Acta Neurochir (2014) 156:505–514 DOI 10.1007/s00701-014-1996-x
CLINICAL ARTICLE - VASCULAR
Utility of 320-detector row CT for diagnosis and therapeutic strategy for paraclinoid and intracavernous aneurysms Satoshi Inoue & Kohkichi Hosoda & Atsushi Fujita & Yoshiharu Ohno & Masahiko Fujii & Kazuro Sugimura & Eiji Kohmura
Received: 27 October 2013 / Accepted: 9 January 2014 / Published online: 22 January 2014 # Springer-Verlag Wien 2014
Abstract Background The aim of this study was (1) to assess the diagnostic accuracy of 320-detector row computed tomography (CT) for paraclinoid and intracavernous aneurysms, and (2) to investigate whether this method provides sufficient information for surgery. Methods A total of 14 patients with 16 unruptured proximal ICA aneurysms underwent three-dimensional CT angiography (3D-CTA) fusion imaging, which was created by superimposing 3D-CT venography data and/or 3D-bone data onto 3D-CTA data using 320-detector row CT, magnetic resonance imaging (MRI), and 3D digital subtraction angiography (DSA). The images of each modality were assessed using intraoperative findings as the reference standard. Results All aneurysms were clearly visualized on 320detector row CT. Bone subtraction and arterio-venous discrimination were accurate. On 3D-CTA fusion images, 11 aneurysms were diagnosed as “extracavernous” and five as “intracavernous”. No discordance in aneurysm location between the 3D-CTA fusion images and the intraoperative findings was found. In contrast, discordance between MRI and intraoperative findings were found in five of the 16 cases (31 %), which was significantly more frequent than with 3D-CTA (p=0.043). The findings DSA, which was performed in nine patients, were also in excellent agreement with the intraoperative findings. However, 3D-CTA fusion imaging provided more comprehensive information, including venous and osseous structures, than 3D-DSA. The 320-detector row S. Inoue : K. Hosoda (*) : A. Fujita : E. Kohmura Department of Neurosurgery, Kobe University Graduate School of Medicine, 7-5-1 Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan e-mail: [email protected] Y. Ohno : M. Fujii : K. Sugimura Department of Radiology, Kobe University Graduate School of Medicine, Kobe, Japan
CTA after surgery demonstrated a clear relationship between the clip and aneurysmal neck with notably few artifacts, which suggested the utility of this modality for postoperative assessment. Conclusions The 320-detector row CT provided high accuracy for the diagnosis of paraclinoid and intracavernous aneurysms. This technique also provided comprehensive depiction of the aneurysms and surrounding structures. Therefore, this modality might be useful for the diagnosis of the paraclinoid and intracavernous aneurysms and for developing a surgical treatment plan. Keywords Area-detector computed tomography . Dural ring . Intracavernous aneurysm . Paraclinoid aneurysm . Three-dimensional CT angiography . Surgery
Introduction Recent advances in imaging techniques, including multidetector row computed tomography (MDCT), high-field magnetic resonance imaging (MRI), and three-dimensional (3D) rotational catheter angiography, have enabled more accurate diagnoses of cerebral aneurysms. However, some critical issues remain in the clinical evaluation of proximal internal carotid artery (ICA) aneurysms, such as paraclinoid aneurysms and intracavernous aneurysms. It is still difficult to visualize the relationship of an aneurysm to the adjacent dural, osseous, and venous structures, especially the distal dural ring (DDR), proximal dural ring (PDR), anterior clinoid process, optic strut, and cavernous sinus. Obscure visualization causes difficulty in differentiating extracavernous aneurysms from intracavernous aneurysms. This differentiation is critical for the management of proximal ICA aneurysms. Extracavernous aneurysms are thought to
506
carry a risk of subarachnoid hemorrhage and are usually treated surgically or endovascularly [9]. The relationship of an aneurysm to adjacent structures is also critical when considering neurosurgical management. One of the important issues in a clipping procedure is whether proximal control can be obtained without removing the anterior clinoid process, opening the cavernous sinus, and exposing the cervical ICA. It is still difficult to obtain the above-mentioned comprehensive information on proximal ICA aneurysms in a single examination with a conventional imaging technique. Although digital subtraction angiography (DSA) has been the gold standard for imaging cerebral aneurysms, it is an invasive technique [11, 15]. Three-Tesla (3 T) magnetic resonance angiography (MRA) provides high spatial resolution [7] and can depict dural structures, including the DDR [23, 24]. However, MRA does not provide osseous and venous information, and is sensitive to laminar flow-related artifacts, especially in large or giant aneurysms. Three-dimensional computed tomographic angiography (3D-CTA) has been used as an alternative to DSA for the preoperative examination of proximal ICA aneurysms [19]. In contrast with MRA and DSA, 3D-CTA provides both osseous and venous information, which is important when assessing the exact location of aneurysms. However, aneurysms are not always clearly demonstrated on 3D-CTA because they tend to be embedded in the osseous and venous structures [15]. To solve this problem, several techniques to subtract bone [2, 12, 13, 18, 19, 21, 26] and to separate arteries from surrounding venous structures have been reported [14], but the separation technique is still uncommon. Recently, a new generation of CT systems with 320detector row has become clinically available, which makes volumetric imaging possible [16]. Several scans of volumetric CT produce the time-resolved volume data for “dynamic fourdimensional CT angiography (dynamic 4D-CTA)” during the time course of injected contrast medium that passes the brain’s vasculature. Therefore, a single examination of dynamic 4DCTA consists of a series of continuous 3D images of the vasculature, which includes a non-contrast phase, arterial phase, early venous phase, and late venous phase. After the subtraction process, 3D images of the arterial, venous, and osseous structures can be acquired as a “3D-CTA image”, a “3D computed tomographic venography (3D-CTV) image”, and a “3D-bone image”, respectively. The obtained 3D images can be mutually superimposed on 3D workstations. We call this fusion image the “3D-CTA fusion image”. The objective of the current study was to assess the diagnostic accuracy of 320-detector row CT for paraclinoid aneurysms and intracavernous aneurysms using the intraoperative findings as the reference standard. We also investigated whether this method provides sufficient information for surgery.
Acta Neurochir (2014) 156:505–514
Methods Patients with proximal ICA aneurysms were prospectively recruited between July 2008 and January 2013 in our hospital. The inclusion criteria of the current study were unruptured paraclinoid and intracavernous aneurysms and surgical treatment for the aneurysms after a CTA. Exclusion criteria included any contraindications for contrast medium, ruptured aneurysms, coil embolization after CTA, and prior treatment for the aneurysms. The data were retrospectively analyzed. The institutional review board approved this study, and written informed consent was obtained from all patients. CT studies Dynamic CT examinations were performed with a 320detector row CT scanner (Aquilion ONE, Toshiba Medical Systems Corporation, Tokyo, Japan) using volumetric cine scanning without helical imaging. The technical parameters were as follows: 160×0.5 mm detector width; 0.25 mm reconstruction interval; 512×512 matrix; 180–240 mm field of view; tube voltage, 80 to 100 kV; tube current, 100 to 200 mAs; detector width, 80 mm; and total 20 scans. The scanning delay was automatically adjusted using a test injection method. Eighteen subsequent dynamic scans were acquired with continuous scanning at one scan rotation per 1 s, followed by two intermediate dynamic scans at one scan rotation per 5 s in the late venous phase. Twenty-five milliliters of non-ionic contrast material (370 mgI/ml) were injected into a right cubital vein at a rate of 5 ml/s, followed by 25 ml of saline flush. Post-processing The bone subtraction process was performed automatically on the CT console. Non-enhanced acquisition in the sequence of the 22 scans was used both as a bone image and as a mask image for bone subtraction. The obtained data were transferred to a 3D workstation (Zio M900; Ziosoft, Tokyo, Japan). Then, dynamic 4D-CTA images were generated using a volume rendering technique. From a series of 22 scans of dynamic 4D-CTA, we selected an arterial phase image that provided the best depiction of the aneurysm (3D-CTA) and a venous phase image that provided the best depiction of the cavernous sinus and other surrounding venous structures (3DCTV). Non-enhanced phase images were used to obtain 3Dbone images. The 3D-CTA fusion images were created by superimposing 3D-CTV data and/or 3D-bone data onto 3DCTA data. To create accurate 3D-CTV images, we selected and fused data from two different venous phase images from dynamic 4D-CTA if necessary. Placement of the regions of interest on the ICA or the cavernous sinus was not needed to obtain 3D-CTA and 3D-CTV images because we selected the
Acta Neurochir (2014) 156:505–514
most appropriate arterial and venous phases from the early arterial phase and the late venous phase in a series of dynamic 4D-CTAs. The 3D-CTA and 3D-CTA fusion image We identified each aneurysm and measured its size on 3DCTA. Based on the projection and location, each aneurysm was classified into one of six groups according to previous reports on paraclinoid aneurysms: anterior wall, ventral paraclinoid, true ophthalmic, carotid cave, transitional, or intracavernous [1, 9, 10]. Then, on the basis of the visualized relationship between the aneurysms and the cavernous sinus on 3D-CTA fusion images, the location of each aneurysm was classified as follows: “extracavernous”, in which the aneurysm neck and sac were located distal to and outside of the cavernous sinus; “transcavernous”, in which the aneurysm neck or sac was located partly in the cavernous sinus; or “intracavernous”, in which the aneurysm neck and sac were located inside the cavernous sinus. In addition, the optic strut was used as an anatomic landmark for the PDR [5, 6]. According to the relationship between the aneurysms and the optic struts on non-subtracted axial source images of 3D-CTA, the aneurysms were classified as follows: “distal to PDR” if the aneurysm neck and sac were located distal to the optic strut; “on PDR” if the aneurysm neck or sac straddled the optic strut; or “proximal to PDR” if the aneurysm neck and sac were located proximal to the optic strut. Thus, “distal to PDR” corresponds to “extracavernous”, “on PDR” to “transcavernous”, and “proximal to PDR” to “intracavernous” on 3D-CTA fusion images, respectively. MRI Studies MRI was performed using a 3T MR scanner (Achieva 3.0 T system with Quasar Dual gradients, Philips Healthcare, Best, the Netherlands). Three-dimensional time-of-flight MRA was performed with the following parameters: TR/TE/FA = 25 ms/ 3.45 ms/20º, FOV = 240 mm (RFOV = 90 %), slice thickness = 1.2 (−0.6) mm,matrix = 255 (P) × 400 (F) (reconstruction, 800×800), resolution = 0.94 mm (P) × 0.6 mm (F) (reconstruction, 0.3 mm×0.3 mm), number of slices = 150, SENSE factor = 2.2, NEX = 1, and scan time = 5 min. To locate the aneurysm, the origin of the ophthalmic artery was used as the anatomic landmark of the DDR according to the DSA classification [23]. A horizontal line passing through the origin of the ipsilateral ophthalmic artery was chosen. The aneurysms were classified as follows: “distal to DDR” if the aneurysm neck and sac were above the ophthalmic line; “on DDR” if the aneurysm neck or sac was straddling the ophthalmic line; or “proximal to DDR” if the aneurysm neck and sac were
507
beneath the ophthalmic line. If the ipsilateral ophthalmic artery was hypoplastic or not visible, we used the contralateral ophthalmic artery as a landmark. MR coronal T2-weighted balanced fast field-echo images (coronal T2-weighted bFFE images) were obtained with the above-referenced 3T MR scanner with the following parameters: TR/TE/FA = 6.2 ms/2.6 ms/45º, FOV = 160 mm (RFOV = 100 %), slice thickness = 1.0 (−0.5) mm, Matrix = 326 (P) × 256 (F) (reconstruction, 336×336), resolution = 0.49 mm (P) × 0.63 mm (F) (reconstruction, 0.48 mm×0.48 mm), number of slices = 100, NEX = 3, and scan time = 5 min. On the basis of the relationship between the aneurysm and DDR on MR coronal T2-weighted bFFE images, the aneurysms were classified as follows: “distal to DDR” if the aneurysm neck and sac were located above the DDR; “on DDR” if the aneurysm neck or sac was straddling the DDR; or “proximal to DDR” if the aneurysm neck and sac were located below the DDR [24]. DSA DSA was performed for only nine patients because the other five patients refused DSA. A standard biplane DSA unit with rotational capabilities and a matrix resolution of 1024×1024 (Artis Zee BA twin; Siemens Medical Solutions, Erlangen, Germany) was used. Non-ionic contrast medium (370 mg) was injected with a power injector. Selective angiograms consisted of one anteroposterior, one lateral, and one or two oblique views. The 3D-DSA acquisition was performed for each ICA with aneurysms. Assessment of CT, MRI, and DSA findings Two experienced neurosurgeons (SI and KH) and one neuroradiologist (MF) blinded to the clinical data independently reviewed the CT, MRI, and DSA data. They performed the initial evaluations independently, and disagreements regarding the final interpretations were resolved by consensus. Then, the interpretations were verified using the intraoperative findings as the reference standard. The Fisher's exact test was used for proportion analysis to compare the accuracy of each modality.
