Early Clinical Experience with a Dedicated Portable Computed Tomography Scanner in a Biplane Neuroangiography Suite: Assessment of Feasibility and Utility in Neurointerventional Practice

Summary

The ability to have on-site access to cross-sectional imaging in a biplane neuroangiography suite has tremendous potential for enhancing current neurointerventional practice. Although a few prototypical multimodality suites have been created, several problems/limitations have prevented widespread implementation. Recently`, a portable CT scanner has been developed, which may overcome previous shortcomings. We review our recent clinical experience with this new modality, exploring numerous adjunctive diagnostic and therapeutic applications. Forty-one patients underwent periprocedural CT using the Tomoscan M/EG portable CT (Philips). The portable CT scanner is kept at the “head-end” of the biplane neuroangiography suite, being moved into position as needed before, after; or during a procedure. A pivoting angiographic table permits excellent z-plane mobility for rapid gantry to fluoroscopy positioning. Five mm slices at five mm increments were obtained.

High quality images were obtained in all cases. The portable CT scanner could be quickly positioned and activated within five min. Total scanning time for a typical case, including initial positioning and set-up was 10-12 min. Twelve of 41 cases were performed adjunctively during diagnostic angiography; 29/41 were performed in an interventional setting. Twenty of 29 scan evaluated baseline or post-therapeutic status of the brain (e.g., Guglielmi detachable coil aneurysm obliteration, arteriovenous malformation (AVM) embolisation, local thrombolysis); 9/29 provided cross-sectional guidance to various interventions (direct puncture embolisation, percutaneous vertebroplasty, spinal biopsy, discography).

Use of the portable CT scanner permitted rapidly accessible high quality cross sectional imaging within the biplane neuroangiography suite, which augmented diagnostic and therapeutic decision-making, and therapeutic intervention.

Introduction

The scope and depth of interventional neuroradiology practice continues to grow as a result of many rapidly evolving technical and technological innovations that have occurred within the last decade. With this expansion in clinical activities come increasing expectations and demands for improving both technical and clinical outcomes, many of which may derive from enhanced efficiency or optimization of the precision and/or specificity of therapeutic decision-making and execution.

Consequently, there is currently interest in developing specialized image-guided procedure/operating rooms that integrate multiple modalities for approaching a variety of pathologies affecting the brain, spinal cord, skull, neck and spinal column. Since much of conventional neurointerventional practice has been and will continue to be directed at endovascular surgical procedures, such “hybrid” facilities must at the minimum have a fixed site, monoplane (and preferably biplane) C-arm with high resolution fluoroscopy and digital subtraction angiography (DSA) imaging chain. The additional component most commonly needed to be incorporated into these types of hybrid facilities is some type of cross sectional imaging modality (i.e. usually CT or MRI).

In the recent past both scaled down, and full size fixed site CT or MRI units have been built into the same or immediately adjacent room housing a monoplane fixed site angiography room. In such configurations, the two modalities usually share some type of modified table that can be manually or mechanically positioned into either device^2,3,5. Although the preliminary experience with these prototype systems has been promising, many limitations and compromises (e.g. structural incompatibilities, impractical logistics, economic obstacles, etc) have been encountered that have severely limited their widespread implementation and utilization.

However, a potentially more flexible and practical solution to the need for “on demand” and “on site” cross sectional imaging in a neuroangiography suite may be derived from the use of a portable or mobile CT scanner. Such a device has recently become commercially available in which preliminary clinical use in the operating room and intensive care settings have been very favorable ^1,4,6. Our group has recently acquired such a device for unique dedicated application within a modern biplane neuroangiography suite, where a wide spectrum of neurointerventional services is rendered. We review our initial experience with the mobile CT scanner to assess its feasibility and utility for enhancing current neurointerventional practice.

Material and Methods

A dedicated third generation mobile gantry CT system, (Tomoscan M/EG, Philips Medical, Eindhoven, Netherlands) was placed into an existing biplane neuroangiography suite (BN3000, Philips Medical) (figure 1). The mobile CT scanner has been described in detail previously¹. In brief, the device consists of a separate mobile gantry scanner and operator console (Sun SPARC 5 UNIX operating system) that is on wheels for easy transportation. Unlike fixed site systems, the mobile CT uses precise mechanical translation of the gantry over a fixed table position for obtaining contiguous slice acquisition.

