Surgical Neurology 66 (2006) 581 – 592
www.surgicalneurology-online.com
Technique-Imaging
Intraoperative navigated 3-dimensional ultrasound
angiography in tumor surgery
Ola M. Rygh, MDa,b,d,4, Toril A. Nagelhus Hernes, PhDb,c,d, Frank Lindseth, PhDc,d,
Tormod Selbekk, MScc,d, Tomm Brostrup Mqller, PhD, MDa,d,
Geirmund Unsgaard, PhD, MDa,b,d
a
Department of Neurosurgery, St. Olav University Hospital, 7006 Trondheim, Norway
b
The Norwegian University of Science and Technology, 7006 Trondheim, Norway
c
SINTEF Health Research, 7491 Trondheim, Norway
d
National Centre for 3D ultrasound in Surgery, St. Olav University Hospital, 7006 Trondheim, Norway
Received 20 January 2006; accepted 23 May 2006
Abstract
Background: Avoiding damage to blood vessels is often the concern of the neurosurgeon during
tumor surgery. Using angiographic image data in neuronavigation may be useful in cases where
vascular anatomy is of special interest. Since 2003, we have routinely used 3D ultrasound
angiography in tumor surgery, and between January 2003 and May 2005, 62 patients with different
tumors have been operated using intraoperative 3D ultrasound angiography in neuronavigation.
Methods: An ultrasound-based neuronavigation system was used. In addition to 3D ultrasound
tissue image data, 3D ultrasound angiography (power Doppler) image data were acquired at different
stages of the operation. The value and role of navigated 3D ultrasound angiography as judged by the
surgeon were recorded.
Results: We found that intraoperative ultrasound angiography was easy to acquire and interpret, and
that image quality was sufficient for neuronavigation. In 26 of 62 cases, ultrasound angiography was
found to be helpful by visualizing hidden vessels adjacent to and inside the tumor, facilitating
tailored approaches and safe biopsy sampling.
Conclusions: Intraoperative 3D ultrasound angiography is straightforward to use, image quality is
sufficient for image guidance, and it adds valuable information about hidden vessels, increasing
safety and facilitating tailored approaches. Furthermore, with updated 3D ultrasound angiography
imaging, accuracy of neuronavigation may be maintained in cases of brain shift.
D 2006 Elsevier Inc. All rights reserved.
Keywords:
Neuronavigation; Image-guided surgery; Ultrasound; Doppler ultrasonography
1. Introduction
During tumor surgery, avoiding damage to blood vessels is
one of many concerns for the neurosurgeon. Neuronavigation
may help the neurosurgeon to locate the position of important
structures such as vessels; thus, tumor resection may be
Abbreviations: 3D, 3-dimensional; AVM, arteriovenous malformation;
BFI, blood flow imaging; CT, computed tomography; DSA, digital
subtraction angiography; FMRI, functional magnetic resonance imaging;
MRI, magnetic resonance imaging; PICA, posterior inferior cerebellar artery.
4 Corresponding author. St Olav University Hospital, Trondheim N7005 Trondheim Norway. Tel.: +47 73867596; fax: +47 73867977.
E-mail address: ola.rygh@ntnu.no (O.M. Rygh).
0090-3019/$ – see front matter D 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.surneu.2006.05.060
performed more safely and more radically. However, brain
shift as well as registration errors may limit the overall
accuracy of a neuronavigation system [3,7,9,16]. To compensate for brain shift, intraoperative imaging has been
introduced in neuronavigation [1,4], thus maintaining accuracy [26,29]. In addition, residual tumor may be detected.
In our clinic, we have used neuronavigation with
integrated intraoperative ultrasound imaging for guidance
of tumor resection for several years [4-6,25-28]. The
equipment also has capability of Doppler imaging with
power Doppler for visualization of vessels. Thus, 3D
ultrasound angiography image data can be acquired intraoperatively for neuronavigation. The aim of the present
582
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
study was to evaluate neuronavigation with intraoperative
3D ultrasound angiography (power Doppler) for imageguided resection of intracranial tumors by visualizing the
vascular anatomy adjacent to and inside the tumor.
2. Patients and methods
Between January 2003 and May 2005, we have used
intraoperative ultrasound angiography in 62 tumor cases.
Cases selected for use of neuronavigation with 3D
ultrasound angiography were those wherein neuronavigation (with 3D ultrasound) was already requested and
wherein the surgeon assumed that ultrasound angiography
might be of value.
We used an intraoperative ultrasound-based neuronavigation system, SonoWand (MISON AS, Trondheim, Norway), with a 4- to 8-MHz flat phased array probe with
optimal focusing properties at 3 to 6 cm and with the
capability of acquiring power Doppler (ultrasound angiography) images. On the ultrasound angiography images,
vessels were displayed as an overlay in shades of red over
the tissue image according to the power of the Doppler
signal. For neuronavigation with 3D ultrasound, only power
Doppler image data were applied in the present study. The
ultrasound platform (a GE Vingmed Vivid FiVe unit) in
SonoWand also has triplex imaging, but this cannot be
imported and used for neuronavigation with 3D ultrasound.
For tracking of the ultrasound probe, a tracking frame
was attached. Neuronavigation was performed with a
tracked pointer device or the CUSA (Valleylab, Boulder,
Colo, USA) with an attached tracking frame. Tracking of the
CUSA enabled continuous monitoring of the position of the
tip and the trajectory of this instrument, bonline resection.Q
A biopsy forceps with tracking was also available for image
guidance of biopsies. For acquisition of 3D ultrasound
image data for navigation, the ultrasound probe was placed
on the dura and tilted or translated over the area of interest
by a freehand movement. The 2D ultrasound images
acquired were reconstructed to a 3D ultrasound data set
ready for navigation. Image acquisition and reconstruction
typically took about 1 minute. The ultrasound probe was not
in the operating field after the acquisition of a 3D ultrasound
volume, unless additional 3D ultrasound data sets were
acquired or real-time 2D imaging was needed.