Results During the study period, 22 patients with 26 unruptured proximal ICA aneurysms underwent 320-detector row CT scanning. Fourteen patients with 16 aneurysms underwent surgery, which fulfilled the criteria described above. Thirteen patients were female and one was male. The mean age was 60.3 years and ranged from 44 to 72 years. Clinical and imaging data are summarized in Table 1.
72/F
65/F
4
5a
44/F
14
9
9
5
10
16
Ventral
Ventral
Ventral
Intracavernous
Ant wall
Ventral
Intracavernous
True ophth
Intracavernous
Intracavernous
Ant wall
Ventral
Ant wall
Cave
Cave
Intracavernous
Classification
Extracavernous
Extracavernous
Extracavernous
Intracavernous
Extracavernous
Extracavernous
Intracavernous
Extracavernous
Intracavernous
Intracavernous
Extracavernous
Extracavernous
Extracavernous
Extracavernous
Extracavernous
Intracavernous
Location (CS)
Distal to PDR
Distal to PDR
Distal to PDR
Proximal to PDR
Distal to PDR
Distal to PDR
Proximal to PDR
Distal to PDR
Proximal to PDR
Proximal to PDR
Distal to PDR
Distal to PDR
Distal to PDR
Distal to PDR
Distal to PDR
Proximal to PDR
3D-CTA location (OS)
7 15 10 6
− − − −
9
21a
−
−
2
−
9
5
−
−
7
−
25a
10
−
5
2
−
−
4
−
−
19a
Size (mm)
3T-MRI
−
Discordance on CT
Distal to DDR
Distal to DDR
Distal to DDR
Proximal to DDR
Distal to DDR
Distal to DDR
Proximal to DDR
Distal to DDR
Proximal to DDR
Proximal to DDR
Distal to DDR
Distal to DDR
Distal to DDR
Proximal to DDR
On DDR
Proximal to DDR
Location
−
−
−
−
+
−
−
−
−
−
+
−
+
+
+
−
Discordance on MRI
Ventral
Ventral
ND
Intracavernous
Ant wall
Ventral
Intracavernous
True ophth
Intracavernous
ND
ND
ND
ND
Intracavernous
DSA
Clipping
Clipping
Clipping
Bypass + trapping
Clipping
Clipping
Bypass + ligation
Clipping
Bypass + ligation
Exploration
Clipping
Clipping
Clipping
Exploration
Clipping
Bypass + ligation
Surgery
Extracavernous/distal to DDR/ventral
Extracavernous/distal to DDR/ventral
Extracavernous/distal to DDR/ventral
Intracavernous
Extracavernous/on DDR/ant wall
Extracavernous/distal to DDR/ventral
Intracavernous
Extracavernous/distal to DDR/true ophth
Intracavernous
Intracavernous
Extracavernous/on DDR/ant wall
Extracavernous/distal to DDR/ventral
Extracavernous/on DDR/ant wall
Extracavernous/distal to DDR/cave
Extracavernous/distal to DDR/cave
Intracavernous
Operative findings
a
Aneurysms depicted only partially on MRA
3T-MRI 3 Tesla magnetic resonance imaging, 3D-CTA three-dimensional computed tomographic angiography, Ant wall anterior wall, cave carotid cave, CS cavernous sinus, DDR distal dural ring, ND not done, OS optic strut, true ophth true ophthalmic, ventral ventral paraclinoid
70/F
13
24
10
58/F
72/F
9
48/F
70/F
8
11
58/M
7
12
6
50/F
6
7
3
20
73/F
5b
6
6
8
65/F
4
3
45/F
2a
19
2
54/F
1a
Size (mm)
3D-CTA fusion image
2b
Age (y)/Sex
Case
Table 1 Validation for the location of proximal ICA aneurysms CTA, MRI, and DSA with intraoperative findings
508 Acta Neurochir (2014) 156:505–514
Acta Neurochir (2014) 156:505–514
Ten of 11 “extracavernous (distal to PDR)” aneurysms were treated with surgical neck clipping, and one underwent surgical exploration only (case 2b). Four “intracavernous (proximal to PDR)” aneurysms were treated with bypass and ligation or trapping of the ICA because they were symptomatic. One intracavernous aneurysm coexisted with an extracavernous aneurysm. In these cases with intracavernous aneurysms, no intradural portion of the aneurysm was confirmed during surgery. All procedures were performed successfully, and no patients in this series suffered aneurysm rupture during the follow-up period.
509
aneurysms, one “on DDR” aneurysms, and six “proximal to DDR” aneurysms. Validation of CTA and MRI findings with intraoperative findings
Volume-rendered 3D-CTA, 3D-CTV, and 3D-bone images were obtained simultaneously in a single examination using dynamic 4D-CTA (Fig. 1a-c). All proximal ICA aneurysms could be clearly identified on 3D-CTA, even though they were surrounded by osseous or venous structures. The maximum diameter of the aneurysms on 3D-CTA ranged from 2.0 to 24 mm (mean, 9.6 mm). 3D-CTV clearly depicted the complex shape of the venous structures, including the cavernous sinus (Fig. 1b). The 3D-CTA fusion images were obtained by superimposing 3D-CTV images and 3D-bone images onto 3D-CTA images (Figs. 1d, 2b, and 3b). The 3D-CTA fusion images facilitated the recognition of complicated anatomic relationships of an aneurysm to the anterior clinoid process, the optic strut, and cavernous sinus. These images were in excellent agreement with the intraoperative views (Figs. 1, 2, and 3).
No discordance in aneurysm location was observed between the preoperative 320-detector row CT findings and intraoperative findings (Table 1). By contrast, discordance between the preoperative MRI findings and intraoperative findings were found in five of 16 (31.3 %) aneurysms (cases 2a, 2b, 3, 5a, 10 in Table 1), which was significantly more frequent than with 3D-CTA (p=0.043). Discordance of the aneurysm location tended to be more frequent for anterior wall aneurysms and carotid cave aneurysms. For example, case 5a (Fig. 1) and case 10 (Fig. 2) were diagnosed as an “extracavernous” aneurysm on 3D-CTA fusion images. In case 5a, a dent in the aneurysm dome suggesting the attachment of DDR was also visualized on 3D-CTA and confirmed during surgery (Fig. 1a, f). On MRI, they were diagnosed as a “distal to DDR” aneurysms because they appeared to arise distal to the ophthalmic artery (case 5a, Fig. 1e) or DDR (case 10, Fig. 2d). However, the intraoperative findings clearly demonstrated that both of the aneurysm necks straddled the DDR and that the clips needed to be applied to both the neck of the aneurysm and the DDR (Fig. 2e and f). The final diagnosis was an “extracavernous/ on DDR/anterior wall” aneurysm. These typical cases showed the superiority of 3D-CTA fusion images over MRI, which was confirmed by intraoperative findings.
