| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Department of Radiological Sciences, David Geffen School of Medicine, University of California Los Angeles, Peter V. Ueberroth Bldg., Ste. 3371, 10945 Le Conte Ave., Los Angeles, CA 90095-7206.
Received March 13, 2007;
accepted after revision May 18, 2007.
J. P. Finn is a consultant for Siemens Medical Solutions and GE
Healthcare.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Twenty-three consenting adult patients were prospectively evaluated. Images acquired were evaluated by two independent observers, and each branch vessel was scored with regard to image quality, presence and grade of stenoses, and artifacts. Interobserver agreement regarding image quality and the presence and degree of stenosis was tested using the kappa coefficient. Differences in quality ratings between the two observers were assessed using the paired Student's t test.
RESULTS. Of 828 vessels analyzed, 92.63% were designated of
diagnostic quality with no significant difference between the observers' image
quality scores (p = 0.63). Good agreement was determined regarding
image quality achieved (
= 0.716). All examinations were free of
artifact sufficient to interfere with confident interpretation. Excellent
correlation was seen with regard to stenosis detection and grading ( =
0.857). Of the external carotid artery systems assessed, 82.6% showed
conventional anatomic vascular branching.
CONCLUSION. High-spatial-resolution, 3D contrast-enhanced MR angiography at 3 T using sagittal source data acquisition and an advanced acceleration factor of 6 allows high-quality (92.63% of arterial segments) visualization of the external carotid artery system, with complete head and neck vascular coverage.
Keywords: external carotid artery • head and neck imaging • MR angiography • vascular system
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Previously, imaging at high spatial resolution required compromises in the field of view or acquisition time, potentially affecting the diagnostic content or quality of the study. With modern radiofrequency hardware architecture, it has become practical to apply parallel imaging techniques to acquire high-spatial-resolution data sets over a large field of view in a fraction of the time required for full k-space sampling [5–7]. Parallel acquisition is a generic tool, applicable to virtually all MR measurements, and it has had some of its most powerful applications in the area of contrast-enhanced MRA [8, 9]. The penalty for fractional data sampling is a reduction in signal-to-noise ratio (SNR) compared with full k-space sampling, but this drawback can be offset by imaging at higher field strengths. Moreover, the increased sensitivity to T1 shortening contrast agents at 3 T further mitigates the SNR penalty [10].
Many MRA clinical protocols today use large fields of view when imaging the head and neck vasculature, allowing craniocaudal visualization from the vertex to the transverse aorta and bilateral visualization to the subclavian or axillary arteries. Given the complexity and caliber of the supraaortic arterial and intracranial vasculature, compromise with regard to spatial resolution is unacceptable. As a result, volume coverage, as reflected in the number of acquired partitions, becomes limiting. Data acquisition in supraaortic MRA conventionally occurs in the coronal plane, so that the facial and occipital regions are incompletely visualized, as illustrated in Figure 1. In evaluating the supraaortic arteries, we have occasionally excluded such vascular anatomic regions from the field of view. Such an example is provided in Figure 2, in which a previously undiagnosed intracranial arteriovenous malformation has been incompletely imaged because of field-of-view constraints. The aim of this study was to assess the feasibility of complete bilateral external carotid artery visualization at 3 T using sagittal acquisition with a sixfold 2D integrated parallel acceleration technique (iPAT) [2].
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
All patients were positioned on the MR table in a supine orientation, with a 20-gauge IV cannula in the antecubital fossa of either upper extremity. The cannula was connected to an electronic power injector (MR Spectris, Medrad) and the patient was moved head-first into the magnet bore.
After acquisition of multiplanar single-shot localizers, transit time was estimated using a timing bolus of 2 mL of gadolinium-based contrast agent ([gadopentetate dimeglumine] Magnevist, Berlex Laboratories) injected at a rate of 1.2 mL/s and flushed with a 20-mL saline bolus at the same rate. High-resolution contrast-enhanced MRA was then performed in the sagittal plane using a fast spoiled gradient-echo sequence, the imaging parameters of which are outlined in Table 1. An asymmetric k-space sampling scheme was applied in all three planes, including the readout direction, to minimize the echo and acquisition times, and zero interpolation was performed to facilitate partial Fourier transform.
|
Two-dimensional parallel imaging, using a generalized autocalibrating partially parallel acquisition algorithm (GRAPPA) with an acceleration factor of 6 (phase-encoding direction x 3, slice-encoding direction x 2), was used, with 24 reference k-space lines for calibration in the anteroposterior phase-encoding direction and 24 references k-space lines in the through-plane right-to-left direction, 160 partitions, and a section thickness of 1.0 mm (interpolated to 0.8 mm) for complete head and neck anatomic coverage. The k-space matrix was 576 x 334 over a 430 x 262 field of view and resulted in voxel dimensions of 0.8 x 0.7 (inplane) x 0.8 mm. The contrast agent was injected at a dose of 0.15 mmol/kg of body weight of gadopentetate and rate of 1.2 mL/s, followed by a 30-mL saline flush at the same flow rate. For optimal visualization of the proximal supraaortic vasculature, patients were instructed to breath-hold for the duration of image acquisition.
