Enhancement of Compounded Ultrasound Images with Spatial Filtering

摘要

Compound ultrasound (US) images benefit from reduced speckle noise, at the expense of blurring. The purpose of this work was to evaluate several different methods for reducing the effects of blurring on compounded ultrasound images. We evaluated 3 different approaches: unsharp masking (USM), Gaussian deconvolution (GD), and system measured point-spread-function (PSF) deconvolution (PD). The compounded images are created from summation of approximately 300 B-Mode US images. We applied the different de-blurring methods to these images and examined the images subjectively as well as measured resolution with an autocorrelation metric. Overall, these methods improved resolution by 1.25 (USM), 1.09 (GD), and 1.27 (PD). Unsharp masking was the best trade-off between runtime performance and image quality.