Relationship of aneurysms to the cavernous sinus and the optic strut on CT
Comparison of 3D-CTA fusion image and DSA based on the intraoperative findings
CT studies
The 3D-CTA fusion images showed 11 “extracavernous” and five “intracavernous” aneurysms. The relationship between aneurysm location and the cavernous sinus was readily observed (Figs. 1d, 2b, and 3b). Using the optic strut as an anatomic landmark, the non-subtracted axial source images on 3D-CTA showed that 11 aneurysms were “distal to PDR” and five were “proximal to PDR”. Accordingly, the location of the aneurysm using these two CT methods was in complete agreement. MRI studies On MRI, the maximum diameter of the aneurysms ranged from 2.0 to 25 mm (mean, 9.8 mm). Visualization of the aneurysm dome was incomplete in three cases (18.8 %) (cases 1a, 6 and 11, the sizes of which were 19, 21, and 25 mm, respectively). MRI, using the origin of ophthalmic artery or DDR as an anatomic landmark, showed nine “distal to DDR”
The DSA and 3D-CTA fusion image findings were compared in nine patients who underwent DSA. Regarding the location of the aneurysms, no discordance with the intraoperative findings was observed between the two modalities. The 3DDSA provided an excellent quality view of the arteries including dents, suggesting the attachment of DDR and blebs, which corresponded well to the intraoperative findings (Figs. 2c, e, f and 3c, d). The 3D-CTA fusion images provided a quality of arteries view that was almost equal to that of the DSA, but also provided high quality images of the osseous and venous structures (Figs. 2b, e, f and 3b, d). Accordingly, the 3DCTA fusion images provided comprehensive information on the arterial, venous, and osseous structures for surgical procedures. Regarding the aneurysmal neck remnant after surgical clipping, postoperative 3D-CTA images were rated as definitely absent in six of ten aneurysms (Figs. 2g and 3e, f), as possibly present in two, and as definitely present in two. Postoperative
510
Acta Neurochir (2014) 156:505–514
Fig. 1 3D-CTA images of case 5 (left ICA anterior wall aneurysm). Left anterior oblique view. a 3D-CTA. A dent in the aneurysm dome suggests attachment of the DDR. A small intracavernous aneurysm (white arrow) is also visualized. b 3D-CTV. c 3D-bone image. d 3D-CTA fusion image with arteries, veins, and bone (semi-transparent). The anatomic relationship among the paraclinoid aneurysm, ACP, and CS is clearly depicted. e time-of-flight MRI demonstrates the origin of the ophthalmic artery (white arrow). The aneurysm appears to be located above the ophthalmic
artery, which means “distal to DDR”. f intraoperative photographs after removing the ACP indicate that the aneurysm neck straddles the DDR. The dent of the aneurysm dome on 3D-CTA corresponds to the DDR in the surgical view. 3D-CTA three-dimensional computed tomographic angiography, 3D-CTV three-dimensional computed tomographic venography, ACP anterior clinoid process, An aneurysm, CS cavernous sinus, DDR distal dural ring, OA ophthalmic artery, ON optic nerve
DSA confirmed the neck remnant in the two patients with a “definitely present” small neck on 3D-CTA fusion imaging.
conventional imaging techniques, such as DSA, conventional MDCT, and MRI/MRA.
Radiation dose
Validation of CTA and MRI findings with intraoperative findings
All dynamic 4D-CTA examinations were completed successfully without any adverse effects. Regarding the radiation dose of each examination, the volumetric CT dose index (CTDIvol) ranged from 114 to 444.6 mGy (mean 352.3 mGy), and the dose-length product (DLP) ranged from 1056.0 to 3574 mGy cm (mean 2253.9 mGy cm).
Discussion The current study demonstrated that 320-detector row CTA achieved high accuracy for the diagnosis of paraclinoid and intracavernous aneurysms. The precise anatomic relationship between the aneurysms and the surrounding structures were recognized on 3D-CTA fusion images. It seems very difficult, if not impossible, to assess these relationships using
In the current study, 3D-CTA fusion images allowed easy identification and visualization of proximal ICA aneurysms, even if they were adjacent to or embedded in the surrounding structures. Time-of-flight MRA may incompletely visualize aneurysms (especially large or giant aneurysms) because of its slow or disturbed flow [25, 28]. Other techniques, such as phase-contrast and contrast-enhanced MRA, can supplement time-of-flight MRA to help improve visualization [25, 29]. However, no aneurysm was incompletely visualized on the 3D-CTA fusion images. Identification of the accurate location of the paraclinoid and extradural intracavernous aneurysms is critical when considering treatment options. In the current study, no discordance in aneurysm location was observed between the preoperative 320-detector row CT findings and intraoperative findings. In contrast, discordance between the preoperative MRI
Acta Neurochir (2014) 156:505–514
511
Fig. 2 Case 10 (left ICA anterior wall aneurysm). a preoperative 3DCTA. b preoperative 3D-CTA fusion image simulating the surgical view, which demonstrates the ACP (white arrow) obscuring the aneurysm neck and the optic strut along the border of the cavernous sinus. The aneurysm neck was distal to the cavernous sinus and the optic strut (“extracavernous” aneurysms). Bleb of the aneurysm dome was also clearly visualized (white arrowhead). c preoperative 3D-DSA demonstrates OA, aneurysm, and bleb, which are in excellent agreement with 3D-CTA. d MR coronal T2-weighted balanced fast field-echo image showing the aneurysm arising from the ICA distal to the DDR (“distal to DDR” aneurysm). e, f intraoperative photographs after removing the ACP (white arrows) (e) and after neck clipping (f), which correspond with
pre- and postoperative 3D-CT images (b and g). According to the intraoperative findings, the aneurysm neck straddled the DDR, and the clip needed to be applied to both the neck of the aneurysm and the dura of the DDR. Black arrowheads demonstrate the bleb, which is identified both on 3D-CTA and 3D-DSA. These findings are in excellent agreement with 3D-CTA. g postoperative 3D-CTA fusion image showing applied titanium-alloy clips with few artifacts. Presence of the remnant neck was rated as definitely absent. 3D-CTA three-dimensional computed tomographic angiography, ACP anterior clinoid process, An aneurysm, DDR distal dural ring, ICA internal carotid artery, OA ophthalmic artery, ON optic nerve
findings and intraoperative findings were found in five of 16 (31.3 %) aneurysms, which was significantly more frequent than with 3D-CTA. Discordance of aneurysm location tended to be more frequent for anterior wall aneurysms and carotid cave aneurysms. Difficulty in localizing the aneurysm on MRI may also be due to anatomic variations of the ophthalmic artery and difficulty visualizing the DDR on 2D images.
Some investigators attempted to identify the PDR and reported on the usefulness of the optic strut on CT angiography [5, 6]. The optic strut is located along the anterior border of the PDR, which defines the roof of the cavernous sinus. Aneurysms above the level of the optic strut are considered intradural, and those below this level are considered extradural and intracavernous. However, the optic strut is an indirect
512
Acta Neurochir (2014) 156:505–514
Fig. 3 Case 9 (left ICA ventral paraclinoid aneurysm). a preoperative 3D-CTA, showing aneurysm and OA. b 3D-CTA fusion image simulating the surgical view, showing the aneurysm, bleb of the aneurysm dome (white arrowhead), ACP (white arrow), and the optic strut. c preoperative 3D-DSA demonstrates OA, aneurysm, and bleb (white arrowhead), which are in excellent agreement with 3D-CTA. d intraoperative photograph after removing the ACP (arrows), which is in excellent agreement
with the preoperative 3D-CTA fusion image. e postoperative 3D-CTA showing two titanium-alloy clips, with few artifacts. Presence of the remnant neck was rated as definitely absent. e postoperative DSA confirmed no residual aneurysm. 3D-CTA three-dimensional computed tomographic angiography, ACP anterior clinoid process, An aneurysm, ICA internal carotid artery, OA ophthalmic artery; ON optic nerve
landmark for the PDR, and the precise site at which the ICA penetrates the dura mater is not clearly identified. Most of the previously reported approaches used only 2D images. Tsuboi et al. attempted to visualize the aneurysms and the cavernous sinus on 2D multi-planar reconstruction images with contrastenhanced MRA [27]. In our current approach with 3D-CTA fusion imaging, we evaluated the relationship among the aneurysms, the cavernous sinus, and the optic strut directly and simultaneously by visualizing them in a 3D manner. Consequently, the images were easy to interpret. The cavernous sinus and the optic strut were visualized directly and in a 3D fashion on the 3D-CTA fusion images. On the other hand, the DDR could not be visualized directly with this method, but the concavity in the paraclinoid segment of the ICA on 3D-CTA coincided with the level of attachment of the DDR [17]. Furthermore, a dent in the aneurysm dome on the 3D-CTA fusion images suggested the location of the DDR in some cases. By using several MRI techniques, some investigators attempted to identify the subarachnoid space, the DDR, and the cavernous sinus [23, 24, 28]. Although direct visualization of the DDR facilitates identification of “supraclinoid intradural” aneurysms, it is less useful for determining the precise location of the carotid cave (intradural)
aneurysms [8, 24]. Furthermore, the current study demonstrated that its correlation with surgical findings was limited (Fig. 2).
Comparison of 3D-CTA fusion imaging and DSA based on the intraoperative findings The 3D-CTA fusion images provided comprehensive information on arterial, venous, and osseous structures. The images provide an almost complete conceptual depiction of the intraoperative view. Furthermore, preoperative identification of the venous flow pattern of the cavernous sinus on 3D-CTA fusion images was helpful for maintaining hemostasis during surgery. The information obtained on 3D-CTA fusion images seems to be superior to DSA, which facilitates the visualization of the arterial structures only. The diameter of perforators is often small, and the spatial resolution of 3D-CTA images is not sufficiently high for their depiction, which is a disadvantage of 320-detector row CTA compared with DSA. Fortunately, perforators do not often become an important problem during surgery for paraclinoid and intracavernous aneurysms.
Acta Neurochir (2014) 156:505–514
513
On postoperative 3D-CTA images, the titanium-alloy clips did not appear to cause significant scatter artifact. Despite the clips, the parent arteries could be reliably evaluated in all ten aneurysm cases. Findings of the postoperative DSA were in good agreement with 3D-CTA.
intraoperative view. Consequently, 320-detector row CT might be useful for the diagnosis of the paraclinoid and intracavernous aneurysms and for developing a surgical treatment plan.
