Tatsuya Nishii MD, Presenter: Nothing to Disclose

Atsushi K. Kono MD,PhD, Abstract Co-Author: Nothing to Disclose

Mizuho Nishio MD, PhD, Abstract Co-Author: Research Grant, Toshiba Corporation

Hiromi Hashimura MD, Abstract Co-Author: Nothing to Disclose

Noriyuki Negi RT, Abstract Co-Author: Nothing to Disclose

Atsushi Fujita, Abstract Co-Author: Nothing to Disclose

Junya Konishi, Abstract Co-Author: Bayer Pharma FUJIFILM RI Pharma Co. Ltd.

Eiji Kohmura MD, PhD, Abstract Co-Author: Nothing to Disclose

Kazuro Sugimura MD, PhD, Abstract Co-Author: Research Grant, Toshiba Corporation Research Grant, Koninklijke Philips NV Research Grant, Bayer AG Research Grant, Eisai Co, Ltd Research Grant, DAIICHI SANKYO Group

CT angiography (CTA) prior to digital subtraction angiography (DSA) is useful for detecting the feeders of spinal dural arteriovenous fistulas (SDAVFs). However, identifying these feeders is sometimes time-consuming because they are small and run close to osseous structures. Non-rigid registration provides precise subtraction that can compensate for organ motion or transformation between two datasets. Thus, we hypothesized that subtracted CTA imaging using non-rigid registration (R-CTA) would facilitate the diagnosis of SDAVF feeders. The aim of this study was to evaluate the utility of R-CTA in patients with SDAVFs in comparison with conventional CTA imaging (C-CTA).

The records of 10 consecutive patients (63±13 years old, 1 female) who had undergone CTA and DSA for clinically suspected SDAVFs were retrospectively reviewed. From repeated CTAs performed at the arterial and late-arterial phases, deformed images of the late-arterial phase were obtained using non-rigid registration adjusted to the arterial phase images. Registration was performed using open-source Advanced Normalization Tools software. Next, R-CTA images were obtained by subtracting the deformed images from the arterial phase images. Both R- and C-CTA were analyzed with DSA results as a reference standard. The time required for detecting the SDAVFs’ feeders, and their detectability, were analyzed for each patient. For each intervertebral foramen, the diagnosis likelihood of the feeders was scored on a 5-point scale (1=definitely negative, 5=definitely positive), and the accuracy was calculated. The difference between R-CTA and C-CTA was assessed by an ANOVA test and McNemar's Chi-square test.

The required time to detection, and the detectability of, feeders using C-CTA vs. R-CTA were 96.1±39.1s vs. 49.3±27.3s (P<0.01), and 60% vs. 80% (P=0.15), respectively. When a feeder was scored ≥4, the accuracy in C-CTA vs. R-CTA was 94.5% (189/200) vs. 97.5% (195/200) (P=0.01), respectively.

R-CTA reduced the time required for detecting SDAVF feeders, and had better accuracy than C-CTA.

Our subtracted CTA imaging technique using non-rigid registration helps clinicians to assess SDAVF feeders more quickly and accurately than the conventional method.

Nishii, T, Kono, A, Nishio, M, Hashimura, H, Negi, N, Fujita, A, Konishi, J, Kohmura, E, Sugimura, K, Novel Subtracted CT Angiography Imaging Using Non-rigid Registration for Better Visualization of Spinal Dural Arteriovenous Fistulas.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14000851.html