The ultrasound angiography images were displayed on
the screen of the navigation system, usually in addition to
the preoperative MRI, which was displayed simultaneously
on the same screen. With a simple pointer device, the safest
surgical route to the tumor was determined with the smallest
risk of vascular damage.
Fig. 1. Workflow in 3D ultrasound angiography. A: 3D ultrasound angiography image data are acquired by tilting or translating the probe over the area of
interest. B: Then, the calibrated CUSA is used for online resection of the tumor. The navigation system displays the position of the CUSA in relation to the
tumor margin and the vessels. C: After some resection, updated 3D ultrasound angiography data may be necessary because of brain shift (arrows) and is again
acquired by tilting or translating the probe over the area of interest.
583
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
Table 1
List of cases
Case
Diagnosis
Location of tumor
Helpfulness of ultrasound angiography
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
Chondrosarcoma
Anaplastic astrocytoma
Glioblastoma
Low-grade astrocytoma
Anaplastic astrocytoma
Chordoma
Meningioma
Metastasis (malignant melanoma)
Metastasis (lung cancer)
Glioblastoma
Chordoma
Metastasis (malignant melanoma)
Low-grade astrocytoma
Metastasis (malignant melanoma)
Trigeminal schwannoma
Glioblastoma
Low-grade astrocytoma
Meningioma (sphenoid wing)
Meningioma (olfactory)
Meningioma (left convexity)
Vestibular schwannoma
Biopsy (result: normal tissue)
Pineoblastoma
Meningioma (sphenoid wing)
Meningioma (olfactory)
Anaplastic astrocytoma
Glioblastoma
Chordoma
Ependymoma
Meningioma (sphenoid wing)
Low-grade astrocytoma
Meningioma (foramen magnum)
Glioblastoma (biopsy only)
Unknown lesion
Meningioma (posterior fossa)
Anaplastic astrocytoma
Meningioma (falx)
Glioblastoma
Meningioma (olfactory)
Pilocytic astrocytoma
Optic glioma
Meningioma (sphenoid wing)
Meningioma (parasagittal)
Low-grade glioma
Low-grade glioma
Glioblastoma
Vestibular schwannoma
Low-grade astrocytoma
Ganglioglioma
Glioblastoma
Meningioma
Ganglioglioma
Meningioma (falx)
Meningioma (convexity)
Meningioma (convexity)
Glioblastoma
Metastasis (breast cancer)
Meningioma (convexity)
Glioblastoma
Meningioma (sphenoid wing)
Vestibular schwannoma
Meningioma (convexity)
Cavernous sinus region (left side)
Left frontal lobe
Left frontal lobe
Right parietotemporal region
Left parietooccipital region
Clivus
Pineal region
Right frontal lobe
Left occipital lobe
Right temporal lobe
Cavernous sinus region (left side)
Cavernous sinus region (right side)
Left medial temporal lobe (hippocampal region)
Occipital lobe
Cavernous sinus region (right side)
Left parietal lobe
Right parietal lobe, close to central sulcus
Sphenoid wing, skull base
Anterior cranial fossa, midline
Anterior cranial fossa (left side)
Cerebellopontine angle (right)
Left temporal lobe, hippocampal region
Vermis of cerebellum
Sphenoid wing (right side)
Anterior cranial fossa, midline
Right medial temporal lobe
Left temporal lobe
Posterior fossa
Posterior fossa, fourth ventricle
Sphenoid wing (left side)
Temporal lobe (left)
Foramen magnum
Parietal lobe (right)
Left temporal lobe
Posterior fossa
Parietal lobe, midline
Falx, parietal region
Right temporal lobe
Anterior cranial fossa, skull base
Posterior cranial fossa
Extending from chiasma
Sphenoid wing (right side)
Occipital convexity (left side, parasagittal)
Right temporal lobe
Right frontal lobe
Right occipital lobe
Cerebellopontine angle (right side)
Left medial temporal lobe
Right temporal lobe
Arising from left thalamus, intraventricular
Sella region
Intraventricular (left side ventricle)
Occipital, arising from falx (left side)
Occipital convexity (right)
Anterior cranial fossa, right convexity
Occipital lobe (right side)
Right temporal lobe, sylvian fissure
Parietal convexity meningioma, right side
Left medial frontal lobe and corpus callosum
Sphenoid wing, skull base
Cerebellopontine angle, right side
Parietal region, right side
Yes
No
Yes
Possibly
No
No
Yes
Yes
No
No
Yes
Yes
Yes
No
Yes
No
Yes
Possibly
Yes
No
No
No
No
Possibly
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Possibly
No
Possibly
Yes
Possibly
Yes
No
No
Yes
Possibly
No
Yes
Possibly
Possibly
No
No
Yes
No
Yes
No
Yes
No
Yes
Possibly
No
No
List of patients with type of lesion, location of lesion, and usefulness of ultrasound angiography as judged by the surgeon.