All image processing was performed by a single radiologist with 5 years' experience, using a dedicated 3D workstation (Leonardo, Siemens Medical Solutions) with standard commercially available software. This radiologist was not involved in subsequent image interpretation. Using the high-resolution partitions acquired as just described, thin maximum-intensity-projection (MIP) multiplanar subvolumes, 10-mm thick and overlapping by 9 mm, and full-thickness multiplanar reconstructions (MPRs) were derived in each case and made available for image analysis. Both the source data and reconstructed volumes were made available for interpretation by each observer.
Image Analysis
Images acquired were independently evaluated by two experienced
radiologists—a neuroradiologist and a cardiovascular imaging
specialist—with 12 years and 5 years of experience, respectively, in MRA
analysis. Reviewers were blinded with regard to patient age, sex, medical
history, and indication for MRA and were instructed to evaluate the
reconstructed MIP and MPR images, using the source data for lesion analysis.
Each external carotid artery system was divided into nine segments including
the main external carotid trunk, the ascending pharyngeal artery, the superior
thyroid artery, the lingual artery, the facial artery, the posterior auricular
artery, the occipital artery, the superficial temporal artery, and the
maxillary artery. The right and left external carotid artery systems were
separately scored in each patient, with resultant scoring of 18 segments in 23
patients, constituting a total of 414 segments scored by each observer.
A 5-point image quality scoring system was used: 0, failure of vessel visualization; 1, poor image quality, inadequate for visualization of vascular detail; 2, suboptimal quality allowing only limited evaluation of vascular detail; 3, good image quality allowing assessment of vascular integrity with confidence; and 4, excellent quality affording highly confident assessment of vascular integrity. The presence and severity of artifacts were also recorded for each study using a 4-point scoring system: 1, minimal or none, not affecting interpretation; 2, mild, exceeding acceptable amount, although again without significant effect on interpretation; 3, moderate and of such degree as to impair image interpretation; and 4, severe, resulting in arterial obscuration and limiting diagnostic interpretation.
The presence of external carotid artery abnormality, such as irregularity, stenosis (graded subjectively by each observer and recorded as a percentage of luminal compromise), dissection, or aneurysmal dilatation, was also recorded when identified. Furthermore, note was made of the anatomic branch pattern observed, and each external carotid artery system was classified according to the system outlined in Figures 3,4,5,6.
|
|
|
|
= 0, poor agreement; =
0.01–0.20, slight agreement; = 0.21–0.40, fair agreement;
= 0.41–0.60, moderate agreement; = 0.61–0.80, good
agreement; and = 0.81–1.00, excellent agreement). Fisher's
analysis of variance was applied to assess relative differences in scores of
each of the individual external carotid artery branches. All statistical tests
were two-tailed, and differences with p <0.05 were regarded as
statistically significant. |
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Image Quality
In all, 92.63% of the 828 vessels analyzed were designated of diagnostic
image quality (score 3–4). Observer 1 designated 29 segments to be of
suboptimal quality, insufficient for diagnostic evaluation, versus 32 segments
for observer 2. These latter figures include vessels not visualized, even in
part, representing 19 of 29 segments in the case of observer 1 and 17 of 32
for observer 2. No significant difference was seen between the observers with
regard to image quality scores assigned in each case (p = 0.63).
Analysis using the kappa coefficient revealed good interobserver agreement
with regard to image quality achieved ( = 0.716). Analysis of variance
confirmed the presence of a statistically significant difference in scores
between the ascending pharyngeal arteries (mean, 3.33 for observer 1 and 3.43
for observer 2) relative to the other external carotid artery branches (F
score = 15.09, p < 0.05), probably a reflection of the extremely
diminutive caliber of this vessel in healthy subjects. Nonetheless, the mean
scores ascribed to these vessels were within the diagnostic quality range.
In only a single case was venous contamination designated as being of mild grade, exceeding an acceptable amount, although not degrading image interpretation; all other cases were scored as "minimal or none." Indeed, the venous contamination observed in this single case was unilateral, representing only 2.17% of the hemicervical regions imaged. All cases were designated as "minimal or none" with regard to noise, with no interference with image interpretation in any instance.