Limitations
Conflicts of interest Drs. Y.O. and K.S. have received research grants from Toshiba Medical Systems, although all authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article. This study was supported in part by a Grant-inAid for Scientific Research (KAKENHI) (90403261), a research grant from the Hyogo Science and Technology Association of Japan and a research grant from the Japan Vascular Disease Research Foundation.
The current study has some limitations. First, the small number of patients made it almost impossible to perform further statistical analysis. Second, all patients did not undergo DSA, which is the gold standard for aneurysm diagnosis, although 320-detector row CT and DSA were in excellent agreement with the intraoperative findings in nine patients who underwent both modalities. Further study with the use of DSA would be beneficial to address the issues of sensitivity and specificity for aneurysm detection and localization, and the value of postclipping CTA. Third, patients with intracavernous aneurysms did not undergo surgical clipping, and thus, the detailed location of these aneurysms in the cavernous sinus was not confirmed with intraoperative inspection. At least, however, no intradural localization of those aneurysms was verified during surgery in the current study. Fourth, dynamic 4D-CTA increases the radiation dose [4, 20]. This is a major limitation of time-resolved CTA imaging. The DLP of the current study was 2253.9 mGy cm. The DLP of 4D-CTA with 320-detector row CT was reported to be 2355.4 mGy in one study [22]. These values are in line with the radiation exposure of conventional CT perfusion with 64detector row CT (DLP: 2330 mGy cm) [3]. However, 3DCTA fusion imaging with 320-detector row CT is likely to deliver more information than conventional MDCT, which may justify the increased radiation exposure. Future investigations should focus on minimizing the radiation dose while maintaining image quality. Another limitation of the current approach with dynamic 4D-CTA is the enormous amount of data. A massive amount of data, up to 3,200 axial images (160 slices, 20 scans), is produced at the time of a single examination, which leads to data storage issues.
Conclusions The results of the current study demonstrated that the findings of 320-detector row CT were in excellent agreement with intraoperative findings. This modality provided a high accuracy of diagnosis for paraclinoid and intracavernous aneurysms. The technique also facilitated comprehensive depiction of the aneurysms and surrounding structures, and therefore provided an almost complete conceptual drawing of the
References 1. al-Rodhan NR, Piepgras DG, Sundt TM Jr (1993) Transitional cavernous aneurysms of the internal carotid artery. Neurosurgery 33: 993–998 2. Buerke B, Puesken M, Wittkamp Stehling C, Stehling C, Ditt H, Seidensticker P, Wessling J, Heindel W, Kloska SP (2010) Bone subtraction CTA for transcranial arteries: intra-individual comparison with standard CTA without bone subtraction and TOF-MRA. Clin Radiol 65:440–446 3. Cohnen M, Wittsack HJ, Assadi S, Muskalla K, Ringelstein A, Poll LW, Saleh A, Mödder U (2006) Radiation exposure of patients in comprehensive computed tomography of the head in acute stroke. AJNR Am J Neuroradiol 27:1741–1745 4. Diekmann S, Siebert E, Juran R, Roll M, Deeg W, Bauknecht HC, Diekmann F, Klingebiel R, Bohner G (2010) Dose exposure of patients undergoing comprehensive stroke imaging by multidetector-row CT: comparison of 320-detector row and 64detector row CT scanners. AJNR Am J Neuroradiol 31:1003–1009 5. Gonzalez LF, Walker MT, Zabramski JM, Partovi S, Wallace RC, Spetzler RF (2003) Distinction between paraclinoid and cavernous sinus aneurysms with computed tomographic angiography. Neurosurgery 52:1131–1137, discussion 1138–1139 6. Hashimoto K, Nozaki K, Hashimoto N (2006) Optic strut as a radiographic landmark in evaluating neck location of a paraclinoid aneurysm. Neurosurgery 59:880–895, discussion 896–897 7. Hiratsuka Y, Miki H, Kiriyama I, Kikuchi K, Takahashi S, Matsubara I, Sadamoto K, Mochizuki T (2008) Diagnosis of unruptured intracranial aneurysms: 3T MR angiography versus 64-channel multidetector row CT angiography. Magn Reson Med Sci 7:169–178 8. Hitotsumatsu T, Natori Y, Matsushima T, Fukui M, Tateishi J (1997) Micro-anatomical study of the carotid cave. Acta Neurochir (Wien) 139:869–874 9. Iihara K, Murao K, Sakai N, Shindo A, Sakai H, Higashi T, Kogure S, Takahashi JC, Hayashi K, Ishibashi T, Nagata I (2003) Unruptured paraclinoid aneurysms: a management strategy. J Neurosurg 99:241– 247 10. Kinouchi H, Mizoi K, Nagamine Y, Yanagida N, Mikawa S, Suzuki A, Sasajima T, Yoshimoto T (2002) Anterior paraclinoid aneurysms. J Neurosurg 96:1000–1005 11. Leffers AM, Wagner A (2000) Neurologic complications of cerebral angiography. A retrospective study of complication rate and patient risk factors. Acta Radiol 204–210 12. Lell MM, Ditt H, Panknin C, Sayre JW, Ruehm SG, Klotz E, Tomandl BF, Villablanca JP (2007) Bone-subtraction CT angiography: evaluation of two different fully automated image-registration procedures for interscan motion compensation. AJNR Am J Neuroradiol 28:1362–1368
514 13. Lell MM, Ruehm SG, Kramer M, Panknin C, Habibi R, Klotz E, Villablanca JP (2009) Cranial computed tomography angiography with automated bone subtraction: a feasibility study. Invest Radiol 44:38–43 14. Matsumoto M, Kodama N, Sakuma J, Sato S, Oinuma M, Konno Y, Suzuki K, Sasaki T, Suzuki K, Katakura T, Shishido F (2005) 3D-CT arteriography and 3D-CT venography: the separate demonstration of arterial-phase and venous-phase on 3D-CT angiography in a single procedure. AJNR Am J Neuroradiol 26:635–641 15. McKinney AM, Palmer CS, Truwit CL, Karagulle A, Teksam M (2008) Detection of aneurysms by 64-section multidetector CT angiography in patients acutely suspected of having an intracranial aneurysm and comparison with digital subtraction and 3D rotational angiography. AJNR Am J Neuroradiol 29:594–602 16. Murayama K, Katada K, Nakane M, Toyama H, Anno H, Hayakawa M, Ruiz DSM, Murphy K (2009) Whole-brain perfusion CT performed with a prototype 256-detector row CT system: initial experience. Radiology 250:202–211 17. Murayama Y, Sakurama K, Satoh K, Nagahiro S (2001) Identification of the carotid artery dural ring by using threedimensional computerized tomography angiography. Technical note. J Neurosurg 95:533–536 18. Romijn M, Gratama van Andel HA, van Walderveen MA, Sprengers ME, van Rijn JC, van Rooij WJ, Venema HW, Grimbergen CA, den Heeten GJ, Majoie CB (2008) Diagnostic accuracy of CT angiography with matched mask bone elimination for detection of intracranial aneurysms: comparison with digital subtraction angiography and 3D rotational angiography. AJNR Am J Neuroradiol 29:134–139 19. Sakamoto S, Kiura Y, Shibukawa M, Ohba S, Arita K, Kurisu K (2006) Subtracted 3D CT angiography for evaluation of internal carotid artery aneurysms: comparison with conventional digital subtraction angiography. AJNR Am J Neuroradiol 27:1332–1337 20. Shankar JJ, Lum C, Sharma M (2010) Whole-brain perfusion imaging with 320-MDCT scanner: reducing radiation dose by increasing sampling interval. AJR Am J Roentgenol 195:1183–1186 21. Siebert E, Bohner G, Dewey M, Bauknecht C, Klingebiel R (2010) Dose related, comparative evaluation of a novel bone-subtraction
Acta Neurochir (2014) 156:505–514
22.
23.
24.
25.
26.
27.
28.
29.