584
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
Fig. 2. Visualization of vessels. Snapshots from the neuronavigation system. A: Low-grade glioma in medial part of temporal lobe (case 26). Magnetic
resonance imaging (T1 with gadolinium) (top) and ultrasound angiography (bottom). Vessels in the circle of Willis are visualized by ultrasound angiography;
the tumor is partially resected. B: Glioma in right temporal lobe (case 38). Magnetic resonance imaging (T1 with gadolinium) (top) and ultrasound angiography
(bottom). Vessels surrounding the tumor, probably both arterial and venous, are displayed on ultrasound angiography. The surgeon suspected that branches of
the middle cerebral artery were among these vessels. From power Doppler alone, it is not possible to tell which vessels are arterial and which are venous. Being
superficial, the vessels give a strong power Doppler signal. C: Meningioma originating from posterior part of falx (case 53). Magnetic resonance imaging (T1
with gadolinium) (top) and ultrasound angiography (bottom). Multiple veins inside the meningioma drain into the straight sinus. D: Thalamic glioma (case 50).
Magnetic resonance imaging (T1 with gadolinium) (top) and ultrasound angiography. The thalamostriate vein is revealed by ultrasound angiography, crossing
the tumor surface. E: Meningioma, foramen magnum (case 32). Magnetic resonance imaging (T1, top) and ultrasound angiography (bottom). Notice branches
from vertebral arteries adjacent to the tumor. F: Pilocytic astrocytoma in posterior fossa (case 40). Ultrasound tissue (top) and ultrasound angiography (bottom).
The vein of Galen is visualized (arrow) by ultrasound angiography.
Additional ultrasound and ultrasound angiography image
data were acquired after some resection, if required, because
of suspected brain shift or for resection control. The
workflow with the equipment is summarized in Fig. 1.
Additional resection was usually carried out if residual
tumor was detected, except in cases where this was decided
against because of unacceptable risk of neurological damage
to the patient.
585
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
Table 2
Cases where navigated 3D ultrasound angiography was found helpful
Case
Diagnosis
Location of tumor
Vessel visualized
Role of navigation
1
Chondrosarcoma
Internal carotid artery
Visualization of hidden vessel
3
7
Glioblastoma
Meningioma
Cavernous sinus
region (left side)
Left frontal lobe
Pineal region
Middle cerebral artery
Great cerebral vein
8
Metastasis
(malignant melanoma)
Chordoma
Right frontal lobe
Pericallosal arteries
Cavernous sinus
region (left side)
Cavernous sinus
region (right side)
Internal carotid artery
Visualization of hidden
Visualization of hidden
online CUSA resection
Visualization of hidden
online CUSA resection
Visualization of hidden
Left medial temporal
lobe (hippocampal region)
Cavernous sinus
region (right side)
Right parietal lobe,
close to central sulcus
Anterior cranial fossa,
midline
Anterior cranial fossa,
midline
Right medial temporal lobe
Middle cerebral artery
Visualization of hidden vessel,
facilitation of approach,
online CUSA resection
Visualization of hidden vessel
Internal carotid artery
Visualization of hidden vessel
Branches from middle
cerebral artery
Anterior cerebral artery
Visualization of hidden vessel
11
12
Metastasis
(malignant melanoma)
13
Low-grade astrocytoma
15
Trigeminal schwannoma
17
Low-grade astrocytoma
19
26
Meningioma
(olfactory)
Meningioma
(olfactory)
Anaplastic astrocytoma
29
Ependymoma
30
Meningioma
(sphenoid wing)
Low-grade astrocytoma
25
31
32
Posterior fossa,
fourth ventricle
Sphenoid wing
(left side)
Left temporal lobe
33
Meningioma
(foramen magnum)
Glioblastoma (biopsy only)
34
Unknown lesion
Left medial
temporal lobe
40
42
45
Pilocytic astrocytoma
Meningioma
(sphenoid wing)
Low-grade glioma
Posterior cranial fossa
Sphenoid wing
(right side)
Right frontal lobe
48
Low-grade astrocytoma
53
Meningioma (falx)
55
Meningioma
(convexity)
Metastasis
(breast cancer)
Glioblastoma
Left medial
temporal lobe
Occipital, arising
from falx (left side)
Right anterior
cranial fossa
Right temporal
lobe in sylvian fissure
Left medial frontal
lobe and corpus callosum
57
59
Internal carotid artery
Anterior cerebral artery
Middle cerebral artery
and internal cerebral vein
Branches from PICA
Anterior and middle
cerebral arteries
Middle cerebral artery
Foramen magnum
Vertebral arteries
Right parietal lobe
Smaller vessels
Small cortical arteries
on medial surface of
temporal lobe
Vein of Galen
Middle cerebral artery
and internal carotid
Pericallosal artery
The procedures were performed by 3 surgeons, and most
of the procedures (51 cases) were performed by the senior
author (GU). The value and role of navigated ultrasound
angiography as judged by the surgeon were recorded
postoperatively in most cases. In addition, we have
retrospectively reviewed all cases. The subjective judgment
Middle cerebral artery
Posterior cerebral artery
and straight sinus
Pericallosal arteries
Middle cerebral artery
Pericallosal arteries
Visualization of hidden
online CUSA resection
Visualization of hidden
online CUSA resection
Visualization of hidden
tailored approach
Visualization of hidden
vessel
vessel,
vessel,
vessel
vessel,
vessel,
vessel,
vessel
Visualization of hidden vessel,
online CUSA resection
Tailored approach, visualization
of hidden vessel, online
CUSA resection
Visualization of hidden vessel,
online CUSA resection
Safe biopsy sampling,
visualization of hidden vessel
Tailored approach,
visualization of hidden vessel
Visualization of hidden vessel
Visualization of hidden vessel,
online CUSA resection
Visualization of hidden vessel,
online CUSA resection
Facilitation of approach
Visualization of hidden vessel,
online CUSA resection
Visualization of hidden vessel,
online CUSA resection
Visualization of hidden vessel
Visualization of hidden vessel,
facilitation of approach,
online CUSA resection
by the surgeon of the value of navigated ultrasound
angiography was graded into 3 categories: helpful, possibly
helpful, and not helpful. The degree of helpfulness was
graded as helpful when navigated ultrasound angiography
was used actively during the whole procedure and was
considered to contribute significantly to the safe completion
586
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
of the procedure. The degree of helpfulness was considered
possibly helpful when navigated ultrasound angiography
was considered useful for orientation but not considered
essential for safe navigation. When ultrasound angiography
did not give additional valuable information for image
guidance, it was deemed not helpful.
