Validation of Optimal 2D Sparse Arrays in Focused Mode: Phantom Experiments

International audienceBackground, Motivation and ObjectiveIn parallel with the increasing interest for 3D ultrasound imaging, different design techniques have been investigated to find the best configurations of 2D sparse arrays to scan an entire volume of interest [Trucco, IEEE UFFC99], [Diarra, IEEE TBME13]. In particular, we recently addressed the issue of driving a full 2D array of 1024 elements with a reduced number of channels (128, 192 or 256): the optimal arrays (opti128, opti192 and opti256) were obtained using simulated annealing to sculpt the radiated wideband pressure field at multiple depths [Roux, IEEE UFFC17]. The aim of the present work is to experimentally validate these optimal configurations by performing 3D focused imaging on phantoms.Statement of Contribution/MethodsThe 1024 elements of a 3 MHz array made by Vermon were individually driven by four synchronized Verasonics Vantage 256 systems. The systems were programmed to transmit 3-cycle sine bursts at 3 MHz focused at 25 mm and to scan a 3D sector with span ±30° in 31×29 steered beams in azimuth and elevation, respectively. Six arrays were compared: the optimal arrays (opti128, opti192 and opti256), an array whose active elements where randomly selected (rand256) and two references (REF716 and REF1024). The circular dense array REF716 corresponds to the full array REF1024 array without the corner elements. The comparison criteria were the lateral resolution (full width at half maximum - FWHM) and the contrast to noise ratio (CNR), measured on the images obtained by scanning a Gammex (Sono410 SCG) and CIRS (054GS) phantoms respectively.Results/DiscussionThe results are reported in Table 1. The opti256 performs the best among all the compared sparse arrays because it presents the same resolution performance as the REF1024 and an acceptable loss of CNR while using only 25% of the active elements of REF1024. The REF716 array is very competitive with only 70% of the active elements of REF1024