A constrained adaptive beamformer for medical ultrasound: initial results

Adaptive beamforming has been widely used as a way to improve image quality in medical ultrasound applications by correcting phase and amplitude aberration errors resulting from tissue inhomogeneity. A less-studied concern in ultrasound beamforming is the deleterious contribution of bright off-axis targets. This paper describes a new approach, the constrained adaptive beamformer (CAB), which builds on classic array processing methods from radar and sonar. Given a desired frequency response for the mainlobe beam, the CAB reduces off-axis signals by imposing an optimal set of weights on the receive aperture. A linearly constrained adaptive filter dynamically adjusts the aperture weights in response to the incoming data. Initial results show a factor of two improvement in point target resolution and a 60% contrast improvement for low echogenicity cysts. The CAB could considerably improve cardiac and abdominal image quality. We address implementation issues and discuss future work.