This study showed that the use of fusion image guidance is associated with significant reductions in overall operative time, contrast medium volume, time and radiation dose for fluoroscopy in ISF-TEVAR. In addition, it demonstrated that perfect fusion accuracy was achieved in 8 out of 15 cases (44%), with a satisfactory deviation in most patients. The key advantage of fusion guidance is to provide safe and accurate perforation localization, which is strong evidence to ensure successful fenestration. To the best of our knowledge, this is the first comparative study in which this image fusion technique is used for ISF-TEVAR.
Previous studies have demonstrated the feasibility and accuracy of fusion imaging for standard TEVAR procedures without LSA revascularization16,18,25. Performing a subgroup analysis, Dias et al. first showed in a comparative study28 that image fusion guidance could significantly reduce the total radiation dose in standard TEVAR procedures, even in only ten patient samples. Subsequently, Hiraoka et al. further demonstrated that significant reductions in exposure and contrast medium could be achieved by the use of image fusion in standard TEVAR procedures involving larger populations29. For complex aortic disease, image fusion guidance has been primarily described in endovascular repair procedures. A meta-analysis suggested that image fusion could significantly lower the contrast volume, fluoroscopy time and operative time in complex (fenestrated/branched) EVAR24. Regarding ISF, Leger et al. first reported a cohort study demonstrating the feasibility and benefit of ISF for complex EVAR guided by 3D image fusion. Image fusion guidance has rarely been used for ISF-TEVAR requiring LSA revascularization. By providing real-time visualization of 3D vascular structure, the current comparative study indicated that exposure and operative time could also be reduced with ISF-TEVAR.
Problems associated with inaccuracies have been thoroughly discussed in Schulz’s17 and Sailer’s22 studies. The reasons could be grouped into three main categories: different patient positions between CTA and intraoperative CBCT, patient movement, and straightening and deformation of vessels. In response to the first problem, CTA acquisition needs to be performed with normal breathing followed by breath holding, which is similar to the case of general anaesthesia with low tidal volume ventilation. Additionally, patient repositioning was avoided after CBCT acquisition to eliminate inaccuracies in patient movement. The most challenging inaccuracies originate from deformations caused by the insertion of stiff devices, including guidewires and introducers. In our centre, we proposed a novel registration method that aligns CTA and CBCT images based on contours of soft tissue or calcified tissue landmarks within the region of interest rather than the traditionally used bony landmarks. A closer alignment to the region of interest could reduce the mismatch caused by respiratory diseases, especially for cases such as TEVAR procedures. In addition, multiplanar registration brought to a more accurate three-dimensional alignment instead of two-dimensional manual adjustment after overlaying. Therefore, this registration method greatly improved the accuracy of fusion. To better understand the improvement of soft tissue-based method compared to traditional bony-based method, we measured the mismatch reduced of the center of LSA ostium after using soft-tissue based method. Through computing the difference of LSA center deviation in preoperative CTA and intraoperative CBCT imaging between these two methods (Fig. 5), the median deviation was 19.7 mm (range 13.0–28.6), which indicated the mismatch corrected by soft tissue-based method. Image contour extraction, synthesis of reconstruction and distance measurement were performed using ImageJ (version 1.48v, National Institutes of Health). Compared to the accuracy reported in Schulz’s17 study for the standard TEVAR procedures without LSA revascularization, the median deviation in our study on ISF-TEVAR was eight times more accurate (1.45 mm, range 0.0-8.4 vs. 11.7 mm, range 0.0-37.2). The main reason could be that the 3D-3D registration method used in this study was based on soft tissue instead of bony structures.
Fusion-guided ISF-TEVAR may provide several benefits for operators. i) By reconstructing and overlaying the 3D vascular anatomy of the region of interest, image fusion guidance could help operators increase the visibility of important structures. Instead of the guidance of traditional 2D angiograms, which require multiple rounds of confirmation from different angles, the guidewires and devices can be guided in a more intuitive way through 3D visualization. Accordingly, fluoroscopy time could be significantly reduced by accelerating the process of fenestration. ii) The overlaid vascular anatomy provided visualization of the true and false lumen, which could facilitate introducing the guidewire into the true lumen and reduce the disastrous risk if mispuncture occurred to the false lumen. In this way, the contrast medium for confirming the position of devices could also be reduced significantly. iii) Image fusion could assist operators in obtaining the optimal working angle under fluoroscopy and angiography in a simpler manner and in less time. Determining the optimal working angle is key to precise stent deployment and perforation. An optimal working angle satisfied the following criteria: 1) the LSA and LCCA could be fully visualized, which helps the stent to deploy accurately, and 2) the plane of the LSA ostia is on top of the aortic arch such that the perforation site is displayed in a more intuitive way. In this way, image fusion with a three-dimensional aortic arch display provided a simpler way to find the optimal working angle, thereby reducing the unnecessary radiation dose for both patients and operators. iv) The image fusion technique is especially beneficial to novice operators when intuitive visualization can advance their understanding of DSA imaging with anatomy, potentially hastening the learning curve of the procedure. However, the relatively small sample size of the current study is also a limitation, and prospective randomized controlled trials with greater numbers of cases are needed in the future.