algorithm in 64-row cervico-cranial CT angiography. Eur J Radiol 73:168–174 Siebert E, Bohner G, Dewey M, Hoffmann KT, Mews J, Engelken F, Bauknecht HC, Diekmann S, Klingebiel R (2009) 320-slice CT neuroimaging: initial clinical experience and image quality evaluation. Br J Radiol 82:561–570 Thines L, Gauvrit JY, Leclerc X, Le Gars D, Delmaire C, Pruvo JP, Lejeune JP (2008) Usefulness of MR imaging for the assessment of nonophthalmic paraclinoid aneurysms. AJNR Am J Neuroradiol 29: 125–129 Thines L, Lee SK, Dehdashti AR, Agid R, Willinsky RA, Wallace CM, Terbrugge KG (2009) Direct imaging of the distal dural ring and paraclinoid internal carotid artery aneurysms with high-resolution T2 turbo-spin echo technique at 3-T magnetic resonance imaging. Neurosurgery 64:1059–1064 Thomas B, Sunaert S, Thamburaj K, Wilms G (2005) Spurious absence of signal on 3D time-of-flight MR angiograms on 1 and 3 tesla magnets in cerebral arteries associated with a giant ophthalmic segment aneurysm: the need for alternative techniques. JBR-BTR 88: 241–244 Tomandl BF, Hammen T, Klotz E, Ditt H, Stemper B, Lell M (2006) Bone-subtraction CT angiography for the evaluation of intracranial aneurysms. AJNR Am J Neuroradiol 27:55–59 Tsuboi T, Tokunaga K, Shingo T, Itoh T, Mandai S, Kinugasa K, Date I (2007) Differentiation between intradural and extradural locations of juxta-dural ring aneurysms by using contrast-enhanced 3dimensional time-of-flight magnetic resonance angiography. Surg Neurol 67:381–387 Watanabe Y, Nakazawa T, Yamada N, Higashi M, Hishikawa T, Miyamoto S, Naito H (2009) Identification of the distal dural ring with use of fusion images with 3D-MR cisternography and MR angiography: application to paraclinoid aneurysms. AJNR Am J Neuroradiol 30:845–850 Wikström J, Ronne-Engström E, Gal G, Enblad P, Tovi M (2008) Three-dimensional time-of-flight (3D TOF) magnetic resonance angiography (MRA) and contrast-enhanced MRA of intracranial aneurysms treated with platinum coils. Acta Radiol 49:190–196
CLINICAL ARTICLE - VASCULAR
Utility of 320-detector row CT for diagnosis and therapeutic strategy for paraclinoid and intracavernous aneurysms Satoshi Inoue & Kohkichi Hosoda & Atsushi Fujita & Yoshiharu Ohno & Masahiko Fujii & Kazuro Sugimura & Eiji Kohmura
Received: 27 October 2013 / Accepted: 9 January 2014 / Published online: 22 January 2014 # Springer-Verlag Wien 2014
Abstract Background The aim of this study was (1) to assess the diagnostic accuracy of 320-detector row computed tomography (CT) for paraclinoid and intracavernous aneurysms, and (2) to investigate whether this method provides sufficient information for surgery. Methods A total of 14 patients with 16 unruptured proximal ICA aneurysms underwent three-dimensional CT angiography (3D-CTA) fusion imaging, which was created by superimposing 3D-CT venography data and/or 3D-bone data onto 3D-CTA data using 320-detector row CT, magnetic resonance imaging (MRI), and 3D digital subtraction angiography (DSA). The images of each modality were assessed using intraoperative findings as the reference standard. Results All aneurysms were clearly visualized on 320detector row CT. Bone subtraction and arterio-venous discrimination were accurate. On 3D-CTA fusion images, 11 aneurysms were diagnosed as “extracavernous” and five as “intracavernous”. No discordance in aneurysm location between the 3D-CTA fusion images and the intraoperative findings was found. In contrast, discordance between MRI and intraoperative findings were found in five of the 16 cases (31 %), which was significantly more frequent than with 3D-CTA (p=0.043). The findings DSA, which was performed in nine patients, were also in excellent agreement with the intraoperative findings. However, 3D-CTA fusion imaging provided more comprehensive information, including venous and osseous structures, than 3D-DSA. The 320-detector row S. Inoue : K. Hosoda (*) : A. Fujita : E. Kohmura Department of Neurosurgery, Kobe University Graduate School of Medicine, 7-5-1 Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan e-mail: [email protected] Y. Ohno : M. Fujii : K. Sugimura Department of Radiology, Kobe University Graduate School of Medicine, Kobe, Japan
CTA after surgery demonstrated a clear relationship between the clip and aneurysmal neck with notably few artifacts, which suggested the utility of this modality for postoperative assessment. Conclusions The 320-detector row CT provided high accuracy for the diagnosis of paraclinoid and intracavernous aneurysms. This technique also provided comprehensive depiction of the aneurysms and surrounding structures. Therefore, this modality might be useful for the diagnosis of the paraclinoid and intracavernous aneurysms and for developing a surgical treatment plan. Keywords Area-detector computed tomography . Dural ring . Intracavernous aneurysm . Paraclinoid aneurysm . Three-dimensional CT angiography . Surgery
Introduction Recent advances in imaging techniques, including multidetector row computed tomography (MDCT), high-field magnetic resonance imaging (MRI), and three-dimensional (3D) rotational catheter angiography, have enabled more accurate diagnoses of cerebral aneurysms. However, some critical issues remain in the clinical evaluation of proximal internal carotid artery (ICA) aneurysms, such as paraclinoid aneurysms and intracavernous aneurysms. It is still difficult to visualize the relationship of an aneurysm to the adjacent dural, osseous, and venous structures, especially the distal dural ring (DDR), proximal dural ring (PDR), anterior clinoid process, optic strut, and cavernous sinus. Obscure visualization causes difficulty in differentiating extracavernous aneurysms from intracavernous aneurysms. This differentiation is critical for the management of proximal ICA aneurysms. Extracavernous aneurysms are thought to
506
carry a risk of subarachnoid hemorrhage and are usually treated surgically or endovascularly [9]. The relationship of an aneurysm to adjacent structures is also critical when considering neurosurgical management. One of the important issues in a clipping procedure is whether proximal control can be obtained without removing the anterior clinoid process, opening the cavernous sinus, and exposing the cervical ICA. It is still difficult to obtain the above-mentioned comprehensive information on proximal ICA aneurysms in a single examination with a conventional imaging technique. Although digital subtraction angiography (DSA) has been the gold standard for imaging cerebral aneurysms, it is an invasive technique [11, 15]. Three-Tesla (3 T) magnetic resonance angiography (MRA) provides high spatial resolution [7] and can depict dural structures, including the DDR [23, 24]. However, MRA does not provide osseous and venous information, and is sensitive to laminar flow-related artifacts, especially in large or giant aneurysms. Three-dimensional computed tomographic angiography (3D-CTA) has been used as an alternative to DSA for the preoperative examination of proximal ICA aneurysms [19]. In contrast with MRA and DSA, 3D-CTA provides both osseous and venous information, which is important when assessing the exact location of aneurysms. However, aneurysms are not always clearly demonstrated on 3D-CTA because they tend to be embedded in the osseous and venous structures [15]. To solve this problem, several techniques to subtract bone [2, 12, 13, 18, 19, 21, 26] and to separate arteries from surrounding venous structures have been reported [14], but the separation technique is still uncommon. Recently, a new generation of CT systems with 320detector row has become clinically available, which makes volumetric imaging possible [16]. Several scans of volumetric CT produce the time-resolved volume data for “dynamic fourdimensional CT angiography (dynamic 4D-CTA)” during the time course of injected contrast medium that passes the brain’s vasculature. Therefore, a single examination of dynamic 4DCTA consists of a series of continuous 3D images of the vasculature, which includes a non-contrast phase, arterial phase, early venous phase, and late venous phase. After the subtraction process, 3D images of the arterial, venous, and osseous structures can be acquired as a “3D-CTA image”, a “3D computed tomographic venography (3D-CTV) image”, and a “3D-bone image”, respectively. The obtained 3D images can be mutually superimposed on 3D workstations. We call this fusion image the “3D-CTA fusion image”. The objective of the current study was to assess the diagnostic accuracy of 320-detector row CT for paraclinoid aneurysms and intracavernous aneurysms using the intraoperative findings as the reference standard. We also investigated whether this method provides sufficient information for surgery.
Acta Neurochir (2014) 156:505–514
Methods Patients with proximal ICA aneurysms were prospectively recruited between July 2008 and January 2013 in our hospital. The inclusion criteria of the current study were unruptured paraclinoid and intracavernous aneurysms and surgical treatment for the aneurysms after a CTA. Exclusion criteria included any contraindications for contrast medium, ruptured aneurysms, coil embolization after CTA, and prior treatment for the aneurysms. The data were retrospectively analyzed. The institutional review board approved this study, and written informed consent was obtained from all patients. CT studies Dynamic CT examinations were performed with a 320detector row CT scanner (Aquilion ONE, Toshiba Medical Systems Corporation, Tokyo, Japan) using volumetric cine scanning without helical imaging. The technical parameters were as follows: 160×0.5 mm detector width; 0.25 mm reconstruction interval; 512×512 matrix; 180–240 mm field of view; tube voltage, 80 to 100 kV; tube current, 100 to 200 mAs; detector width, 80 mm; and total 20 scans. The scanning delay was automatically adjusted using a test injection method. Eighteen subsequent dynamic scans were acquired with continuous scanning at one scan rotation per 1 s, followed by two intermediate dynamic scans at one scan rotation per 5 s in the late venous phase. Twenty-five milliliters of non-ionic contrast material (370 mgI/ml) were injected into a right cubital vein at a rate of 5 ml/s, followed by 25 ml of saline flush. Post-processing The bone subtraction process was performed automatically on the CT console. Non-enhanced acquisition in the sequence of the 22 scans was used both as a bone image and as a mask image for bone subtraction. The obtained data were transferred to a 3D workstation (Zio M900; Ziosoft, Tokyo, Japan). Then, dynamic 4D-CTA images were generated using a volume rendering technique. From a series of 22 scans of dynamic 4D-CTA, we selected an arterial phase image that provided the best depiction of the aneurysm (3D-CTA) and a venous phase image that provided the best depiction of the cavernous sinus and other surrounding venous structures (3DCTV). Non-enhanced phase images were used to obtain 3Dbone images. The 3D-CTA fusion images were created by superimposing 3D-CTV data and/or 3D-bone data onto 3DCTA data. To create accurate 3D-CTV images, we selected and fused data from two different venous phase images from dynamic 4D-CTA if necessary. Placement of the regions of interest on the ICA or the cavernous sinus was not needed to obtain 3D-CTA and 3D-CTV images because we selected the
Acta Neurochir (2014) 156:505–514
most appropriate arterial and venous phases from the early arterial phase and the late venous phase in a series of dynamic 4D-CTAs. The 3D-CTA and 3D-CTA fusion image We identified each aneurysm and measured its size on 3DCTA. Based on the projection and location, each aneurysm was classified into one of six groups according to previous reports on paraclinoid aneurysms: anterior wall, ventral paraclinoid, true ophthalmic, carotid cave, transitional, or intracavernous [1, 9, 10]. Then, on the basis of the visualized relationship between the aneurysms and the cavernous sinus on 3D-CTA fusion images, the location of each aneurysm was classified as follows: “extracavernous”, in which the aneurysm neck and sac were located distal to and outside of the cavernous sinus; “transcavernous”, in which the aneurysm neck or sac was located partly in the cavernous sinus; or “intracavernous”, in which the aneurysm neck and sac were located inside the cavernous sinus. In addition, the optic strut was used as an anatomic landmark for the PDR [5, 6]. According to the relationship between the aneurysms and the optic struts on non-subtracted axial source images of 3D-CTA, the aneurysms were classified as follows: “distal to PDR” if the aneurysm neck and sac were located distal to the optic strut; “on PDR” if the aneurysm neck or sac straddled the optic strut; or “proximal to PDR” if the aneurysm neck and sac were located proximal to the optic strut. Thus, “distal to PDR” corresponds to “extracavernous”, “on PDR” to “transcavernous”, and “proximal to PDR” to “intracavernous” on 3D-CTA fusion images, respectively. MRI Studies MRI was performed using a 3T MR scanner (Achieva 3.0 T system with Quasar Dual gradients, Philips Healthcare, Best, the Netherlands). Three-dimensional time-of-flight MRA was performed with the following parameters: TR/TE/FA = 25 ms/ 3.45 ms/20º, FOV = 240 mm (RFOV = 90 %), slice thickness = 1.2 (−0.6) mm,matrix = 255 (P) × 400 (F) (reconstruction, 800×800), resolution = 0.94 mm (P) × 0.6 mm (F) (reconstruction, 0.3 mm×0.3 mm), number of slices = 150, SENSE factor = 2.2, NEX = 1, and scan time = 5 min. To locate the aneurysm, the origin of the ophthalmic artery was used as the anatomic landmark of the DDR according to the DSA classification [23]. A horizontal line passing through the origin of the ipsilateral ophthalmic artery was chosen. The aneurysms were classified as follows: “distal to DDR” if the aneurysm neck and sac were above the ophthalmic line; “on DDR” if the aneurysm neck or sac was straddling the ophthalmic line; or “proximal to DDR” if the aneurysm neck and sac were
507
beneath the ophthalmic line. If the ipsilateral ophthalmic artery was hypoplastic or not visible, we used the contralateral ophthalmic artery as a landmark. MR coronal T2-weighted balanced fast field-echo images (coronal T2-weighted bFFE images) were obtained with the above-referenced 3T MR scanner with the following parameters: TR/TE/FA = 6.2 ms/2.6 ms/45º, FOV = 160 mm (RFOV = 100 %), slice thickness = 1.0 (−0.5) mm, Matrix = 326 (P) × 256 (F) (reconstruction, 336×336), resolution = 0.49 mm (P) × 0.63 mm (F) (reconstruction, 0.48 mm×0.48 mm), number of slices = 100, NEX = 3, and scan time = 5 min. On the basis of the relationship between the aneurysm and DDR on MR coronal T2-weighted bFFE images, the aneurysms were classified as follows: “distal to DDR” if the aneurysm neck and sac were located above the DDR; “on DDR” if the aneurysm neck or sac was straddling the DDR; or “proximal to DDR” if the aneurysm neck and sac were located below the DDR [24]. DSA DSA was performed for only nine patients because the other five patients refused DSA. A standard biplane DSA unit with rotational capabilities and a matrix resolution of 1024×1024 (Artis Zee BA twin; Siemens Medical Solutions, Erlangen, Germany) was used. Non-ionic contrast medium (370 mg) was injected with a power injector. Selective angiograms consisted of one anteroposterior, one lateral, and one or two oblique views. The 3D-DSA acquisition was performed for each ICA with aneurysms. Assessment of CT, MRI, and DSA findings Two experienced neurosurgeons (SI and KH) and one neuroradiologist (MF) blinded to the clinical data independently reviewed the CT, MRI, and DSA data. They performed the initial evaluations independently, and disagreements regarding the final interpretations were resolved by consensus. Then, the interpretations were verified using the intraoperative findings as the reference standard. The Fisher's exact test was used for proportion analysis to compare the accuracy of each modality.
