Low-complexity adaptive beamforming for acoustic pulse-echo imaging

For years, the research community has viewed adaptive beamforming as a promising approach to enhance image quality in pulse-echo imaging applications such as medical ultrasound and sonar. Methods like the minimum-variance beamformer offer the potential for improved resolution and contrast, enabling sharper representation of key features in the image. However, a significant challenge lies in the high computational cost, stemming from the construction and manipulation of the spatial covariance matrix, which is typically calculated on a per-pixel basis. To address this, several low-complexity methods have been developed that retain the essential benefits of the minimum-variance beamformer while achieving computational efficiency suitable for real-time applications. A fundamental aspect is how to universally apply such beamformers, as the pulse-echo nature allows double adaptiveness over the transmit and/or receive beamforming steps. This presentation provides an overview of the theory and practical applications of these approaches, drawing on over 15 years of research at the University of Oslo and elsewhere. We start with the foundational principles of the minimum-variance beamformer, discuss the motivation for transitioning to low-complexity adaptive beamforming, assess the performance in real-world scenarios, and conclude with recent findings and future directions for this field.