Ultrasound Obstet Gynecol 2012; 40: 576581

Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/uog.11134

Three-dimensional power Doppler ultrasound for the study

of cervical cancer and precancerous lesions
P. BELITSOS, D. PAPOUTSIS, A. RODOLAKIS, S. MESOGITIS and A. ANTSAKLIS
1st Department of Obstetrics and Gynecology, University of Athens, Alexandra Hospital, Athens, Greece

K E Y W O R D S: cervical cancer; cervical intraepithelial neoplasia; power Doppler; three-dimensional

ABSTRACT
Objective To evaluate the blood flow characteristics of
the cervix in normal women and in women with cervical
precancerous lesions or cervical cancer.
Methods We studied 165 women with three-dimensional
power Doppler ultrasound (3D-PDU), of whom 71 had
cervical cancer, 61 had precancerous lesions and 33 were
healthy controls. The cervix was manually traced in the
stored volumes using 15 rotation steps and the following
3D-PDU indices were calculated: vascularization index
(VI), flow index (FI) and vascularization flow index (VFI).
These indices were compared among the study groups and
were also correlated with features of the precancerous
lesions group and cancer group.
Results The three indices were all statistically significantly
higher in the cervical cancer group and precancerous
lesions group than in controls (P < 0.001). In addition,
significantly higher values of all indices were found
in the cervical cancer group than in the precancerous
lesions group (P < 0.001). Further analysis according
to patient characteristics in the cancer group showed
that VI, FI and VFI were not significantly different in
relation to grade, histology, presence of positive lymph
nodes or lymphovascular space involvement (P > 0.05).
However, VI was significantly higher in patients with
Stages IIIBIV cancer than in patients with less advanced
disease (P = 0.045). In the cervical cancer group there was
a significant positive correlation between 3D-PDU indices
and cervical volume.
Conclusion 3D-PDU assessment of the cervix reveals
significant differences in all indices studied between
women with cervical precancerous lesions or cancer
and healthy women. In women with cervical cancer,
an advanced stage is associated with higher VI, but
3D-PDU indices are not related to other pathological
characteristics. Copyright 2012 ISUOG. Published by
John Wiley & Sons, Ltd.

INTRODUCTION
Angiogenesis plays a critical role in the development
of tumors, as well as their invasion and metastasis1 3 .
One method of studying the process of angiogenesis
in tumors, including cervical cancer, is by assessing
histological sections for the intratumoral microvessel
density4 . However, counting the vessels depends on the
area examined in each section and therefore the overall
assessment of vascularity is inconsistent and does not give
a precise estimation of blood flow inside the tumor5 .
Three-dimensional power Doppler ultrasound (3DPDU) has emerged in the last decade as a new
method for assessing the vascularity of solid organs,
allowing non-invasive assessment of the vasculature
and blood flow in various tissues. In-vivo studies have
shown that the vascularization index (VI) correlates
positively with microvessel density as determined by
immunohistochemical techniques6 .
We hypothesized that differences in vascular density
and blood flow in cervical cancer and precancerous
lesions, as compared with those parameters in normal
women, could be assessed quantitatively by 3D-PDU. The
aim of our study was to identify and quantify any such
changes. This could aid in setting reference values for the
vascular indices studied in both normal and pathological
conditions of the cervix, monitoring changes in these
indices in cervical cancer patients receiving chemotherapy
or radiotherapy and assessing the effect on these indices
of surgical treatments for cervical precancer.

METHODS
Patients
Between November 2008 and June 2010 we prospectively
evaluated 165 women at the 1st Department of Obstetrics
and Gynecology of the University of Athens at Alexandra
Hospital, of whom 71 had cervical cancer, 61 had lowand high-grade intraepithelial cervical lesions and 33

Correspondence to: Dr P. Belitsos, 150 Maikina Street, Zografou, Athens 15771, Greece (e-mail: panbelitsos@yahoo.com)
Accepted: 27 January 2012

Copyright 2012 ISUOG. Published by John Wiley & Sons, Ltd.

ORIGINAL PAPER

3D-PDU for cervical cancer

comprised the control group with no pathology of the
cervix. All women gave informed consent before the
examination, and the study had the approval of the ethical
committee of the hospital.
The women with cervical cancer were all treated in
the gynecologic oncology ward of our hospital. The
recruitment process included all cancer patients willing to
participate in the study. These women were referred with
a diagnosis of cervical cancer. The diagnosis was based
on previously performed colposcopically guided cervical
biopsies. However, previous conization was an exclusion
criterion, since it was considered that this could alter the
vascularization, blood flow and volume of the cervix. The
women were staged according to International Federation
of Gynecology and Obstetrics guidelines7 , had complete
diagnostic work-up (patient history, cervical biopsy,
pelvic magnetic resonance imaging, abdominal computed
tomography, chest X-ray or computed tomography and,
where indicated, cystoscopy/sigmoidoscopy and positron
emission tomography) and were accordingly treated either
by surgery followed by adjuvant therapy (chemoradiation)
or by chemoradiation alone. From the biopsies, or
following surgery if performed, the histological type
and grade were assigned by the pathologist according to
modified Broders criteria for squamous cell carcinomas8,9
and architectural features as well as nuclear features
for adenocarcinomas10,11 . To facilitate statistical analysis
(since there was only one patient with Grade 1 carcinoma)
we divided patients into two subgroups: those with
poorly differentiated tumors (Grade 3) and those with
well and moderately differentiated tumors (Grade 1
and Grade 2, respectively). We also compared patients
according to the presence of operable (Stage IIIA) or
inoperable (Stage IIBIV) tumors. Lymphovascular space
involvement (LVSI) was assessed by hematoxylineosin
staining of the samples.
Women with precancerous lesions were examined at the
colposcopy unit of our hospital and had colposcopically
directed biopsies of their lesions. An inclusion criterion
was normal appearance of the uterus and cervix on twodimensional (2D) ultrasound examination; this was in
order to exclude other types of uterine pathology such
as myomas, endometriosis or adenomyosis that could
possibly alter the vascularization and overall blood flow
of the uterus and especially the cervix. An exclusion
criterion was prior cervical conization. The recruitment
process for the women with precancerous lesions included
all women fulfilling these criteria.
The control group was recruited from women attending
the outpatient department of our hospital for screening
procedures (Pap smears, clinical breast examination,
mammography). Inclusion criteria for the control group
were: a negative cervical cytology examination within
the previous 12 months, no observable pathology of
the vagina or cervix during speculum examination (in
order to exclude pathological conditions such as clinically
obvious inflammation or abnormal hemorrhage, trauma
or other pathology), no pathological findings at bimanual
examination, normal findings of the cervix and uterus on