Figure 1.

A Tomoscan M portable CT scanner (Philips, Medical) is placed at the head end of a fixed biplane neuroangiography suite (BN3000, Philips Medical), allowing for rapid alternating digital subtraction angiography/fluoroscopy/cross-sectional imaging capability.

When not immediately in use the portable CT scanner was positioned at the “head end” of the biplane neuroangiography suite. Positioning for scanning and fluoroscopic/angiographic procedures was accomplished by a combination of moving the patient's position on the angiographic table (typically by moving the targeted anatomy as close to the front end of the table as possible), the gantry, and angiographic table. In all cases, the lateral C-arm would have to be moved from the operating field, while the base of the frontal plane C-arm was rotated 90 degrees “away” from the longitudinal vertex position. For cases involving the head, neck, cervical or thoracic spine, the patient was placed “head first” upon the angiographic table. In cases involving the lumbar spine or pelvis, the patient was placed in a “feet first” position.

The mobile CT gantry was positioned as close to the head end of the table as the rotated frontal C-arm would allow. Scans of the head and upper cervical spine could be performed in the “A” pivot position with the angiographic table capable of extending to its greatest caudal position, while scans of the lower cervical and thoracic spine required translation of the angiographic table into the “B” pivot position, where the table may be extended to its greatest cephalad position. Scans of the lumbar spine and pelvis generally required the “B” pivot position. The free-floating angiographic table was locked at an appropriate height and lateral position for the anatomic site to be studied. With the mobile gantry within its most caudal position on its console, the table was longitudinally translated into a position where the most caudal aspect of the field of view was aligned with the axial scan plane of the gantry. Patient movement is not needed at any time during the set-up or scanning.

To perform contiguous slice scanning, the anatomic region of interest would first be imaged with either a lateral or frontal planar scanogram. Most head scans were planned from a scanogram in the lateral projection acquired at 120 kV, 30 ma, 17.1 s, and a field of view (FOV) of 250 mm. Using the Sun SPARC 5 UNIX workstation, the axial FOV, slice thickness, slice interval, slice number, image and reconstruction matrix, kv, ma, scan time, and scan angle could be planned. Initially, scans of the head were performed at a 512 x 512 matrix, 120 kev, 45 ma, scan time 4 s, FOV of 201 mm, slice thickness of 5 mm, slice interval 5 mm, and a gantry angle of 9-1 1 degrees (approaching the cantho-meatal line). At these initial settings, only 12 slices of the head were possible secondary to the charge limitations of the gantry's capacitor. This required several minutes of re-planning the remainder of the region of interest, and allowing the capacitor to recharge for the next scan session to obtain complete coverage of the head. Reducing the ma to 30 resulted in complete coverage of the head in a single scan session, without a significant subjective reduction in image quality. Predefined scan parameters for the cervical, thoracic and lumbar spine were generally used, with appropriate changes in the matrix size, slice thickness, slice interval, and gantry angle for the prescribed region of interest. In general, scans of thicker segments of the body could be performed at the same kV and scan time, with an increase in the ma (i.e. 50 ma was used for scans of the lumbar spine).

Results

43 portable CT scans have been performed on 41 patients undergoing a wide range of diagnostic or therapeutic evaluations by the in-terventional neuroradiology team. In all cases, the CT studies obtained with the portable CT scanner were of high quality (figures 2-5). The spatial resolution and contrast of these images subjectively approached those obtained by geographically fixed scanners within our department. All studies obtained by the device exceeded our minimum expectations for diagnostic utility.

Figure 2.

A) Axial CT image demonstrates deposition of embolic material within the nidus of a right frontal AVM. B) Axial CT image, following suspected intraprocedural vessel perforation during superselective catheterization, confirms subarachnoid extravasation of contrast and blood. The patient remained clinically stable and asymptomatic.

The portable CT scanner could be quickly positioned and activated within five minutes or less. The set-up time continued to decrease as the experience of both the angiography technologists (n=4) and physicians (n=3) increased. As expected, there was also a learning curve for both technologists and physicians for planning and executing scans. However, after performing only a few studies, the average total scanning time for a typical head case, including initial positioning, set-up and scanning, decreased to 10-12 minutes. Single slice scans could be performed with 2, 4, or 6 second acquisitions with cycling times of 7, 9, and 11 seconds respectively. Computer reconstruction of 512 x 512 matrix images occurred with a duration of 6 seconds for each slice.