Results During the study period, 22 patients with 26 unruptured proximal ICA aneurysms underwent 320-detector row CT scanning. Fourteen patients with 16 aneurysms underwent surgery, which fulfilled the criteria described above. Thirteen patients were female and one was male. The mean age was 60.3 years and ranged from 44 to 72 years. Clinical and imaging data are summarized in Table 1.
72/F
65/F
4
5a
44/F
14
9
9
5
10
16
Ventral
Ventral
Ventral
Intracavernous
Ant wall
Ventral
Intracavernous
True ophth
Intracavernous
Intracavernous
Ant wall
Ventral
Ant wall
Cave
Cave
Intracavernous
Classification
Extracavernous
Extracavernous
Extracavernous
Intracavernous
Extracavernous
Extracavernous
Intracavernous
Extracavernous
Intracavernous
Intracavernous
Extracavernous
Extracavernous
Extracavernous
Extracavernous
Extracavernous
Intracavernous
Location (CS)
Distal to PDR
Distal to PDR
Distal to PDR
Proximal to PDR
Distal to PDR
Distal to PDR
Proximal to PDR
Distal to PDR
Proximal to PDR
Proximal to PDR
Distal to PDR
Distal to PDR
Distal to PDR
Distal to PDR
Distal to PDR
Proximal to PDR
3D-CTA location (OS)
7 15 10 6
− − − −
9
21a
−
−
2
−
9
5
−
−
7
−
25a
10
−
5
2
−
−
4
−
−
19a
Size (mm)
3T-MRI
−
Discordance on CT
Distal to DDR
Distal to DDR
Distal to DDR
Proximal to DDR
Distal to DDR
Distal to DDR
Proximal to DDR
Distal to DDR
Proximal to DDR
Proximal to DDR
Distal to DDR
Distal to DDR
Distal to DDR
Proximal to DDR
On DDR
Proximal to DDR
Location
−
−
−
−
+
−
−
−
−
−
+
−
+
+
+
−
Discordance on MRI
Ventral
Ventral
ND
Intracavernous
Ant wall
Ventral
Intracavernous
True ophth
Intracavernous
ND
ND
ND
ND
Intracavernous
DSA
Clipping
Clipping
Clipping
Bypass + trapping
Clipping
Clipping
Bypass + ligation
Clipping
Bypass + ligation
Exploration
Clipping
Clipping
Clipping
Exploration
Clipping
Bypass + ligation
Surgery
Extracavernous/distal to DDR/ventral
Extracavernous/distal to DDR/ventral
Extracavernous/distal to DDR/ventral
Intracavernous
Extracavernous/on DDR/ant wall
Extracavernous/distal to DDR/ventral
Intracavernous
Extracavernous/distal to DDR/true ophth
Intracavernous
Intracavernous
Extracavernous/on DDR/ant wall
Extracavernous/distal to DDR/ventral
Extracavernous/on DDR/ant wall
Extracavernous/distal to DDR/cave
Extracavernous/distal to DDR/cave
Intracavernous
Operative findings
a
Aneurysms depicted only partially on MRA
3T-MRI 3 Tesla magnetic resonance imaging, 3D-CTA three-dimensional computed tomographic angiography, Ant wall anterior wall, cave carotid cave, CS cavernous sinus, DDR distal dural ring, ND not done, OS optic strut, true ophth true ophthalmic, ventral ventral paraclinoid
70/F
13
24
10
58/F
72/F
9
48/F
70/F
8
11
58/M
7
12
6
50/F
6
7
3
20
73/F
5b
6
6
8
65/F
4
3
45/F
2a
19
2
54/F
1a
Size (mm)
3D-CTA fusion image
2b
Age (y)/Sex
Case
Table 1 Validation for the location of proximal ICA aneurysms CTA, MRI, and DSA with intraoperative findings
508 Acta Neurochir (2014) 156:505–514
Acta Neurochir (2014) 156:505–514
Ten of 11 “extracavernous (distal to PDR)” aneurysms were treated with surgical neck clipping, and one underwent surgical exploration only (case 2b). Four “intracavernous (proximal to PDR)” aneurysms were treated with bypass and ligation or trapping of the ICA because they were symptomatic. One intracavernous aneurysm coexisted with an extracavernous aneurysm. In these cases with intracavernous aneurysms, no intradural portion of the aneurysm was confirmed during surgery. All procedures were performed successfully, and no patients in this series suffered aneurysm rupture during the follow-up period.
509
aneurysms, one “on DDR” aneurysms, and six “proximal to DDR” aneurysms. Validation of CTA and MRI findings with intraoperative findings
Volume-rendered 3D-CTA, 3D-CTV, and 3D-bone images were obtained simultaneously in a single examination using dynamic 4D-CTA (Fig. 1a-c). All proximal ICA aneurysms could be clearly identified on 3D-CTA, even though they were surrounded by osseous or venous structures. The maximum diameter of the aneurysms on 3D-CTA ranged from 2.0 to 24 mm (mean, 9.6 mm). 3D-CTV clearly depicted the complex shape of the venous structures, including the cavernous sinus (Fig. 1b). The 3D-CTA fusion images were obtained by superimposing 3D-CTV images and 3D-bone images onto 3D-CTA images (Figs. 1d, 2b, and 3b). The 3D-CTA fusion images facilitated the recognition of complicated anatomic relationships of an aneurysm to the anterior clinoid process, the optic strut, and cavernous sinus. These images were in excellent agreement with the intraoperative views (Figs. 1, 2, and 3).
No discordance in aneurysm location was observed between the preoperative 320-detector row CT findings and intraoperative findings (Table 1). By contrast, discordance between the preoperative MRI findings and intraoperative findings were found in five of 16 (31.3 %) aneurysms (cases 2a, 2b, 3, 5a, 10 in Table 1), which was significantly more frequent than with 3D-CTA (p=0.043). Discordance of the aneurysm location tended to be more frequent for anterior wall aneurysms and carotid cave aneurysms. For example, case 5a (Fig. 1) and case 10 (Fig. 2) were diagnosed as an “extracavernous” aneurysm on 3D-CTA fusion images. In case 5a, a dent in the aneurysm dome suggesting the attachment of DDR was also visualized on 3D-CTA and confirmed during surgery (Fig. 1a, f). On MRI, they were diagnosed as a “distal to DDR” aneurysms because they appeared to arise distal to the ophthalmic artery (case 5a, Fig. 1e) or DDR (case 10, Fig. 2d). However, the intraoperative findings clearly demonstrated that both of the aneurysm necks straddled the DDR and that the clips needed to be applied to both the neck of the aneurysm and the DDR (Fig. 2e and f). The final diagnosis was an “extracavernous/ on DDR/anterior wall” aneurysm. These typical cases showed the superiority of 3D-CTA fusion images over MRI, which was confirmed by intraoperative findings.
