to any one element's beamformer delay. Application of a priori knowledge allows the phase of the sinusoid, and hence the delay aberration value for a given element, to be determined with only two shots. An additional shot allows brightness trends due to "motion noise" to be estimated and removed. Appropriate separation of the delays used for the shots reduces susceptibility to thermal noise. Compared to the iterative algorithm, this new approach is at least as fast and typically much faster, the number of transmits is deterministic and independent of delay quantization, and it has reduced susceptibility to thermal and motion noise. The effectiveness of the algorithm was tested experimentally using an artificial aberrator attached to a phased array. Comparisons between aberrated and corrected phantom images confirm the ability of the N-shot algorithm to determine delay aberrations with only two ROI firings for each array element.