Table 1 summarizes the types of cases encountered in our series. 12/43 cases were performed adjunctively during diagnostic angiography. In many cases, the CT scanner was brought into the scanning position at the completion of angiography, resulting in the performance of a delayed contrast enhanced CT scan. These studies often provided both confirmatory and complementary cross-sectional diagnostic imaging. For example, several patients with clinically suspected atypical high flow vascular malformations were found to lack arteriovenous shunting in association with either completely absent or subtle pathoanatomic findings on angiography. However, the supplemental use of CT immediately following angiography in these cases enabled us to both confirm the localization of intraparenchymal pathology, and often provide a higher degree of confidence in diagnosing angiographically occult ve-nous or cavernous malformations of the brain. The addition of CT to many of the diagnostic angiographic studies was found to be particularly useful in several situations, including but not limited to subjectively assessing the volume of vascular malformations or enhancing neoplasms, identifying subtle microvascular enhancement occasionally seen with certain vascular diseases (e.g. vasculitis), and determining the presence or absence of intraluminal thrombus within intracranial aneurysms.

Table 1.

Current applications for portable CT scans

PROCEDURE	PATIENTS	DIAGNOSIS/ INTERVENTION	PROCEDURE CODE (I-intervention, D-diagnosis, cip-combined interventional procedure)
1	KG	CA VM stage I embolisation	I
2	KG	CA VM stage I embolisation	I
3	MS	CAVM vs venous malformation	D
4	TD	CAVM vs venous malformation	D
5	SH	PTA evaluation/intraarterial lysis	I
6	MR	ICH; R/O CAVM	D
7	GM	SAH; sentinal headache	D
8	CW	Basilar summit aneurysm; GDC embolisation	I
9	TR	Mycotic aneurysm; NBCA embolisation	I
10	DA	facial venous malformation; DPSTx w/ Sotradecol; combination CT/fluoro/DSA	I (cip)
11	GW	SAH; Basilar summit aneurysm/GDC	I
12	JD	severe closed head injury	D
13	RL	R O intracranial vasculitis	D
14	PK	CAVM s/p resection; r/o remnant	D
15	DE	ECA AVM embolisation combined DPAE and intra-arterial NBCA	I (cip)
16	DH	ADST; s/p venous UK and thrombolysis	I
17	GC	failed aneurysm clipping x 2; GDC f/u	I
18	JT	CT/fluoro guided drainage	I (cip)
19	RK	s/p aneurysm clipping; r/o remnant	D
20	BR	ectopic glomus tumour nasopharyngeal; DPAE w/ combine CT and fluoroscopy	I (cip)
21	JS	JAF; considered for combined CT/fluoro	I
22	GB	ICH; r/o AVM	D
23	SH	stroke in evolution; Intra-arterial thrombolysis	I
24	JI	CAVM; Stage I embolisation	I
25	SH	failed aneurysm clipping w/ rebleed; GDC; IA papaverine	I
26	RK	rCBS	I
27	VL	CT/fluoroscopic guided biopsy	I (cip)
28	HB	Rt frontal lobe AVM, embolisation NBCA	I
29	PC	spinal ABC; DPAE	I (cip)
30	SE	GDC; inadvertant parent artery occlusion	I
31	JJ	CT guided discography	I (cip)
32	SP	CT guided discography	I (cip)
33	RS	CAVM; s/p stereotactic radiotherapy	D
34	RV	CT guided discography	I (cip)
35	DC	SCA aneurysm; GDC	I
36	SV	CAVM embolisation; NBCA	I
37	LR	PCoA aneurysm; GDC; complicated	I
		with intraprocedural rebleed
38	CV	cerebellar CAVM; NBCA embo stage I	I
39	SW	SAH; ACoA aneurysm; frontal haematoma	D
40	SL	SAH; GDC complicated w/ rebleed	I
41	MT	sacral lesion; CT/fluoro guided biopsy	I (cip)
42	GL	multiple intracranial aneurysms; haematoma	D
43	BA	LICA aneurysm; GDC; coil perforation with rupture	I
CAVM = cerebral AVM; PTA = percutaneous transluminal angioplasty; ICH = intracranial haemorrhage; GDC = Gugliemi detach- able coil; SAH = subarachnoid haemorrhage; NBCA = n-butyl cyanoacrylate; DPSTx = direct puncture sotradecol therapy; ECA = ex- ternal carotid artery; DPAE = direct puncture acrylic embolisation; ADST = acute dural sinus thrombosis; UK = urokinase; JAF = ju- venile angiofibroma; IA = intra-arterial; CBS = carotid blowout syndrome; SCA = superior cerebellar artery; PCoA = posterior com- municating artery; ACoA = anterior communicating aneurysm; LICA = left internal carotid artery.

