Avaliação de algoritimos para geração de imagens por ultrassom utilizando técnicas de beamformers adaptativos e sparse arrays

Abstract

The modern diagnosis ultrasound systems (US) based on the transmission of plane waves are capable to generate images with elevated frame rates allowing sophisticated exams and making some previously complex and risky procedures simple and non-invasive. The majority of US systems use the Delay and Sum (DAS) technique due to its easy implementation and low computational cost, although they provide low quality images. Adaptive beamformer techniques are capable of reconstructing images with elevate resolution and contrast, but with complex implementation and high processing cost. In this work is suggested the use of adaptive techniques such as Minimum Variance (MV), Generalized Sidelobe Canceler (GSC) and Eigenspace Generalized Sidelobe Canceler (EGSC) based on plane waves and combined with Sparse Arrays to evaluate the efficiency of these methods with decrease in the number of active elements of the transducer on reception, and consequent decrease in the amount of data generated. The evaluation of the proposed methods was performed using: a simulated dataset; actual data from acquisitions on an ultrasound biological tissue mimic phantom; and data collected in vivo. The tests were performed with 128 elements for transmission and 128, 65, 44 and 23 elements sparsely distributed at reception. Performance evaluation metrics have been done using the Full Width at Half Maximum (FWHM) for the verification of lateral/axial resolutions and the contrast ratio (CNR). The results showed that the images generated by the proposed method, with reduced number of active elements at reception, were close to those provided by the DAS in terms of spatial resolutions and contrast, indicating that the adaptive beamform methods combined with proposed sparse arrays are suitable for B-mode images reconstruction.