Masked Autoencoders for Ultrasound Signals: Robust Representation Learning for Downstream Applications