In several situations, a non-contrast CT was obtained prior to diagnostic angiography. Typically, this was performed in an acute care setting on relatively unstable or high clinical grade patients to rapidly assess the extent and/or progression of neuropathology and the possible need for acute therapeutic intervention.

The remaining 31/43 cases were performed in a variety of interventional and/or therapeutic settings. In the majority of cases (21/43) CT scans were obtained as adjunctive studies either before or after therapeutic intervention. Specifically, they were used to assess results following aneurysm, AVM, or tumour embolisation, providing additional cross-sectional information essential for assessing the following: correlation of patient's condition with pathology, progress of therapy, or the development of clinically unsuspected complications. For example, in a case of endosaccular embolisation of a previously ruptured berry aneurysm, significant additional subarachnoid haemorrhage (SAH) was identified on the post-therapeutic CT study, indicating clinically unsuspected rehaemorrhage that was likely the result of aneurysm wall perforation during endovascular surgery. In this case, the ventricular size, severity of cerebral oedema and mass effect were rapidly assessed without extensive delays related to availability of a fixed site CT scanner and patient transportation. This resulted in rapid selection of appropriate therapeutic interventions (e.g. ventriculostomy) for managing this complication. Another example demonstrated the development of a moderate amount of focal subarachnoid haemorrhage following perforation of a feeding artery during endovascular embolisation of a right frontal AVM (figure 2). This patient experienced the sudden onset of a severe focal frontal headache during the procedure. DSA confirmed patency of the draining vein and failed to demonstrate active extravasation. Rapid CT evaluation confirmed initial SAH in the absence of significant cerebral oedema, mass effect, or ventricular extension or enlargement. The session was concluded and the patient was closely monitored for SAH-re¯ lated complications. No additional intervention was required and the patient remained otherwise clinically asymptomatic.

In 10/43 cases, the CT scanner provided invaluable cross sectional guidance to various diagnostic and therapeutic interventions. For several cases, a combined CT fluoroscopic-angiographic guided approach to therapeutic intervention was used. Such cases included direct percutaneous puncture of craniocervical and spinal tumours, and head & neck vascular malformations (figure 3). In this setting the CT gantry was maintained at the head end of the angiographic table throughout the duration of the procedure. This allowed for rapid repositioning of the patient for CT or single-plane fluoroscopy/digital subtraction angiography throughout the procedure by simple longitudinal translation of the angiographic table. In these patients (DA, DE, and BR; table 1), CT images were acquired to identify non-embolised compartments of the pathologic entities, direct and confirm placement of needles within these compartments, and intermittently assess the progress of therapy. Fluoroscopy and DSA were also performed throughout the procedures to provide live guidance for needle placement, assessment and confirmation of vascular compartment access, real time monitoring of embolisations, and post-embolisation control angiography. These patients were successfully managed in a single treatment session, obviating the need for additional endovascular therapeutic interventions prior to surgical excision.

Figure 3.

A) RAO digital subtraction angiography of the right submandibular space demonstrates direct percutaneous positioning of the needle within a vascular compartment of a large facial venous malformation. B) RAO radiograph illustrates intravascular compartment deposition of embolic material. C) Axial CT of the region confirms intravascular deposition of embolic material within the anterior aspect of the malformation and identifies additional non-embolised compartments posteriorly.

CT and fluoroscopic guided discography were also performed in several cases, which permitted rapid acquisition of axial cross-sectional imaging of the injected discs (figure 4). Also the on site cross-sectional imaging obtained with the mobile CT scanner helped plan access to the often difficult to reach L5-Sl disc space.

