3D spatial compounding of ultrasound images using image-based nonrigid registration

Medical ultrasound images are often distorted enough to significantly limit resolution during compounding (i.e., summation of images from multiple vie ws). A new, volumetric image registration technique has been used successfu lly to enable high spatial resolution in three-dimensional (3D) spatial com pounding of ultrasound images. Volumetric ultrasound data were acquired by scanning a linear matrix array probe in the elevational direction in a foca l lesion phantom and in a breast in vitro. To obtain partly uncorrelated vi ews, the volume of interest was scanned at five different transducer tilt a ngles separated by 4 degrees to 6 degrees. Pairs of separate views were reg istered by an automatic procedure based on a mutual information metric, usi ng global full affine and thin-plate spline warping transformations. Regist ration accuracy was analyzed automatically in the phantom data, and manuall y in vivo, yielding average registration errors of 0.31 mm and 0.65 mm, res pectively. In the vicinity of the warping control points, registrations obt ained with warping transformations were significantly more accurate than fu ll affine registrations. Compounded images displayed the expected reduction in speckle noise and increase in contrast-to-noise ratio (CNR), as well as better delineation of connective tissues and reduced shadowing. Compoundin g also revealed some apparent low contrast lobulations that were not visibl e in the single-sweep images. Given expected algorithmic and hardware enhan cements, nonrigid, image-based registration shows great promise for reducin g tissue motion and refraction artifacts in 3D spatial compounding. (C) 200 1 World Federation for Ultrasound in Medicine & Biology.