The cases are listed in Table 1.
3. Results
Between 2003 and May 2005, in 62 of 108 patients
operated with image guidance, 3D ultrasound angiography
was applied and evaluated. Among these patients, there
were 20 meningiomas, 24 gliomas, 5 metastases, and 11
tumors of other types. There were also 2 frameless
Fig. 3. Chordoma in cavernous sinus region (case 11). A: Sagittal MRI. Green arrows outline the tumor, red arrow directed at internal carotid artery. B: Coronal
MRI. Green arrows outline the tumor, red arrows directed at internal carotid artery. C: Photograph from the microscope. During surgery, the internal carotid
artery was revealed. White arrows directed at the internal carotid artery. D: MRI image slice from a snapshot from the navigation system during surgery. Green
arrows directed at the tumor borders. E: Ultrasound angiography from the same snapshot as panel (D) visualizing the internal carotid artery (in shades of red),
which in this case was adjacent to the medial parts of the tumor. The tumor in this case was hypoechogenic. Red arrow directed at artery. A flash artifact is
visible upward to the right of the artery (white arrow). F: Photograph from the microscope. The pointer is directed at the internal carotid artery (indicated by
white and black arrows). The snapshot from which panels (D) and (E) originate was taken about the same time this photograph was taken, demonstrating good
accuracy of the ultrasound angiography navigation, as the position of the pointer directed at the artery seem to correspond with the position as indicated by the
navigation system. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
stereotactic biopsies performed with the use of navigated 3D
ultrasound angiography.
3D ultrasound angiography image data were successfully
acquired and visualized in all cases.
Both arteries and veins (Fig. 2) were visualized via ultrasound angiography. Vessels in range of the probe gave a
strong Doppler signal (Fig. 2). In some cases, the simultaneous display of preoperative MRI helped to decide the
identity of vessels, by giving anatomical overview, as the 3D
ultrasound angiography only displayed the part of the anatomy that was covered in the 3D ultrasound image acquisition.
Image guidance with 3D ultrasound angiography was
performed with a pointer device or tracked surgical instruments, such as the tracked CUSA or a tracked biopsy forceps.
With tracking of the CUSA, the surgeon was able to see the
position of the tip of the CUSA in relation to the tumor
capsule or margin, as well as the vessels of interest (Fig. 1).
We found this to be a useful safety measure during resection
when there were vessels adjacent to or inside the tumor.
In some of the cases (cases 26, 31, 34, 39, 44, 49, and
50), the role of navigation with 3D ultrasound angiographic
587
imaging was to tailor an approach with minimal invasiveness. With angiographic image guidance, a safe route to the
tumor could be chosen, with less need for dissecting and
visually inspecting vessels.
With tracking of the biopsy forceps, frameless stereotactic biopsies were acquired with image guidance. Angiographic imaging with ultrasound was found to be helpful in
deciding sites for biopsy sampling with minimal risk of
bleeding. We found that ultrasound angiography sufficiently
displayed small vessels (about z 0.5 mm) in the brain tissue
for safe image-guided biopsy sampling. One patient in our
series (case 57) had an infarction of the middle cerebral
artery, which passed close to a metastasis. Except this case,
we have not recorded any vascular complications in
our series.
The cases where navigated ultrasound angiography was
found to be helpful and the role of navigated ultrasound
angiography in those cases are presented in Table 2.
In 26 (42%) of the cases, ultrasound angiography was
found to be helpful, in 10 (16%) possibly helpful, and in 26
(42%) not helpful.
Fig. 4. Large olfactory meningioma (case 25). Green arrows outline tumor. A: CT image showing the tumor. B: DSA showing branches from the anterior
cerebral artery being displaced by the tumor (white arrows). C: Ultrasound angiography showing the same branches as indicated in panel (B) from the anterior
cerebral artery (white arrows) adjacent to the tumor. D: Ultrasound angiography before the resection. E: Ultrasound angiography during resection. F: Ultrasound
angiography after tumor removal. The branches from the anterior cerebral artery are clearly visualized during these phases of the operation. Updated 3D
ultrasound angiography image data ensure that accuracy is good even if brain shift has occurred. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article).
588
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
4. Case descriptions
4.1. Case 11
A 58-year-old woman complained of hoarseness and
swallowing difficulties. A left recurrent laryngeal nerve
paresis was found on clinical examination, and MRI
investigation showed a tumor in the left middle cranial
fossa and jugular fossa with close anatomical relation to the
internal carotid artery (Fig. 3A and B). She was operated
using a pterional approach. The tumor was found to be a
chordoma by frozen sections, confirmed later by histopathological analysis. Before reaching the tumor, 3D ultrasound
and 3D ultrasound angiography image data were acquired.
During surgery, navigating with the pointer and the
calibrated CUSA showed the position of the internal carotid
artery in relation to the tip of the instrument. This was found
valuable for safe resection because the artery was not visible
in this phase of the operation. Further dissection revealed
the carotid artery (Fig. 3C), and we could verify that the
navigation system showed the correct position of the artery
when directing the pointer at the artery (Fig. 3E and F).
Using neuronavigation with ultrasound angiography
images, surgery was performed without any complications,
and the patient had an uneventful recovery without any new
neurological deficits.
