Vein grafts are placed to bypass diseased arteries in the lower limb when impaired circulation causes such symptoms as claudication and tissue necrosis. Frequent surveillance of vein grafts is recommended since early detection and revision of lesions improves the likelihood of long-term graft patency. The current clinical standards for vein graft monitoring are arteriograms and conventional ultrasound imaging. Both of these techniques, however, produce only two-dimensional (2D) views of vessels. Lesions at sites of complex geometry are difficult to assess with 2D methods, and spatial relationships over time are not preserved. Therefore, we are developing three-dimensional (3D) ultrasound imaging methods to monitor lower-limb grafts. Three-dimensional imaging can produce a complete representation of the vessel geometry, allowing measurement of progressive changes at specific sites within the graft.
Our 3D ultrasound imaging system is based on a magnetic tracking system integrated with a conventional ultrasound imaging unit. A series of cross-sectional images are collected in a region of interest while the tracking system records the location and orientation of the ultrasound scanhead. ECG triggering ensures acquisition at the same phase of the cardiac cycle. Computer reconstructions of the vessel are derived from the 2D images after they are registered in a 3D reference coordinate system. Interactive visualization software allows the user to freely rotate the reconstructions to view the 3D geometry of the vessel.
Surface reconstructions can be generated from manual outlines on a series of images, or the 2D image data can be reconstructed directly by inserting image pixels into a regular 3D volume grid. Using the volume method 3D reconstructions can be generated without user interaction, saving considerable analysis time. Volume data sets also lend themselves to flexible visualization and analysis techniques, including orthogonal and arbitrary slicing, contour and surface extraction, and volume rendering. In the case of power Doppler imaging, volume reconstructions also preserve the amplitude information of the signal.
Of particular clinical interest is the quantification of vein graft remodeling. We have measured cross-sectional area changes over time in several subjects who have undergone vein patch angioplasty to treat a developing stenosis within a graft. Patches have been observed at time points ranging from 2 weeks to 8 months post-operative. The regions of interest are registered across visits by matching external fiducial points on the limb; changes in cross-sectional area are then compared at corresponding locations along the length of the patch. For example, one subject examined at 2 weeks and 30 weeks post-operative demonstrated cross-sectional area reduction ranging from 30% at the distal end to 70% at the proximal end of the patch. Studies such as these can provide a measure of vein graft remodeling rate at specific locations within the graft. Quantitative monitoring of vein graft morphology may enable us to distinguish normal remodeling from pathologic changes which lead to re-stenosis.