Application of a 2D interpolation technique to reduce the number of steering angles required for high-quality ultrasound images in plane-wave compounding

Plane wave imaging (PWI) can reach frame rates in the kHz range. However, PWI suffers from reduced image quality due to the lack of focusing. To tackle this limitation, plane wave compounding (PWC) was introduced. PWC is based on recombining multiple PWI images generated using different steering angles. Alas, the number of steering angles required to achieve reasonable image quality significantly reduces the frame rate. In this study, we propose a two-dimensional (2D) interpolation technique, based on radial basis functions, which allows reducing the number of steering angles in PWC without degrading image quality. The idea is to reduce the number of steering angles and then apply 2D interpolation, along the angle dimension and directly to the RF data, to reconstruct the data related to those angles for which data were not collected. The full dataset is then utilised to generate an image. To compare this technique with standard PWC, we utilised the dataset from the Plane-wave Imaging Challenge in Medical UltraSound (PICMUS). Results showed that the number of steering angles can be reduced by factor 3, thus tripling the frame rate, while achieving comparable performance with standard PWC (with 75 angles) in terms of contrast and spatial-resolution.