4.2. Case 25
A 52-year-old man presented with progressive dementia.
A CT scan revealed a large interhemispheric tumor in the
anterior cranial fossa (Fig. 4A). Preoperative DSA showed
branches of the anterior cerebral artery passing close to the
tumor (Fig. 4B). An image-guided extirpation of the tumor
via the bifrontal approach was performed. During surgery,
intraoperative 3D ultrasound angiography showed branches
from the anterior cerebral arteries and their close relation to
the tumor (Fig. 4C). Debulking and resection of the tumor
were performed with tracked CUSA, which enabled monitoring of the CUSA tip in relation to the vessels, clearly
revealed by 3D ultrasound angiography. Several 3D ultrasound angiography image acquisitions were performed
during different stages of the operation (Fig. 4D-F), ensuring
that inaccuracy due to brain shift was minimal. Neuronavigation with intraoperative ultrasound angiography thus
served to increase the safety of the procedure. The postoperative recovery was uneventful. At follow-up, the patient
had a significant improvement in cognitive function.
5. Discussion
In image-guided surgery, angiographic imaging may be
valuable in addition to tissue imaging. In image guidance of
aneurysm and AVM surgery, preoperative angiographic
image data, MR Angiography and CT Angiography, have
successfully been imported and used in neuronavigation
[2,13,20,21,27] and have been found to be valuable for
localization of aneurysms and AVM feeders. In addition, in
tumor surgery, preoperative angiographic image data (CT
angiography and MR angiography) may be used in neuronavigation. Advanced neuronavigation with CT angiography has been found to be valuable for tailoring surgical
approaches and identification of hidden vessels (ie, vessels
not exposed in the surgical field) during resection [17].
However, in cases with brain shift, preoperative images may
become inaccurate during surgery [3,7,16]. Intraoperative
imaging, tissue and angiographic, may therefore be necessary to maintain the accuracy in neuronavigation. Both
intraoperative MRI and ultrasound are modalities for
acquiring intraoperative angiographic image data for use
in neuronavigation. Intraoperative ultrasound integrated
with neuronavigation is now a convenient alternative, as
modern ultrasound Doppler technology is capable of
depicting intracranial vasculature with sufficient image
quality [5,6,19,24,30,31].
Sure et al [23] has reported land-marking of vessels in a
conventional neuronavigation system with the help of a
tracked 2D ultrasound probe using color flow Doppler and
found that marking the position of vessels adjacent to a
tumor facilitated image-guided tumor resection. The solution we have used enables navigation directly in the 3D
ultrasound angiography images without need for landmarking on a preoperative MRI data set. In addition,
updated 3D ultrasound angiographic image data may
quickly be acquired and used in neuronavigation.
Not surprisingly, we found navigated 3D ultrasound
angiography to be of value during surgery on tumor cases
with large and important vessels adjacent to or embedded in
the tumor (Table 2). Five of these cases (patients 19, 25, 30,
32, and 40) were patients with large meningiomas arising
from the skull base, displacing important arteries. With
online resection using the CUSA, monitoring the position of
the tip of the CUSA relative to the position of vessels helped
the surgeon to do a more complete resection of the
meningiomas before starting the dissection of the capsule,
probably reducing the traumatization of the cerebral cortex
during this phase. Updated 3D ultrasound angiography
image data gave angiographic images of sufficient quality
for navigation during the different stages of the resection. In
addition, in 9 cases (patients 10, 13, 22, 26, 31, 34, 44, 48,
and 49) with tumors in the medial part of the temporal lobe,
we found that visualization of vessels in the basal cistern
(eg, the great cerebral vein and the middle cerebral artery)
was helpful for safer resection and biopsy sampling.
In other cases, ultrasound angiography was not found
helpful, such as in superficial gliomas not close to larger
vessels. Nevertheless, even in such cases, unexpected
vessels may give the surgeon unpleasant surprises, and
confirmation that important vessels were not present close to
the tumor was found reassuring by the surgeon.
In skull base surgery, the vessels are usually attached
to the skull base and are therefore not subject to brain
shift [3,22]. Nevertheless, in MRI-based neuronavigation,
inaccuracy in image registration may still lead to inaccurate
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
targeting of skull base vessels. As no image registration is
needed in neuronavigation based on 3D ultrasound angiography [28], this technique is still valuable because image
registration errors are avoided.
One limitation of the 3D ultrasound–based neuronavigation is that covering the whole area of interest during 3D
ultrasound image data acquisition can be difficult, for
example, in cases with large tumors. Furthermore, in skull
base surgery, the skull base itself may hinder ultrasound
imaging of the entire tumor, for example, in cases where a
large skull base tumor is invading both the middle and the
posterior cranial fossa. Still, it is our experience that careful
planning of the surgical approach and keeping in mind
optimal positioning of the ultrasound probe may reduce
such problems to a minimum. Moreover, simultaneous
display of MRI and 3D ultrasound in neuronavigation may
be helpful, giving overview and anatomical orientation in
cases with large tumors that are difficult to cover entirely
with ultrasound.
589
6. Technical considerations
For angiographic imaging with the Doppler technique,
both color flow and power Doppler technique may be
used. Power Doppler displays the vessels because of the
power of the Doppler signal in shades of red as an overlay
on the ultrasound tissue image. In contrast to color flow
Doppler imaging, power Doppler does not have flow
velocity or direction information but is less angle-dependent, there is no aliasing and the modality is more flow
sensitive [15,18]. Because of less angle dependence, vessel
continuity is also better with power Doppler than with
color flow Doppler, and vessels will be visualized even if
the ultrasound beam hits the vessel at angles close to 908.