Relationship of aneurysms to the cavernous sinus and the optic strut on CT
Comparison of 3D-CTA fusion image and DSA based on the intraoperative findings
CT studies
The 3D-CTA fusion images showed 11 “extracavernous” and five “intracavernous” aneurysms. The relationship between aneurysm location and the cavernous sinus was readily observed (Figs. 1d, 2b, and 3b). Using the optic strut as an anatomic landmark, the non-subtracted axial source images on 3D-CTA showed that 11 aneurysms were “distal to PDR” and five were “proximal to PDR”. Accordingly, the location of the aneurysm using these two CT methods was in complete agreement. MRI studies On MRI, the maximum diameter of the aneurysms ranged from 2.0 to 25 mm (mean, 9.8 mm). Visualization of the aneurysm dome was incomplete in three cases (18.8 %) (cases 1a, 6 and 11, the sizes of which were 19, 21, and 25 mm, respectively). MRI, using the origin of ophthalmic artery or DDR as an anatomic landmark, showed nine “distal to DDR”
The DSA and 3D-CTA fusion image findings were compared in nine patients who underwent DSA. Regarding the location of the aneurysms, no discordance with the intraoperative findings was observed between the two modalities. The 3DDSA provided an excellent quality view of the arteries including dents, suggesting the attachment of DDR and blebs, which corresponded well to the intraoperative findings (Figs. 2c, e, f and 3c, d). The 3D-CTA fusion images provided a quality of arteries view that was almost equal to that of the DSA, but also provided high quality images of the osseous and venous structures (Figs. 2b, e, f and 3b, d). Accordingly, the 3DCTA fusion images provided comprehensive information on the arterial, venous, and osseous structures for surgical procedures. Regarding the aneurysmal neck remnant after surgical clipping, postoperative 3D-CTA images were rated as definitely absent in six of ten aneurysms (Figs. 2g and 3e, f), as possibly present in two, and as definitely present in two. Postoperative
510
Acta Neurochir (2014) 156:505–514
Fig. 1 3D-CTA images of case 5 (left ICA anterior wall aneurysm). Left anterior oblique view. a 3D-CTA. A dent in the aneurysm dome suggests attachment of the DDR. A small intracavernous aneurysm (white arrow) is also visualized. b 3D-CTV. c 3D-bone image. d 3D-CTA fusion image with arteries, veins, and bone (semi-transparent). The anatomic relationship among the paraclinoid aneurysm, ACP, and CS is clearly depicted. e time-of-flight MRI demonstrates the origin of the ophthalmic artery (white arrow). The aneurysm appears to be located above the ophthalmic
artery, which means “distal to DDR”. f intraoperative photographs after removing the ACP indicate that the aneurysm neck straddles the DDR. The dent of the aneurysm dome on 3D-CTA corresponds to the DDR in the surgical view. 3D-CTA three-dimensional computed tomographic angiography, 3D-CTV three-dimensional computed tomographic venography, ACP anterior clinoid process, An aneurysm, CS cavernous sinus, DDR distal dural ring, OA ophthalmic artery, ON optic nerve
DSA confirmed the neck remnant in the two patients with a “definitely present” small neck on 3D-CTA fusion imaging.
conventional imaging techniques, such as DSA, conventional MDCT, and MRI/MRA.
Radiation dose
Validation of CTA and MRI findings with intraoperative findings
All dynamic 4D-CTA examinations were completed successfully without any adverse effects. Regarding the radiation dose of each examination, the volumetric CT dose index (CTDIvol) ranged from 114 to 444.6 mGy (mean 352.3 mGy), and the dose-length product (DLP) ranged from 1056.0 to 3574 mGy cm (mean 2253.9 mGy cm).
Discussion The current study demonstrated that 320-detector row CTA achieved high accuracy for the diagnosis of paraclinoid and intracavernous aneurysms. The precise anatomic relationship between the aneurysms and the surrounding structures were recognized on 3D-CTA fusion images. It seems very difficult, if not impossible, to assess these relationships using
In the current study, 3D-CTA fusion images allowed easy identification and visualization of proximal ICA aneurysms, even if they were adjacent to or embedded in the surrounding structures. Time-of-flight MRA may incompletely visualize aneurysms (especially large or giant aneurysms) because of its slow or disturbed flow [25, 28]. Other techniques, such as phase-contrast and contrast-enhanced MRA, can supplement time-of-flight MRA to help improve visualization [25, 29]. However, no aneurysm was incompletely visualized on the 3D-CTA fusion images. Identification of the accurate location of the paraclinoid and extradural intracavernous aneurysms is critical when considering treatment options. In the current study, no discordance in aneurysm location was observed between the preoperative 320-detector row CT findings and intraoperative findings. In contrast, discordance between the preoperative MRI
Acta Neurochir (2014) 156:505–514
511
Fig. 2 Case 10 (left ICA anterior wall aneurysm). a preoperative 3DCTA. b preoperative 3D-CTA fusion image simulating the surgical view, which demonstrates the ACP (white arrow) obscuring the aneurysm neck and the optic strut along the border of the cavernous sinus. The aneurysm neck was distal to the cavernous sinus and the optic strut (“extracavernous” aneurysms). Bleb of the aneurysm dome was also clearly visualized (white arrowhead). c preoperative 3D-DSA demonstrates OA, aneurysm, and bleb, which are in excellent agreement with 3D-CTA. d MR coronal T2-weighted balanced fast field-echo image showing the aneurysm arising from the ICA distal to the DDR (“distal to DDR” aneurysm). e, f intraoperative photographs after removing the ACP (white arrows) (e) and after neck clipping (f), which correspond with
pre- and postoperative 3D-CT images (b and g). According to the intraoperative findings, the aneurysm neck straddled the DDR, and the clip needed to be applied to both the neck of the aneurysm and the dura of the DDR. Black arrowheads demonstrate the bleb, which is identified both on 3D-CTA and 3D-DSA. These findings are in excellent agreement with 3D-CTA. g postoperative 3D-CTA fusion image showing applied titanium-alloy clips with few artifacts. Presence of the remnant neck was rated as definitely absent. 3D-CTA three-dimensional computed tomographic angiography, ACP anterior clinoid process, An aneurysm, DDR distal dural ring, ICA internal carotid artery, OA ophthalmic artery, ON optic nerve
findings and intraoperative findings were found in five of 16 (31.3 %) aneurysms, which was significantly more frequent than with 3D-CTA. Discordance of aneurysm location tended to be more frequent for anterior wall aneurysms and carotid cave aneurysms. Difficulty in localizing the aneurysm on MRI may also be due to anatomic variations of the ophthalmic artery and difficulty visualizing the DDR on 2D images.
Some investigators attempted to identify the PDR and reported on the usefulness of the optic strut on CT angiography [5, 6]. The optic strut is located along the anterior border of the PDR, which defines the roof of the cavernous sinus. Aneurysms above the level of the optic strut are considered intradural, and those below this level are considered extradural and intracavernous. However, the optic strut is an indirect
512
Acta Neurochir (2014) 156:505–514
Fig. 3 Case 9 (left ICA ventral paraclinoid aneurysm). a preoperative 3D-CTA, showing aneurysm and OA. b 3D-CTA fusion image simulating the surgical view, showing the aneurysm, bleb of the aneurysm dome (white arrowhead), ACP (white arrow), and the optic strut. c preoperative 3D-DSA demonstrates OA, aneurysm, and bleb (white arrowhead), which are in excellent agreement with 3D-CTA. d intraoperative photograph after removing the ACP (arrows), which is in excellent agreement
with the preoperative 3D-CTA fusion image. e postoperative 3D-CTA showing two titanium-alloy clips, with few artifacts. Presence of the remnant neck was rated as definitely absent. e postoperative DSA confirmed no residual aneurysm. 3D-CTA three-dimensional computed tomographic angiography, ACP anterior clinoid process, An aneurysm, ICA internal carotid artery, OA ophthalmic artery; ON optic nerve
landmark for the PDR, and the precise site at which the ICA penetrates the dura mater is not clearly identified. Most of the previously reported approaches used only 2D images. Tsuboi et al. attempted to visualize the aneurysms and the cavernous sinus on 2D multi-planar reconstruction images with contrastenhanced MRA [27]. In our current approach with 3D-CTA fusion imaging, we evaluated the relationship among the aneurysms, the cavernous sinus, and the optic strut directly and simultaneously by visualizing them in a 3D manner. Consequently, the images were easy to interpret. The cavernous sinus and the optic strut were visualized directly and in a 3D fashion on the 3D-CTA fusion images. On the other hand, the DDR could not be visualized directly with this method, but the concavity in the paraclinoid segment of the ICA on 3D-CTA coincided with the level of attachment of the DDR [17]. Furthermore, a dent in the aneurysm dome on the 3D-CTA fusion images suggested the location of the DDR in some cases. By using several MRI techniques, some investigators attempted to identify the subarachnoid space, the DDR, and the cavernous sinus [23, 24, 28]. Although direct visualization of the DDR facilitates identification of “supraclinoid intradural” aneurysms, it is less useful for determining the precise location of the carotid cave (intradural)
aneurysms [8, 24]. Furthermore, the current study demonstrated that its correlation with surgical findings was limited (Fig. 2).
Comparison of 3D-CTA fusion imaging and DSA based on the intraoperative findings The 3D-CTA fusion images provided comprehensive information on arterial, venous, and osseous structures. The images provide an almost complete conceptual depiction of the intraoperative view. Furthermore, preoperative identification of the venous flow pattern of the cavernous sinus on 3D-CTA fusion images was helpful for maintaining hemostasis during surgery. The information obtained on 3D-CTA fusion images seems to be superior to DSA, which facilitates the visualization of the arterial structures only. The diameter of perforators is often small, and the spatial resolution of 3D-CTA images is not sufficiently high for their depiction, which is a disadvantage of 320-detector row CTA compared with DSA. Fortunately, perforators do not often become an important problem during surgery for paraclinoid and intracavernous aneurysms.
Acta Neurochir (2014) 156:505–514
513
On postoperative 3D-CTA images, the titanium-alloy clips did not appear to cause significant scatter artifact. Despite the clips, the parent arteries could be reliably evaluated in all ten aneurysm cases. Findings of the postoperative DSA were in good agreement with 3D-CTA.
intraoperative view. Consequently, 320-detector row CT might be useful for the diagnosis of the paraclinoid and intracavernous aneurysms and for developing a surgical treatment plan.
