Standardized ultrasound evaluation of carotid stenosis for clinical trials: University of Washington Ultrasound Reading Center.

INTRODUCTION: Serial monitoring of patients participating in clinical trials of carotid artery therapy requires noninvasive precision methods that are inexpensive, safe and widely available. Noninvasive ultrasonic duplex Doppler velocimetry provides a precision method that can be used for recruitment qualification, pre-treatment classification and post treatment surveillance for remodeling and restenosis. The University of Washington Ultrasound Reading Center (UWURC) provides a uniform examination protocol and interpretation of duplex Doppler velocity measurements. METHODS: Doppler waveforms from 6 locations along the common carotid and internal carotid artery path to the brain plus the external carotid and vertebral arteries on each side using a Doppler examination angle of 60 degrees are evaluated. The UWURC verifies all measurements against the images and waveforms for the database, which includes pre-procedure, post-procedure and annual follow-up examinations. Doppler angle alignment errors greater than 3 degrees and Doppler velocity measurement errors greater than 0.05 m/s are corrected. RESULTS: Angle adjusted Doppler velocity measurements produce higher values when higher Doppler examination angles are used. The definition of peak systolic velocity varies between examiners when spectral broadening due to turbulence is present. Examples of measurements are shown. DISCUSSION: Although ultrasonic duplex Doppler methods are widely used in carotid artery diagnosis, there is disagreement about how the examinations should be performed and how the results should be validated. In clinical trails, a centralized reading center can unify the methods. Because the goals of research examinations are different from those of clinical examinations, screening and diagnostic clinical examinations may require fewer velocity measurements.

Measurement of anastomosis geometry in lower extremity bypass grafts with 3-D ultrasound imaging.

The attachment sites of lower extremity bypass grafts are known to exhibit a wide range of geometries. Factors that determine the geometry of a given anastomosis include graft material, graft site, native vessel size, graft size and individual patient anatomy. Therefore, it is difficult to specify a standard anastomosis geometry before surgery and difficult to predict the effect of the geometry on long-term graft patency. We have used 3-D ultrasound imaging to study 46 proximal anastomoses of lower limb bypass grafts. We have developed methods to characterize the 3-D geometry of the anastomosis in terms of component sizes and angles. These detailed geometric measurements describe a range of anastomosis geometries and establish standardized parameters across cases that can be used to relate anastomosis geometry to outcome.

Display of spatially-registered Doppler spectral waveforms and three-dimensional vein graft geometry.

A method has been developed to display Doppler spectral waveforms in lower extremity vein grafts in conjunction with 3-D vessel geometry. Doppler spectral waveforms and cross-sectional images of the vein graft are collected with a custom 3-D ultrasound imaging system. Computer processing generates a display of the Doppler sample volumes registered in 3-D space with a surface reconstruction of the vein graft lumen. An interactive computer interface displays spectral waveforms at user-selected sites in the graft. Summary displays combining spectral waveforms, maximum velocity and cross-sectional area provide a pictorial record of the state of the vein graft along its full length. The method is demonstrated for two patient studies, each at two time points after graft revisions. The graphic display of both hemodynamics and geometry allows rapid assessment of vein graft changes over time.

Remodeling in peripheral vein graft revisions: serial study with three-dimensional ultrasound imaging.

OBJECTIVE: Remodeling of vein grafts in the lower limb can lead to stenotic lesions that threaten long-term graft patency. Progressive changes in vein graft geometry were measured at sites of repaired stenoses with three-dimensional (3D) ultrasound imaging. METHODS: Ten vein graft revisions with patch angioplasty were followed up for 31 to 47 weeks. Four revisions were at valve sites, and six were at sites of diffuse intimal hyperplasia. Sets of spatially registered two-dimensional (2D) cross-sectional ultrasound images were assembled to create 3D computer models of each vein graft. Cross-sectional area measurements in planes normal to the vessel center axis were calculated from the 3D surface reconstructions. Data sets from serial studies were registered in a common coordinate system, and cross-sectional area measurements were compared at matched sites. RESULTS: Three of the four vein graft revisions at valve sites changed by less than 18%, and one decreased in cross-sectional area by 61%. Five of the six revisions at sites of diffuse intimal hyperplasia demonstrated significant decreases in lumen area ranging from 26% to 61%, and one revision exhibited no significant change in cross-sectional area. Reproducibility of the cross-sectional area measurements derived from the 3D imaging technique was 6.9%. CONCLUSIONS: Sequential area measurements from 3D ultrasound scans demonstrated different remodeling patterns and rates of change among revision sites within the vein grafts. Lumen narrowing documented with 3D scanning was not associated with consistent flow velocity changes on conventional duplex graft surveillance scans.