In an application with free-hand 3D ultrasound image
acquisition, the ultrasound beam will almost always hit a
vessel with several different angles; therefore, there is
minimal risk of missing a vessel when acquiring 3D ultrasound angiography image data.
Fig. 5. Multimodal visualization. A: Glioma (same patient as in Fig 2B, case 38). Ultrasound image slice placed in a 3D scene with a segmented representation
of the MR-imaged tumor (yellow). Ultrasound angiography (green) and MR angiography (red) show the vessels in relation to the tumor. The ultrasound
angiography and MR angiography do not completely overlap, and this may be due to brain shift or registration error. The ultrasound angiography displays more
vessels than the MR angiography, and the vessels appear thicker. Still, the vascular anatomy is recognized on both ultrasound angiography and MR
angiography. The ultrasound angiography only shows parts of the anatomy. B: Meningioma (same patient as in Fig 2C, case 53). Ultrasound angiography
(segmented) here shown in red. Image slices from T1 MRI and ultrasound give orientation and show the tumor. C: Foramen magnum meningioma (same
patient as in Fig 2E, case 32). A volume rendered representation of the tumor is shown in grayscale, whereas MR angiography is shown in red and ultrasound
angiography in green. The visualization in a 3D scene demonstrates the spatial relationships between the tumor and vessels. (For interpretation of the references
to color in this figure legend, the reader is referred to the web version of this article).
590
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
In contrast to MR angiography, ultrasound angiography
shows both arteries and veins at the same time. We
consider this an advantage, as one usually is interested in
preserving both arteries and veins during surgery. It is
however not possible to discriminate between arteries and
veins using the power Doppler imaging technique alone.
Still, adjusting the blow velocity rejectQ setting on the
ultrasound scanner may be helpful for showing only vessels
with high flow velocities, in which case, most veins will be
bfiltered out.Q Using triplex display may also be of help to
discriminate between arteries and veins by evaluating the
Doppler spectrum. Using the ultrasound probe as a pointer
will enable targeting of a vessel with neuronavigation while
using triplex display, and in this way, it is possible to
evaluate the targeted vessel with triplex. We did however
use triplex in the present study. With larger vessels, it was
often possible to decide whether the targeted vessel was an
artery or a vein by using anatomical knowledge and
comparing with arterial MR angiography in cases where
this was available.
In ultrasound angiography imaging, the ultrasound data
are used for simultaneous imaging of both tissue and
angiographic imaging. We have not experienced wherein the
quality of the ultrasound tissue part of the image is
significantly decreased on ultrasound angiography images.
There are, however, some drawbacks of 3D ultrasound
angiography with power Doppler. The high sensitivity of
power Doppler may lead to the visualization of too many
vessels, as small and surgically less important vessels are
depicted. Sometimes, this may result in confusing images
that are difficult to interpret. In our experience, reducing the
gain setting on the ultrasound scanner may reduce this
problem by filtering out the smaller vessels. For better
orientation, simultaneous display of preoperative MRI and
MR angiography is also useful in such cases.
Blooming may also be a problem with power Doppler, as
the power Doppler signal tends to expand beyond the vessel
walls; consequently, the vessels may seem larger and thicker
than they really are. In addition, the 2D ultrasound image
plane itself has a certain thickness that varies with the depth;
thus, the resolution in the elevation direction (normal to the
scan plane) of the ultrasound beam affects the image quality
in 3D ultrasound. Therefore, the vessels in reconstructed
(ultrasound angiography) images tend to appear bulkier
when viewed on slices orthogonal to the scan plane.
Finally, power Doppler is relatively sensitive to flash
artifacts, which occur as red fields or stripes (Fig. 2). Flash
artifacts may, for example, occur when the probe is in a
cavity filled with saline, when motion in the saline is
detected by the power Doppler (Fig. 3C). This is another
consequence of the high motion sensitivity of power
Doppler, which can occur relatively often during free-hand
3D ultrasound angiography acquisition. Flash artifacts can
be reduced with gentle movement of the probe during image
acquisition and by adjustments on the ultrasound scanner,
such as adjusting the lower limit for detectable flow.
The overall clinical accuracy of the SonoWand system
may be as low as 2 mm in a clinical setting when using
intraoperative imaging to compensate for brain shift [11].
We did not systematically measure accuracy in this study.
Nonetheless, in each case a vessel was exposed in the
surgical field, we directed the pointer at the artery for
accuracy testing, as in Fig. 3F. In each case, we observed
that the position of the pointer corresponded with the
position indicated by the navigation system, as shown in
Fig. 3E. Although not a precise measurement of accuracy,
we found these accuracy tests to indicate that the clinical
accuracy of the navigation system was satisfactory.
7. Limitations of this study
This technical evaluation study of navigated 3D
ultrasound angiography demonstrates the straightforward
application of neuronavigation with this imaging modality
in 62 cases. To prove any benefit for the patient, a study
with a different design is needed. The cases selected for
evaluation of ultrasound angiography were cases wherein
ultrasound-based neuronavigation was chosen in advance.
Small, extra-axial tumors that were easy to localize were
not included in this material because they were not
operated with image guidance. In other tumor cases, which
were operated with image guidance, ultrasound angiography was not used because the surgeon felt that it would not
be of value. Most of the operations, and thus also the
evaluation of the helpfulness of ultrasound angiography,
were performed by one surgeon. Still, the 3 surgeons who
participated seemed to judge the usefulness of 3D
ultrasound angiography similarly. Nevertheless, different
surgeons may judge the helpfulness of navigated ultrasound
angiography differently, for example, some surgeons would
not trust the navigation system enough to leave out
dissection of important vessels. All this has to be kept in
mind when interpreting the results of this study.