Limitations
Conflicts of interest Drs. Y.O. and K.S. have received research grants from Toshiba Medical Systems, although all authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article. This study was supported in part by a Grant-inAid for Scientific Research (KAKENHI) (90403261), a research grant from the Hyogo Science and Technology Association of Japan and a research grant from the Japan Vascular Disease Research Foundation.
The current study has some limitations. First, the small number of patients made it almost impossible to perform further statistical analysis. Second, all patients did not undergo DSA, which is the gold standard for aneurysm diagnosis, although 320-detector row CT and DSA were in excellent agreement with the intraoperative findings in nine patients who underwent both modalities. Further study with the use of DSA would be beneficial to address the issues of sensitivity and specificity for aneurysm detection and localization, and the value of postclipping CTA. Third, patients with intracavernous aneurysms did not undergo surgical clipping, and thus, the detailed location of these aneurysms in the cavernous sinus was not confirmed with intraoperative inspection. At least, however, no intradural localization of those aneurysms was verified during surgery in the current study. Fourth, dynamic 4D-CTA increases the radiation dose [4, 20]. This is a major limitation of time-resolved CTA imaging. The DLP of the current study was 2253.9 mGy cm. The DLP of 4D-CTA with 320-detector row CT was reported to be 2355.4 mGy in one study [22]. These values are in line with the radiation exposure of conventional CT perfusion with 64detector row CT (DLP: 2330 mGy cm) [3]. However, 3DCTA fusion imaging with 320-detector row CT is likely to deliver more information than conventional MDCT, which may justify the increased radiation exposure. Future investigations should focus on minimizing the radiation dose while maintaining image quality. Another limitation of the current approach with dynamic 4D-CTA is the enormous amount of data. A massive amount of data, up to 3,200 axial images (160 slices, 20 scans), is produced at the time of a single examination, which leads to data storage issues.
Conclusions The results of the current study demonstrated that the findings of 320-detector row CT were in excellent agreement with intraoperative findings. This modality provided a high accuracy of diagnosis for paraclinoid and intracavernous aneurysms. The technique also facilitated comprehensive depiction of the aneurysms and surrounding structures, and therefore provided an almost complete conceptual drawing of the
References 1. al-Rodhan NR, Piepgras DG, Sundt TM Jr (1993) Transitional cavernous aneurysms of the internal carotid artery. Neurosurgery 33: 993–998 2. Buerke B, Puesken M, Wittkamp Stehling C, Stehling C, Ditt H, Seidensticker P, Wessling J, Heindel W, Kloska SP (2010) Bone subtraction CTA for transcranial arteries: intra-individual comparison with standard CTA without bone subtraction and TOF-MRA. Clin Radiol 65:440–446 3. Cohnen M, Wittsack HJ, Assadi S, Muskalla K, Ringelstein A, Poll LW, Saleh A, Mödder U (2006) Radiation exposure of patients in comprehensive computed tomography of the head in acute stroke. AJNR Am J Neuroradiol 27:1741–1745 4. Diekmann S, Siebert E, Juran R, Roll M, Deeg W, Bauknecht HC, Diekmann F, Klingebiel R, Bohner G (2010) Dose exposure of patients undergoing comprehensive stroke imaging by multidetector-row CT: comparison of 320-detector row and 64detector row CT scanners. AJNR Am J Neuroradiol 31:1003–1009 5. Gonzalez LF, Walker MT, Zabramski JM, Partovi S, Wallace RC, Spetzler RF (2003) Distinction between paraclinoid and cavernous sinus aneurysms with computed tomographic angiography. Neurosurgery 52:1131–1137, discussion 1138–1139 6. Hashimoto K, Nozaki K, Hashimoto N (2006) Optic strut as a radiographic landmark in evaluating neck location of a paraclinoid aneurysm. Neurosurgery 59:880–895, discussion 896–897 7. Hiratsuka Y, Miki H, Kiriyama I, Kikuchi K, Takahashi S, Matsubara I, Sadamoto K, Mochizuki T (2008) Diagnosis of unruptured intracranial aneurysms: 3T MR angiography versus 64-channel multidetector row CT angiography. Magn Reson Med Sci 7:169–178 8. Hitotsumatsu T, Natori Y, Matsushima T, Fukui M, Tateishi J (1997) Micro-anatomical study of the carotid cave. Acta Neurochir (Wien) 139:869–874 9. Iihara K, Murao K, Sakai N, Shindo A, Sakai H, Higashi T, Kogure S, Takahashi JC, Hayashi K, Ishibashi T, Nagata I (2003) Unruptured paraclinoid aneurysms: a management strategy. J Neurosurg 99:241– 247 10. Kinouchi H, Mizoi K, Nagamine Y, Yanagida N, Mikawa S, Suzuki A, Sasajima T, Yoshimoto T (2002) Anterior paraclinoid aneurysms. J Neurosurg 96:1000–1005 11. Leffers AM, Wagner A (2000) Neurologic complications of cerebral angiography. A retrospective study of complication rate and patient risk factors. Acta Radiol 204–210 12. Lell MM, Ditt H, Panknin C, Sayre JW, Ruehm SG, Klotz E, Tomandl BF, Villablanca JP (2007) Bone-subtraction CT angiography: evaluation of two different fully automated image-registration procedures for interscan motion compensation. AJNR Am J Neuroradiol 28:1362–1368
514 13. Lell MM, Ruehm SG, Kramer M, Panknin C, Habibi R, Klotz E, Villablanca JP (2009) Cranial computed tomography angiography with automated bone subtraction: a feasibility study. Invest Radiol 44:38–43 14. Matsumoto M, Kodama N, Sakuma J, Sato S, Oinuma M, Konno Y, Suzuki K, Sasaki T, Suzuki K, Katakura T, Shishido F (2005) 3D-CT arteriography and 3D-CT venography: the separate demonstration of arterial-phase and venous-phase on 3D-CT angiography in a single procedure. AJNR Am J Neuroradiol 26:635–641 15. McKinney AM, Palmer CS, Truwit CL, Karagulle A, Teksam M (2008) Detection of aneurysms by 64-section multidetector CT angiography in patients acutely suspected of having an intracranial aneurysm and comparison with digital subtraction and 3D rotational angiography. AJNR Am J Neuroradiol 29:594–602 16. Murayama K, Katada K, Nakane M, Toyama H, Anno H, Hayakawa M, Ruiz DSM, Murphy K (2009) Whole-brain perfusion CT performed with a prototype 256-detector row CT system: initial experience. Radiology 250:202–211 17. Murayama Y, Sakurama K, Satoh K, Nagahiro S (2001) Identification of the carotid artery dural ring by using threedimensional computerized tomography angiography. Technical note. J Neurosurg 95:533–536 18. Romijn M, Gratama van Andel HA, van Walderveen MA, Sprengers ME, van Rijn JC, van Rooij WJ, Venema HW, Grimbergen CA, den Heeten GJ, Majoie CB (2008) Diagnostic accuracy of CT angiography with matched mask bone elimination for detection of intracranial aneurysms: comparison with digital subtraction angiography and 3D rotational angiography. AJNR Am J Neuroradiol 29:134–139 19. Sakamoto S, Kiura Y, Shibukawa M, Ohba S, Arita K, Kurisu K (2006) Subtracted 3D CT angiography for evaluation of internal carotid artery aneurysms: comparison with conventional digital subtraction angiography. AJNR Am J Neuroradiol 27:1332–1337 20. Shankar JJ, Lum C, Sharma M (2010) Whole-brain perfusion imaging with 320-MDCT scanner: reducing radiation dose by increasing sampling interval. AJR Am J Roentgenol 195:1183–1186 21. Siebert E, Bohner G, Dewey M, Bauknecht C, Klingebiel R (2010) Dose related, comparative evaluation of a novel bone-subtraction
Acta Neurochir (2014) 156:505–514
22.
23.
24.
25.
26.
27.
28.
29.
algorithm in 64-row cervico-cranial CT angiography. Eur J Radiol 73:168–174 Siebert E, Bohner G, Dewey M, Hoffmann KT, Mews J, Engelken F, Bauknecht HC, Diekmann S, Klingebiel R (2009) 320-slice CT neuroimaging: initial clinical experience and image quality evaluation. Br J Radiol 82:561–570 Thines L, Gauvrit JY, Leclerc X, Le Gars D, Delmaire C, Pruvo JP, Lejeune JP (2008) Usefulness of MR imaging for the assessment of nonophthalmic paraclinoid aneurysms. AJNR Am J Neuroradiol 29: 125–129 Thines L, Lee SK, Dehdashti AR, Agid R, Willinsky RA, Wallace CM, Terbrugge KG (2009) Direct imaging of the distal dural ring and paraclinoid internal carotid artery aneurysms with high-resolution T2 turbo-spin echo technique at 3-T magnetic resonance imaging. Neurosurgery 64:1059–1064 Thomas B, Sunaert S, Thamburaj K, Wilms G (2005) Spurious absence of signal on 3D time-of-flight MR angiograms on 1 and 3 tesla magnets in cerebral arteries associated with a giant ophthalmic segment aneurysm: the need for alternative techniques. JBR-BTR 88: 241–244 Tomandl BF, Hammen T, Klotz E, Ditt H, Stemper B, Lell M (2006) Bone-subtraction CT angiography for the evaluation of intracranial aneurysms. AJNR Am J Neuroradiol 27:55–59 Tsuboi T, Tokunaga K, Shingo T, Itoh T, Mandai S, Kinugasa K, Date I (2007) Differentiation between intradural and extradural locations of juxta-dural ring aneurysms by using contrast-enhanced 3dimensional time-of-flight magnetic resonance angiography. Surg Neurol 67:381–387 Watanabe Y, Nakazawa T, Yamada N, Higashi M, Hishikawa T, Miyamoto S, Naito H (2009) Identification of the distal dural ring with use of fusion images with 3D-MR cisternography and MR angiography: application to paraclinoid aneurysms. AJNR Am J Neuroradiol 30:845–850 Wikström J, Ronne-Engström E, Gal G, Enblad P, Tovi M (2008) Three-dimensional time-of-flight (3D TOF) magnetic resonance angiography (MRA) and contrast-enhanced MRA of intracranial aneurysms treated with platinum coils. Acta Radiol 49:190–196
