Ajay Wakhloo

Matthew J. Gounis

C-arm-based conebeam CT (CBCT) has offered tremendous advances in interventional neuroradiology (reviewed¹). One such important development was the ability to simultaneously image cerebrovascular stents and the vessel lumen that provides critical information regarding the relationship of the device to the host artery.² Neurovascular stents, and more recently introduced flow diverters, are challenging targets to image with x-ray-based techniques due to the device materials (eg, nitinol) and the exceptionally small features (as small as 30 μm). To image the vessel lumen and implant in a single acquisition further complicates the problem because the devices that are poorly radio-opaque can easily be obscured with standard contrast injections. Clinically, proper apposition of the implant to the vessel wall is important to avoid thromboembolic events³ and prevent endoleaks in the case of flow diverters.

The first generation of CBCT used the entire detector so as to avoid axial truncation.⁴ Due to restricted data rates and to reduce image noise, the pixels were binned (2×2). Pixel binning also served to reduce the information for filtered back-projection, thereby offering reasonable reconstruction times. Our goal was to develop higher resolution CBCT imaging that ultimately enables visualization of each device strut and to evaluate with high fidelity their apposition to the artery without compromising clinical workflow. To achieve this objective, we reduced the detector format and performed the reconstruction without pixel binning. The truncation problem is resolved using volume of interest tomography.⁵ Combining higher magnification with full-scale reconstruction, our initial phantom experiments demonstrated that we could reconstruct each stent strut. We then optimized the injection protocol in vivo such that we could perform contrast-enhanced CBCT without obscuring the device. A standard practice at the time was to slowly inject contrast, which reduces the quality of the reconstructed dataset due to incomplete opacification of the vessel throughout the cardiac cycle. To overcome these limitations, we diluted the contrast agent in saline and injected at the standard rate that ensures constant filling during the entire CBCT acquisition. Ultimately we found optimal imaging and contrast injection protocols for the assessment of the implant and its relationship to the vessel.

Today, the technology has been commercialized (VasoCT, Philips Healthcare, Best, the Netherlands), and every patient at our institution receiving a stent or flow diverter is imaged with this technique. Depending on the procedure, this imaging protocol is used to determine if additional measures to improve device apposition are necessary (eg, angioplasty).

We have recently quantified the accuracy of the technique as compared with reference-standard histomorphometry for the detection of tissue growth over the implants.⁶ We subsequently found that the technique is capable of imaging microarteriovenous malformations, and we have used these 3D datasets registered with MRI and CT for radiation treatment.⁷ In our research lab, we have employed the technology to directly visualize the location of small silica cannulae for targeted convection-enhanced delivery of gene therapy for the treatment of neurodegenerative disease.⁸ Future work will focus on making the acquisition and data visualization more user-friendly; and further exploration of IV contrast administration will reduce the invasiveness of follow-up imaging. In an era of expanding use of flow diverters, we aspire to see more widescale adoption of the technique, to aid interventionalists in achieving critically important device apposition to the arterial wall.⁹

References

1. Kamran M, Nagaraja S, Byrne JV. C-arm flat detector computed tomography: the technique and its applications in interventional neuro-radiology. Neuroradiology 2010;52:319–27. doi: 10.1007/s00234-009-0609-5
2. Benndorf G, Strother CM, Claus B, et al. Angiographic CT in cerebrovascular stenting. AJNR Am J Neuroradiol 2005;26:1813–18
3. Heller R, Calnan DR, Lanfranchi M, et al. Incomplete stent apposition in Enterprise stent-mediated coiling of aneurysms: persistence over time and risk of delayed ischemic events. J Neurosurg 2013;118:1014–22. doi: 10.3171/2013.2.JNS121427
4. Kudo H, Noo F, Defrise M. Cone-beam filtered-back projection algorithm for truncated helical data. Phys Med Biol 1998;43:2885–2909. doi: 10.1088/0031-9155/43/10/016
5. Wiegert J, Bertram M, Netsch T, et al. Projection extension for region of interest imaging in cone-beam CT. Acad Radiol 2005;12:1010–23. doi: 10.1016/j.acra.2005.04.017
6. Flood TF, van der Bom IMJ, Strittmatter L, et al. P-016 Quantitative assessment of stent induced neointimal hyperplasia with contrast enhanced cone-beam CT: in vivo validation with histomorphometry. J NeuroIntervent Surg 2013;5:A26–27. doi: 10.1136/neurintsurg-2013-010870.48
7. van der Bom IMJ, Gounis MJ, Ding L, et al. Target delineation for radiosurgery of a small brain arteriovenous malformation using high-resolution contrast-enhanced cone beam CT. BMJ Case Reports 2013; Epub ahead of print. doi: 10.1136/bcr-2013-010763
8. van der Bom IMJ, Moser RP, Gao G, et al. Frameless multimodal image guidance of localized convection-enhanced delivery of therapeutics in the brain. J NeuroIntervent Surg 2011;5:69–72. doi: 10.1136/neurintsurg-2011-010170
9. Kizilkilic O, Kocer N, Metaxas GE, et al. Utility of VasoCT in the treatment of intracranial aneurysm with flow-diverter stents. J Neurosurg 2012;117:45–49. doi: 10.3171/2012.4.JNS111660

Read this article at AJNR.org . . .