8. Future prospects
Ultrasound angiography still has potential for significant
improvements, and developments along several venues are
to be expected:
New signal processing methods such as BFI [12] may
better visualize flow inside vessels and hopefully reduce the
problem of blooming. Ultrasound contrast-enhancing agents
have been reported to be helpful in tumor neurosurgery for
assessing the vasculature close to and inside the tumor
[8,14] and may improve the quality of ultrasound angiographic imaging.
New multirow probes will have a more optimal beam
shape with better resolution in the elevation direction (eg,
normal to the scan plane). This will further improve the
image quality of ultrasound angiography because elevation
resolution is a limiting factor for image quality with the
present technology. Real-time 3D probes will be able to
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
acquire 3D ultrasound data sets directly without free hand
movement, and this will minimize flash artifacts.
New multimodal visualization techniques where preoperative MR and intraoperative ultrasound are integrated in
the same 3D scene are already available [10]. This may
probably further enhance the surgeon’s perception of
anatomic and spatial relationships between tumor and
adjacent vessels (Fig. 5). Diffusion tensor imaging of tracts
and FMRI of eloquent cortex may also be visualized.
Furthermore, robust volume-to-volume registration techniques for registration of preoperative MR angiography data
to intraoperative ultrasound angiography data may make it
possible to adjust preoperative MR image data in cases of
brain shift.
9. Conclusions
On the basis of our experience with navigated intraoperative ultrasound angiography, we believe that this
imaging modality is sufficient for intraoperative imaging
of arteries and veins. Image guidance with 3D ultrasound
angiography may increase the safety in surgery of tumor
cases where an important vessel is adjacent to or inside a
tumor. It may facilitate tailored approaches, and frameless
stereotactic biopsies may be performed with increased
safety. Furthermore, with updated 3D ultrasound angiography imaging, accuracy of angio-based neuronavigation may
be maintained in cases of brain shift.
Acknowledgments
This work was supported by the Norwegian Research
Council.
References
[1] Albayrak B, Samdani AF, Black PM. Intra-operative magnetic
resonance imaging in neurosurgery. Acta Neurochir (Wien)
2004;146:543 - 56 [discussion 57].
[2] Coenen VA, Dammert S, Reinges MH, et al. Image-guided microneurosurgical management of small cerebral arteriovenous malformations: the value of navigated computed tomographic angiography.
Neuroradiology 2005;47:66 - 72.
[3] Dorward NL, Alberti O, Velani B, et al. Postimaging brain distortion:
magnitude, correlates, and impact on neuronavigation. J Neurosurg
1998;88:656 - 62.
[4] Gronningsaeter A, Kleven A, Ommedal S, et al. SonoWand, an
ultrasound-based neuronavigation system. Neurosurgery 2000;
47:1373 - 9 [discussion 79-80].
[5] Gronningsaeter A, Unsgard G, Ommedal S, et al. Ultrasound-guided
neurosurgery: a feasibility study in the 3-30 MHz frequency range. Br
J Neurosurg 1996;10:161 - 8.
[6] Hernes TA, Ommedal S, Lie T, et al. Stereoscopic navigationcontrolled display of preoperative MRI and intraoperative 3D
ultrasound in planning and guidance of neurosurgery: new technology
for minimally invasive image-guided surgery approaches. Minim
Invasive Neurosurg 2003;46:129 - 37.
[7] Hill DL, Maurer Jr CR, Maciunas RJ, et al. Measurement of
intraoperative brain surface deformation under a craniotomy. Neurosurgery 1998;43:514 - 26 [discussion 27-28].
591
[8] Kanno H, Ozawa Y, Sakata K, et al. Intraoperative power Doppler
ultrasonography with a contrast-enhancing agent for intracranial
tumors. J Neurosurg 2005;102:295 - 301.
[9] Letteboer MM, Willems PW, Viergever MA, et al. Brain shift
estimation in image-guided neurosurgery using 3-D ultrasound. IEEE
Trans Biomed Eng 2005;52:268 - 76.
[10] Lindseth F, Kaspersen JH, Ommedal S, et al. Multimodal image
fusion in ultrasound-based neuronavigation: improving overview and
interpretation by integrating preoperative MRI with intraoperative 3D
ultrasound. Comput Aided Surg 2003;8:49 - 69.
[11] Lindseth F, Lango T, Bang J, et al. Accuracy evaluation of a 3D
ultrasound-based neuronavigation system. Comput Aided Surg
2002;7:197 - 222.
[12] Loevstakken L, Bjaerum S,Martens D, et al. Real-time blood
motion imaging a 2D blood flow visualization technique. IEEE
International Ultrasonics symposium, Montreal, Canada, Aug. 24-27,
2004. 1: 602.
[13] Muacevic A, Steiger HJ. Computer-assisted resection of cerebral
arteriovenous malformations. Neurosurgery 1999;45:1164 - 70 [discussion 70-1].
[14] Otsuki H, Nakatani S, Yamasaki M, et al. Intraoperative ultrasound
arteriography with the bCoded Harmonic AngioQ technique. Report of
three cases. J Neurosurg 2001;94:992 - 5.
[15] Riccabona M, Resch B, Eder HG, et al. Clinical value of amplitudecoded colour Doppler sonography in paediatric neurosonography.
Childs Nerv Syst 2002;18:663 - 9.
[16] Roberts DW, Hartov A, Kennedy FE, et al. Intraoperative brain shift
and deformation: a quantitative analysis of cortical displacement in
28 cases. Neurosurgery 1998;43:749 - 58 [discussion 58-60].
[17] Rohde V, Spangenberg P, Mayfrank L, et al. Advanced neuronavigation in skull base tumors and vascular lesions. Minim Invasive
Neurosurg 2005;48:13 - 8.