Stenosis Detection
Observer 1 reported the presence of four stenoses (0.48%) in the 828 vessel
segments analyzed, compared with two for observer 2 (0.24%). Of the two
stenoses for which both observers were in agreement, each was assigned a
stenotic grade of 50%. The remaining two stenoses reported by observer 1 were
graded as resulting in a 20% luminal stenosis, having been designated to
result in "irregularity" by the second observer. Overall,
excellent interobserver correlation was seen with regard to stenosis detection
and with regard to the assessment of the hemodynamic significance of stenoses
thus described ( = 0.857).
External Carotid Artery Branch Patterns
Overall, 82.6% of the 46 external carotid artery systems assessed showed
conventional anatomic vascular branching, with separate origins of the
superior thyroid, lingual, facial, and maxillary branches. Of the branch
patterns imaged, 13.05% (n = 6) revealed common origins of the facial
and lingual arteries (truncus linguofacialis), and 4.35% (n =2) were
designated as having common origins of the lingual and superior thyroid
arteries (truncus thyrolingualis). Notably, only in a single patient was the
aberrant branching pattern a bilateral finding, with high-resolution MRA
confirming the presence of bilateral truncus linguofacialis in a 33-year-old
woman. Direct communication between the external carotid and internal carotid
circulations was not visualized during any of these examinations.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The most common clinical indication for external carotid artery evaluation occurs before selective arterial embolization. A number of scenarios exist that may necessitate such intervention, epistaxis being the most frequently encountered in routine practice. Although endovascular embolization has now been adopted in routine clinical practice as the therapy of choice in the management of such patients [9], the incidence of major complications of this technique, such as cerebrovascular accidents, blindness, ophthalmoplegia, seizures, and anaphylaxis to contrast medium used, have been reported to occur in up to 6% of cases [11, 12].
Furthermore, embolization is contraindicated in patients in whom abnormal communications, whether congenital or acquired, exist between the internal and external carotid artery systems and in those with contrast medium allergy [13]. Advanced atherosclerosis also poses a significant challenge when embolization is being considered; the presence of considerable plaque burden complicates this procedure considerably and increases the risk of complications such as distal plaque embolization. The MRA technique described potentially plays a key role in the management of such patients, allowing preoperative noninvasive identification of patients with fistulous external-to-internal carotid artery (ECA-ICA) communication, and allowing assessment of potential offending vessels in patients in whom contrast allergy precludes a conventional approach.
External carotid artery embolization also plays a significant role in the management of a myriad of other disorders, including juvenile nasopharyngeal angiofibroma [13], meningioma [14], paraganglioma (the ascending pharyngeal artery being referred to as the "artery of the paraganglioma") [15, 16], and in the setting of trauma [17]. Again, high-spatial-resolution MRA using the technique described would allow assessment of the relationship of the these structures to the adjacent arterial vasculature and would further preoperative determination of feeding vessels for proposed embolization.
Before the advent of transfemoral endovascular carotid stenting in the management of symptomatic significant internal carotid artery stenosis, ECA-ICA bypass offered a palliative alternative to the management of patients who were unsuited for carotid endarterectomy or who had recurrent stenosis after endarterectomy. However, as follow-up of patients who have undergone this relatively recently developed endovascular approach to carotid stenosis continues, the incidence of in-stent restenosis is becoming apparent, at almost 20% at 18 months for greater than 40% stenosis and with a cumulative incidence over 5 years of clinically significant restenosis (> 80%) of 6.4% [18]. Therefore, a persistent role exists for such ECA-ICA bypass, a procedure for which preoperative external carotid artery evaluation may be of significant benefit. Indeed, Willfort-Ehringer et al. [19] determined a statistically significantly increased rate of atherosclerotic disease in the ipsilateral external carotid artery during the first year after internal carotid artery stenting [19], a pathologic process of particular concern should in-stent restenosis develop, potentially precluding subsequent bypass creation. ECA-ICA bypass has also been successfully applied to the management of other abnormal intracranial processes such as moyamoya disease (Fig. 7).
|
Considerable recent oncologic focus has centered on the promising role of selective intraarterial catheter-directed chemotherapy in the management of head and neck cancers. Tsurumaru et al. [20] reported a favorable experience in the application of this technique in the treatment of 31 patients with head and neck squamous cell carcinomas in a variety of locations. A similar approach has been described by a number of other authors, again with encouraging results, in the management of a variety of head and neck tumors, including squamous cell carcinomas of the lip [21] and larynx [22] and adenoid cystic carcinoma of the parotid gland [23]. Patients undergoing such highly focused catheter-directed chemotherapy may also benefit from preintervention MRA so that tumor feeding vessels, anatomic branch patterns, and vessel caliber may be determined.
