Angular Apodization Estimation Using Independent Component Analysis in Coherent Plane-Wave Compounding

Angular apodization in Coherent Plane-Wave Compounding (CPWC) entails estimation of a set of weights assigned to different angles before summation. Adaptive methods are used to estimate the weights from echo traces acquired from several insonifications at different angles. Herein, the angular apodization estimation is formulated as a blind source separation problem. More specifically, the received signal from each insonification (with a specific angle) has poor resolution and contrast, and is considered as a non-independent observation of the field. The angular weights can then be found by estimating the independent component from the observations. The results on the Plane-wave Imaging Challenge in Medical UltraSound (PICMUS) dataset confirm the improvement in the lateral resolution and contrast compared with the non-adaptive angular apodization.