[18] Rubin JM. Power Doppler. Eur Radiol 1999;9(Suppl 3):S318-22.
[19] Rubin JM, Hatfield MK, Chandler WF, et al. Intracerebral arteriovenous malformations: intraoperative color Doppler flow imaging.
Radiology 1989;170:219 - 22.
[20] Russell SM, Woo HH, Joseffer SS, et al. Role of frameless stereotaxy
in the surgical treatment of cerebral arteriovenous malformations:
technique and outcomes in a controlled study of 44 consecutive
patients. Neurosurgery 2002;51:1108 - 16 [discussion 16-18].
[21] Schmid-Elsaesser R, Muacevic A, Holtmannspotter M, et al. Neuronavigation based on CT angiography for surgery of intracranial
aneurysms: primary experience with unruptured aneurysms. Minim
Invasive Neurosurg 2003;46:269 - 77.
[22] Sure U, Alberti O, Petermeyer M, et al. Advanced image-guided skull
base surgery. Surg Neurol 2000;53:563 - 72 [discussion 72].
[23] Sure U, Benes L, Bozinov O, et al. Intraoperative landmarking of
vascular anatomy by integration of duplex and Doppler ultrasonography in image-guided surgery. Technical note. Surg Neurol 2005;
63:133 - 41.
[24] Tekula F, Pritz MB, Kopecky K, et al. Usefulness of color Doppler
ultrasound in the management of a spinal arteriovenous fistula. Surg
Neurol 2001;56:304 - 7.
[25] Unsgaard G, Gronningsaeter A, Ommedal S, et al. Brain operations
guided by real-time two-dimensional ultrasound: new possibilities as a
result of improved image quality. Neurosurgery 2002;51:402 - 11
[discussion 11-12].
[26] Unsgaard G, Ommedal S, Muller T, et al. Neuronavigation by
intraoperative three-dimensional ultrasound: initial experience during
brain tumor resection. Neurosurgery 2002;50:804 - 12 [discussion 12].
[27] Unsgaard G, Ommedal S, Rygh O, et al. Operation of arteriovenous
malformations assisted by stereoscopic navigation-controlled display
of preoperative magnetic resonance angiography and intraoperative
ultrasound angiography. Neurosurgery 2005;56:281 - 90.
[28] Unsgaard G, Rygh OM, Selbekk T, et al. Intra-operative
3D ultrasound in neurosurgery. Acta Neurochir (Wien) 2006;148:
235 - 53.
592
O.M. Rygh et al. / Surgical Neurology 66 (2006) 581–592
[29] Wirtz CR, Knauth M, Staubert A, et al. Clinical evaluation and
follow-up results for intraoperative magnetic resonance imaging in
neurosurgery. Neurosurgery 2000;46:1112 - 20 [discussion 20-2].
[30] Woydt M, Greiner K, Perez J, et al. Intraoperative color duplex
sonography of basal arteries during aneurysm surgery. J Neuroimaging 1997;7:203 - 7.
[31] Woydt M, Perez J, Meixensberger J, et al. Intra-operative colourduplex-sonography in the surgical management of cerebral AVmalformations. Acta Neurochir (Wien) 1998;140:689 - 98.
Commentary
In the past several years, intraoperative imaging has
gained increasing importance. With the newest medicotechnical achievements such as high-field (1.5–3 T) MRI
and multislice CT, a certain renaissance of intraoperative
imaging has started. However, the setting up of such
diagnostic tool requires a major financial background.
Consequently, Unsgaard et al started to promote the
intraoperative navigated ultrasound technology by introducing a 3D ultrasound navigation system. In previous
publications, this group described a very nice system that
is capable of producing a 3D imaging ultrasound data set
that is used for navigation and can be supported by the
display of preoperatively acquired corresponding MRI
images. In their article, the authors demonstrate the
integration of power Doppler (Duplex or color mode)
signals in their 3D navigation data set. The figures, as
shown by the authors, are of very high quality; let us hope
that the drawbacks of the intraoperative ultrasound imaging
of vessels, such as the exaggeration of the caliper
(blooming), can be reduced in future.
Frequently, for us neurosurgeons, the vascular anatomy is
the most interesting part to recognize in a navigational data
set. Therefore, we believe that particularly the 3D visualization of the vascular anatomy, as presented here, is of
major benefit during surgery of a broad variety of
intracranial pathologies. In our experience, such technology
is of particular benefit when vessels are encased or distorted
by tumors or when evaluation of hemodynamics within a
complex AVM is intended [1]. The usefulness of a similar
technique for treatment of vascular lesions has also been
previously illustrated by the same group [2]. One of the
most interesting questions to answer in the future will be
whether the ultrasound technology will allow to evaluate the
neck of a clipped aneurysm to exclude a remnant.
However, alongside the economic aspect, acquisition
time and quality of intraoperative visualization of the
vascular anatomy are further advantages of the ultrasound
technology when compared with the MRI and/or CT
intraoperative imaging solutions.
Dorothea Miller, MD
Oliver Bozinov, MD
Ulrich Sure, MD
Clinic for Neurosurgery, University Hospital
Philipps-University Marburg
Marburg, 35033 Germany
References
[1] Tirakotai W, Miller D, Heinze S, Benes L, Bertalanffy H, Sure U. A
novel platform for image-guided ultrasound. Neurosurgery 2006;58:
710 - 8.
[2] Unsgaard G, Ommedal S, Rygh OM, Lindseth F. Operation of
arteriovenous malformations assisted by stereoscopic navigationcontrolled display of preoperative magnetic resonance angiography
and intraoperative ultrasound angiography. Neurosurgery 2005;
56(2 Suppl):281 - 90.