Arteriovenous malformations are relatively uncommon in routine practice, whether congenital, spontaneous, or traumatic in origin. However, these complex anomalies demand that the therapeutic algorithm used is closely tailored to the individual patient, depending heavily on the vessels involved, growth rate, location, and symptomatology, among other parameters. Investigators have previously evaluated the potential role of numerous imaging techniques, including angiography and duplex sonography, in the assessment of arteriovenous fistulas involving the external carotid branches, determining each to be of additional diagnostic benefit, although with associated limitations, such as invasive nature and suboptimal assessment of venous outflow, respectively [24, 25].
The potential role of time-resolved MRA at 3 T in the evaluation of intracranial and extracranial vascular malformations has previously been described [26]. However, the emphasis on such sequential imaging remains focused primarily on temporal rather than spatial resolution, often precluding confident separation of overlapping arterial and venous components. Furthermore, although time-resolved imaging may use submillimeter in-plane voxel acquisition, through-plane resolution may be significantly larger. High-spatial-resolution contrast-enhanced MRA, using the technique described herein, may allow identification of morphologic arterial feeding vessels, central nidus, and efferent venous drainage using near-isotropic submillimeter voxels and facilitating MPR images. Application of a similar technique in the evaluation of intracranial aneurysms at 3 T has previously been described [27].
High-spatial-resolution contrast-enhanced MRA has been widely applied to the central and peripheral vasculature. Of relevance to the head and neck, giant cell, or temporal, arteritis is a systemic inflammatory vasculitis of unknown cause that affects medium to large vessels. Untreated, this condition may lead to visual loss or cerebrovascular accident [28]. Bley et al. [29] investigated the potential role of combined high-resolution MRI and MRA at 1.5 T in monitoring the influence of corticosteroid treatment in such patients, determining this technique to be of value in the diagnosis and monitoring of therapeutic response in this patient population. Markl et al. [30] subsequently reported similar usefulness of combined MRI and MRA at higher spatial resolution and 3-T field strength in the assessment of this condition.
Considerable variation exists with regard to anatomic branch patterns of the external carotid arteries. In most patients, this vascular configuration is of no clinical significance and is merely an anatomic curiosity. However, in certain situations, accurate depiction of the course and origin of particular external carotid terminal branch vessels assumes a central role in interventional or surgical planning, particularly in facial and reconstructive techniques. For example, the frontal branch of the superficial temporal artery has been used as a landmark for locating the course of the temporal branch of the facial nerve during rhytidectomy [31]. Temporoparietal, parietooccipital, or forehead flaps commonly used in reconstructive surgical techniques depend on the anatomic course of the superficial temporal artery. For successful flap design, adequate knowledge of the patency and pathway adopted by this vessel is essential [32].
Similarly, the temporalis muscle holds many diverse applications in facial reconstructive surgery. This muscle possesses a dependable blood supply through the middle temporal, anterior deep temporal, and posterior deep temporal branches of the superficial temporal artery, although with considerable variations in supplemental supply from the maxillary artery via its middle meningeal branch [33]. Knowledge of such variable branch patterns and muscular supply may be of value in the preoperative selection of patients suited for such reconstructive procedures. Our study, however, did not contain patients undergoing facial reconstructive surgery.
We observed separate origins of the superior thyroid, lingual, facial, and maxillary arteries in 82.6% of cases and the presence of a truncus linguofacialis in 13.05% and a truncus thyrolingualis in 4.35% of cases (Fig. 7). These results correlate well with those observed by Shima et al. [34], who reported prevalences of 76.6%, 21.7%, and 1.7%, respectively, for the presence of the same anatomic variants [34].
Our study had limitations, including the absence of gold standard conventional angiographic correlation, and this casts some doubt on the accuracy of stenosis measurements in branch vessels. However, most external carotid branches were clearly visualized in an anatomically consistent pattern, suggesting that the technique we describe can be used to map the external carotid circulation noninvasively and with confidence.
In conclusion, our study shows that high-resolution imaging of the external
carotid artery system is technically feasible using the protocol described.
Contrast-enhanced MRA at 3 T allows high-resolution depiction of the external
carotid branch vessels with a high degree of confidence regarding vascular
patency (92.63% of segments within diagnostic range) and an excellent level of
interobserver correlation with regard to the presence and degree of
atherosclerotic stenosis ( = 0.857).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