Figure 4.

A) Lateral radiograph of the L5/Sl disc level following discography suggests posterior extension of contrast without discrete extravasation. B) Axial CT image confirms annular thinning and a small vertical tear, posteriorly.

Other diagnostic or therapeutic interventions using the combination of multiplanar fluoroscopy and CT included percutaneous spinal biopsies, abscess drainages and percutaneous vertebroplasty (figure 5). In patient PC (table 1), transpedicular placement of the needle, intratumoral deposition of embolic material, and the absence of intravascular or spinal canal migration of embolic material were confirmed with greater confidence and sensitivity (subjectively) using axial CT imaging.

Figure 5.

A) AP radiograph demonstrates transpedicular placement of a biopsy needle within the body of the T 12 vertebrae. Extensive embolisation of the collapsed vertebrae with NBCA has been achieved. B) Axial CT image confirms placement of the needle and NBCA embolisation of the vertebral body. In addition, no retroperitoneal or intravascular migration of embolic material is observed.

Discussion

The increasing sophistication and capability of interventional neuroradiology practice is creating greater demands for optimizing and integrating the use of various diagnostic and therapeutic imaging modalities. This trend derives from a variety of developments that have occurred within this decade, including the increasing utilization of endovascular surgical techniques for the management of various acute haemorrhagic (e.g. aneurysm associated SAH) and occlusive cerebrovascular diseases (e.g. thromboembolic stroke, SAH induced vasospasm), the increasing emphasis on both anatomic and functional cross-sectional imaging for peri-therapeutic decision-making, and the development of new diagnostic and therapeutic interventions that may benefit from supplemental cross-sectional imaging guidance^1-6.

These demands may be mostly satisfied by creating multimodality image-guided operation rooms that will provide the broadest possible range of diagnostic and therapeutic capabilities within a convenient proximity. Since much of neurointerventional practice continues (at least for the near future) to rely heavily upon angiographic and/or fluoroscopic-guided diagnosis and therapy, the “centrepiece” imaging modality for such a hybrid room will likely remain a fixed site biplane C-arm fluoroscopy/DSA unit with a highly mobile radiolucent table. Ideally, additional diagnostic and therapeutic imaging guidance could be then provided by the addition of some type of high resolution cross-sectional imaging capability (i.e. CT and/or MR imaging). Unfortunately, prior attempts at placing a fixed site MR or CT scanner within the same proximity as a biplane angiography suite have been met with a variety of problems that limit their potential utility^2,3,5.

For example, owing to shielding and safety requirements related to high magnetic fields, most MR scanners cannot be placed very close to the C-arm image intensifiers. This has the dual consequence of substantially limiting the ability to perform combined modality image guidance of diagnostic and therapeutic interventions, and necessitating inconvenient transportation of the patient away from the angiography equipment whenever MR imaging is desired. Furthermore, a variety of compromises in either advanced imaging (e.g. diffusion/perfusion) or interventional capability will need to be accepted based on current commercially available MRI scanners.

Another problem shared by conventional fixed site CT and MRI scanners placed within an angiography suite is the need for a shared table that must be modified to adapt to the different operational requirements of the imaging systems. This usually results in making significant compromises in functionality of the table for angiographic applications (e.g. restrictions in “free float” movement, reduction in radiolucent surface area, limitations in rotational movements).

Finally, past attempts at creating a combined angiographic and cross-sectional imaging suite have encountered difficulties related to reductions in functional work space within the suite, mobility of the C-arm, and limitations in imaging certain anatomic regions related to patient positioning on the table^1,2,3.

Despite these problems, the limited clinical experience with combined angiography and cross-sectional imaging operation rooms has been mostly positive ^1,2,3, thereby supporting continued efforts to find better solutions to developing such designs. The early use of a mobile CT scanner within neurosurgical operating theatres and intensive care units are promising, demonstrating high diagnostic quality and adequate adaptability within different clinical settings ^1,4,6. We believe that incorporation of a mobile CT scanner into a conventional biplane neuroangiography suite may eliminate some of the problems faced with the fixed unit systems. Certainly based upon our recent experience with the mobile CT scanner, it has become increasingly apparent that single location multimodality imaging capability within a conventional biplane neuroangiography suite offers several logistical and technical advantages for a wide variety of applications. It is likely that these advantages will translate into enhanced clinical diagnosis and decision-making and facilitate the technical performance of various diagnostic and therapeutic interventions.

