Novel Dynamic Four-Dimensional CT Angiography Revealing 2-Type Motions of Cerebral Arteries
Background and Purpose—We developed a novel dynamic 4-dimensional CT angiography to accurately evaluate dynamics in cerebral aneurysm.
Methods—Dynamic 4-dimensional CT angiography achieved high-resolution 3-dimensional imaging with temporal resolution in a beating heart using dynamic scanning data sets reconstructed with a retrospective simulated R-R interval reconstruction algorithm.
Results—Movie artifacts disappeared on dynamic 4-dimensional CT angiography movies of 2 kinds of stationary phantoms (titanium clips and dry bone). In the virtual pulsating aneurysm model, pulsation on the dynamic 4-dimensional CT angiography movie resembled actual movement in terms of pulsation size. In a clinical study, dynamic 4-dimensional CT angiography showed 2-type motions: pulsation and anatomic positional changes of the cerebral artery.
Conclusions—This newly developed 4-dimensional visualizing technique may deliver some clues to clarify the pathophysiology of cerebral aneurysms.
Hemodynamic stress is 1 factor contributing to the initiation of cerebral aneurysm and the rupture process, but simple determination of aneurysm size and location never accounts for hemodynamic forces or the ability of aneurysms to withstand them.1 We recently succeeded in visualizing the dynamics of cerebral aneurysms with electrocardiogram-gated 4-dimensional CT angiography (4D-CTA) in clinical settings.2 However, application of electrocardiogram-gated 4D-CTA to assess the dynamics of aneurysm wall has shown certain limitations, the most significant of which was various visual artifacts.3,4 We therefore developed a novel 4-dimensional visualizing technique, dynamic 4D-CTA (DFA), to more accurately evaluate the dynamics of motion in cerebral aneurysms.
DFA movies were made using dynamic scanning data sets reconstructed with a retrospective simulated R-R interval reconstruction algorithm. The accuracy of DFA was evaluated using stationary and pulsating phantoms and in 4 patients with unruptured cerebral aneurysms (available at http://stroke.ahajournals.org).
Movie artifacts on DFA completely disappeared in a titanium clip phantom (Movie S1) and were much fewer in a dry bone phantom than electrocardiogram-gated 4D-CTA (Movie S2): the CT value was approximately 3000 and 500 Hounsfield units, respectively.
A DFA movie was constructed from 10 phases in 1 cardiac cycle and showed pulsations resembling actual movement of the virtual pulsating aneurysm model (approximately 250 to 300 Hounsfield units; Movie S3). The phase of 3-dimensional CT angiographic images was in the order of absolute time from the newest R wave (Figure 1A). Maximum surface area and volume were observed at the same time, at approximately the 25% R-R interval (Figure 1B). The size of pulsatile motion of the aneurysmal dome in an axial direction was approximately 0.7 mm, close to the real size of the virtual pulsating aneurysm model (Figure 1C).
The DFA movie appeared to depict real movement of arteries, not artifacts (Movie S4). Two types of movements were observed in intracranial arteries: pulsation of the artery itself, probably resulting from contraction and extension of the arterial wall during the cardiac cycle, and motion associated with anatomic positional change (ie, same amplitude for all regions; Movie S5). To clarify contraction and extension of arterial wall, changes in middle cerebral artery (MCA) volume/average MCA volume were analyzed (Figure 2A). In addition, to reveal the relationship between MCA volume variation and flow velocity in the MCA, the approximate curve of MCA volume variation was overlapped with flow velocity as determined by transcranial Doppler ultrasonography of the MCA. Peak MCA volume appeared after maximum flow velocity in the MCA (Figure 2B). Next, positional analysis was performed by the overlap imaging method of diastolic-phase (red) and systolic-phase (blue) DFA images (Figure 3A). The overlap image from an anterior view showed the artery as nearly red and a posterior view as blue (Figure 3B), indicating that the artery moved from an anterior position in the diastolic phase to posteriorly in the systolic phase. Intracranial arteries of both the anterior and posterior circulation moved similarly in another case (Figure 3C). However, occipital bone did not move like arteries (Figure 3D).
Recent reports have revealed the usefulness of 4D-CTA in evaluating cerebral aneurysms.2,5 To precisely evaluate the dynamics in cerebral aneurysm walls, some new techniques including electrocardiogram-gated 4D-CTA2 and the dynamic multiscan technique6 have been developed and shown improvements in the image quality. However, certain artifacts remain problematic although using these techniques (Movie S6). We believe that DFA is the best method for visualizing the dynamics of cerebral aneurysms without movie artifacts.
Yaghmai et al6 reported that artifacts on 4D-CTA result from a lack of cone beam correction. We supposed that start angles of the gantry might give rise to these artifacts even using dynamic multiscan techniques. Half reconstruction corresponding to gantry rotation time is the most important technique to obtain high temporal resolution for image reconstruction. In addition, use of the same start angle for the detector is necessary to reduce movie artifacts during dynamic scanning. Moreover, the temporal dimension in electrocardiogram-gated 4D-CTA is the average order of regular intervals within the R-R interval, whereas DFA uses a rearranged temporal order during the simulated R-R interval. To adopt these techniques, a simulated R-R interval reconstruction algorithm allows the acquisition of higher frame rates in 4D-CT movies without movie artifacts.
In a clinical setting, inconstant CT values of vascular enhancement also cause movie artifacts. We used a variable injection method to make variation of CT values in intracranial arteries on reconstruction images ≤25 Hounsfield units in all phases during the simulated R-R interval. As a result, clinical DFA provided high-quality movies without artifacts, revealing for the first time real 2-type movements of intracranial arteries.
DFA may determine the accurate amplitude of aneurysm wall pulsation, which would be related with the aneurysm wall fragility and the rupture process of cerebral aneurysms. Further studies may prove clinical usefulness of DFA such as identification of aneurysms with a high risk of rupture.
Sources of Funding
This work was supported by a grant-in-aid from the Japan Society for the Promotion of Science (C20591684) to S.M.
F.I. has the patents for dynamic 4-dimensional CT angiography (US 2008/0095307 A1, US Patent and Trademark Office; A000604213, Japan Patent Office).
We express our deep thanks to Mr Masatoshi Kanou (Toshiba Medical Systems Corporation, Japan) for his help in fulfilling this work.
The online-only Data Supplement is available at http://stroke.ahajournals.org/cgi/content/full/STROKEAHA.110.591008/DC1.
- Received July 13, 2010.
- Revision received September 10, 2010.
- Accepted September 29, 2010.
- © 2011 American Heart Association, Inc.
- Hayakawa M,
- Katada K,
- Anno H,
- Izumi S,
- Hayashi J,
- Irie K,
- Negoro M,
- Kato Y,
- Kanno T,
- Sano H