Recent advances in geometrically correct 3D IVUS reconstruction enable the depiction of the true coronary anatomy by combining IVUS data with biplane angiographic images. Further development of the existing 3D IVUS reconstruction software, in conjunction with advancing hardware capabilities, is expected to allow the implementation of real-time 3D IVUS reconstruction within the catheterization laboratories.
TopCurrent Ivus Imaging Deficiencies
While IVUS allows accurate in vivo visualization of the vascular lumen and wall, (Mintz, et al., 2001), (Tobis, et al., 1991), (Yock, Linker, & Angelsen, 1989), (Nissen, et al., 1990), (Liebson & Klein, 1992) it has certain inherent technical limitations. Being a tomographic technique, IVUS produces single plane two-dimensional images, which are cross-sectional to the longitudinal catheter axis. Thus, it provides no direct spatial vascular information and does not clearly depict anatomical relationships (Roelandt, et al., 1994). In an attempt to overcome these deficiencies, the concept of three-dimensional (3D) IVUS reconstruction has emerged as the method of choice for imaging the real coronary anatomy and it is anticipated to extend the diagnostic as well as the interventional potential in the cardiac catheterization laboratory (Rosenfield, et al., 1991), (Klingensmith, et al., 2003).
Linear Versus Geometrically Correct Reconstruction
Traditionally, 3D IVUS reconstruction has been performed in a linear manner by stacking adjacent IVUS frames (von Birgelen, et al., 1997). However, this approach does not take into account the vessel curvature and the axial movements of the IVUS catheter during the pullback, thus the reconstructed model is of limited precision. On the other hand, geometrically correct 3D IVUS reconstruction combines IVUS data with biplane angiographic images (views of the same vessel taken in perpendicular planes) and therefore allows depiction of the spatial trajectory of the IVUS pullback trail (Bourantas, et al., 2005). The latter serves as a “backbone” on which the two-dimensional IVUS tomographic images are aligned and orientated (Coskun, et al., 2003), (Cothren, Shekhar, Tuzcu, Nissen, Cornhill, & Vince, 2000), (Giannoglou, et al., 2006), (Chatzizisis, et al., 2008). The final outcome is a realistic representation of the 3D coronary anatomy which can be used for further analyses (Bourantas, et al., 2008).
Processing Steps For Geometrically Correct 3D Ivus Reconstruction
The methodology for the geometrically correct 3D IVUS reconstruction can be divided in the following discrete steps:
From each of the two perpendicular angiographic projections taken at the initiation of the IVUS procedure, i.e. right anterior oblique (RAO) 30° and left anterior oblique (LAO) 60° a single frame is selected corresponding to the appropriate phase of the cardiac cycle, as determined by simultaneously recorded electrocardiogram (ECG). Generally, frames corresponding to the peak of the R-wave refer to the end-diastolic phase, while frames recorded at the end of the T-wave refer to the end-systolic phase. These images are manually processed for the detection of the IVUS catheter and vessel borders. On the basis of the course of the IVUS catheter in each angiographic projection, the 3D trajectory of the pullback is then reconstructed (Giannoglou, et al., 2006).