The addition of easy on-site CT imaging within the neuroangiography suite permitted greater flexibility and capability in providing rapid cross-sectional diagnosis of pathology affecting the brain, head & neck, and spine. Such information is often of great importance prior to or after both diagnostic and therapeutic neuroangiography. In most medical centres, the conventional practice is to transfer a patient to a geographically fixed CT unit either before or after an angiography procedure has been performed. This has several undesirable consequences, including untimely delays in diagnosing potentially life threatening pathology, increased risks associated with transportation of critically ill patients, delays in instituting urgent endovascular therapy, and increased patient overall procedural time ¹. These problems are either eliminated or greatly minimized with placement of the mobile CT scanner within an existing neuroangiography suite.

The combined imaging capability provided by the mobile CT scanner and high resolution biplane DSA was highly complementary and efficient for a variety of situations. After only a short period of familiarization and practice with positioning and scanning, it was possible to consistently obtain high quality cross sectional imaging of the brain, head & neck and spine in less than 15 minutes. This is considerably less time than would be necessary if a patient were to be transferred to a fixed site CT scanner (particularly in patients undergoing intensive care management or general anaesthesia).

In addition to the conventional use of CT imaging before or after an angiographic procedure was planned or considered (e.g. acute stroke intervention), several new applications were possible, including combined fluoroscopic/CT guided biopsies, discography, and nerve root injections. These procedures were performed rapidly and efficiently, minimizing patient procedure time and discomfort. Although many of these procedures can be performed with a single imaging modality, the utilization of complementary modalities may enhance departmental efficiency and patient management. Minimizing patient delays is particularly important in assessing candidates for acute intra-arterial stroke intervention and it is likely that in the future it will be possible to use the mobile CT scanner for inert xenon cerebral blood flow studies, rapidly assessing at risk regions of hypoperfusion, and allowing rapid mechanical and pharmacological intra-arterial thrombolysis. Intra-arterial CT angiography protocols may be developed and may be particularly valuable in assessing intracranial aneurysms and vascular malformations for endovascular or surgical management.

Therapeutic neurointerventional procedures were also greatly facilitated, including the performance of direct puncture embolisation of tumours and vascular malformations, percutaneous vertebroplasty and therapeutic drainages. The ability to direct and confirm needle placement, target appropriate pathologic compartments, and rapidly assess the progress of therapy may have allowed for greater efficiency in the use of medical resources (equipment and personnel) while maximizing therapeutic outcomes and minimizing patient risks, procedure times, and therapeutic sessions. Furthermore, the ability to rapidly assess the changing neurologic status of patients undergoing endovascular surgery was critical in the decision-making process before, during and immediately after performing certain high risk endovascular surgical procedures (e.g. endosaccular coiling of intracranial aneurysms, therapeutic embolisation of AVMs, superselective chemical and mechanical thrombolysis).

Despite its ease of use and feasibility within the interventional suite, most situations in which the scanner was utilized were adjunctive to the primary procedure of diagnostic or ther-apeutic angiography. Scanning may have been performed at a fixed site unit within the department without clinically significant delays or changes in management. The inability to perform adequate intra-arterial CT angiography and inert xenon blood flow studies with this unit limited the potential utility that could be obtained in the management of aneurysm, AVM, and stroke patients. Adding these features to a mobile unit may require certain software and hardware modifications which may greatly enhance its utility.

Conclusions

The use of a mobile CT scanner within the biplane neuroangiography suite can allow for access to rapid, high quality cross-sectional imaging that can significantly augment diagnostic and therapeutic decision-making and therapeutic intervention. The relative ease, timeliness and diagnostic quality of the device far exceeded our expectations. Despite several technical limitations (inability to perform CT angiography, xenon blood flow), our preliminary clinical experience with this novel approach to creating a multimodality image-guided procedure/operating suite has been positive, with demonstrated and potential utility in a wide variety of applications. An objective assessment of cost-benefit and clinical utility in a hospital or radiology department needs to be performed, yet we believe that dedicated interventional neuroradiologic or endovascular suites of the future may require the presence of multiple modalities, including CT, to optimize patient care and minimize procedural risks.