The influence of speckle statistics on contrast metrics in ultrasound imaging

Quality assessment of ultrasound images is difficult since image quality is subjective to the human observer. Nevertheless, image quality metrics are imperative when benchmarking different beamforming techniques. However, if we do not know how a beamformer alters an image, a quality metric might give an incorrect measurement of the image quality. Using the standard Delay-And-Sum (DAS) beamformer as a reference, we examine many of the most popular adaptive beamformers presented in literature; Capon's Minimum Variance (MV), Eigenspace Based Minimum Variance (EBMV), Coherence Factor (CF), Delay-Multiply-And-Sum (DMAS), Generalized Coherence Factor (GCF), and Phase Coherence Factor (PCF). We show that the speckle statistics for some of these adaptive beamformers are dependent on reflection amplitude, resulting in contradicting measurements of contrast metrics.