Fast interpolation algorithms for real-time three-dimensional cardiac ultrasound

Real-time three-dimensional (RT3D) ultrasound technique based on matrix phased array transducers is likely to become predominant for dynamic screening in cardiac and obstetric practice. With these transducers large data volumes are acquired in spherical coordinates and require resampling to be visualized in cartesian coordinates. A fast 3D resampling method was implemented and five interpolation kernels tested on cardiac RT3D data. Downsizing and smoothing of sampling artifacts were integrated in the resampling process for improvement of the reconstructed data visual quality.