Copyright 2012 ISUOG. Published by John Wiley & Sons, Ltd.

577

2D ultrasound examination performed before 3D-PDU

and no prior operation on the cervix. Recruitment of the
women comprising the control group was consecutive,
providing that they fulfilled the criteria.

Volume acquisition
This was an observational study in which a single
observer (P.B.) performed all 2D ultrasound and 3DPDU examinations using a Voluson 730 Pro ultrasound
machine (GE Medical Systems, Zipf, Austria) equipped
with a 59-MHz vaginal probe (RIC 59). All women
were examined in the lithotomy position after emptying
the bladder. The probe was inserted slowly into the
vagina, without applying pressure, until acquisition of
a satisfactory image of the cervix. After that, the probe
was retracted until the image blurred and then reinserted
in order to obtain again a satisfactory image of the cervix.
An initial 2D ultrasound examination was performed in
order to visualize the uterus and ovaries. The mid-sagittal
section of the uterus was accordingly obtained and the
depth control was set in such a way that only the cervix
and the lower portion of the uterine body appeared on the
screen. The 3D ultrasound mode and the power Doppler
mode were then activated and the power Doppler window
was placed over the cervix to include it entirely.
We used the default power Doppler settings for all
patients, adjusting only pulse repetition frequency, which
was set at 0.6 kHz. The angle of volume acquisition was
also set to its maximum value (90 ) in order to acquire
the whole cervical volume. The following power Doppler
settings were used: gain, 0 dB (power Doppler gain was
set at the highest possible adjustment in order to increase
the sensitivity of the method, especially in normal cases,
but this adjustment should also be appropriately low to
avoid display of random color speckles); quality, normal;
frequency, medium; wall motion filter, low 1. The 3DPDU submenu parameters used were: line filter, 2; gently
color, on; artifact suppression, off; ensemble, 14; line
density, 8; power Doppler map, 5; balance, 160; flow
resolution, 2; smoothing, 5/5.
Holding the vaginal probe as still as possible, we
acquired 3D volumes of the cervix that were stored on a
hard disk for later analysis. If there were any patient
or bowel movements during volume acquisition that
distorted the image, we repeated the procedure until a
satisfactory volume was obtained.

Analysis of acquired volumes

The stored volumes were further analyzed using the
TM
Virtual Organ Computer-aided AnaLysis
(VOCAL)
program, which is part of the 4D-View software version
9.0 (GE Medical Systems). The same investigator (P.B.)
who had acquired the volumes also performed the
analysis. Using manual mode and working on Plane A
(upper left image of the multiplanar image, corresponding
to the mid-sagittal plane of the uterus and cervix) we
defined the volume of interest using 15 rotation steps (12

Ultrasound Obstet Gynecol 2012; 40: 576581.

Belitsos et al.

578

Figure 1 Estimation of cervical volume using Virtual Organ

Computer-aided AnaLysis in a case of cervical cancer.

steps for each image) (Figure 1). Using a 15 rotational

angle has been shown to be as precise as using an angle of
9 in calculating a volume of interest12 . The contour was
manually traced so as to include the whole cervix (and
not just the tumor in cancer patients).
Once all the contours had been drawn, the VOCAL program automatically calculated the volume of the cervix.
Using the histogram modality we then calculated the 3DPDU indices VI, flow index (FI) and vascularization flow
index (VFI), which are semiquantitative measures of the
vascularization and blood flow within an organ13,14 . VI
represents the percentage of power Doppler data (ratio
of color-coded voxels to all voxels) within the volume
of interest, FI the mean intensity of the power Doppler
signals (from all color-coded voxels) and VFI is a combination of the two indices. These indices are thought to
reflect the number of vessels within the volume of interest
(VI), the intensity of flow at the time of volume acquisition
(FI) and both blood flow and vascularization (VFI)14 16 .
We defined the whole cervix as the region of interest,
even in cases of cervical cancer. This has the advantages
that it is not necessary for the volume analysis to
identify the tumor (which is not always possible17 ), and
comparison with precancerous lesions involves also the
whole cervix, since no measurable tumor is found.
The cervical margins were identified according to the
following criteria: the external cervical os was seen at
the outer end of the echogenic endocervical canal. This
was defined as the outer limit of the cervix. A thin hypoor hyperechogenic line could be identified between the
ectocervical surface and the vaginal wall adjacent to
it. Discriminating the posterior margin of the cervix is
based on the fact that the cervical stroma, although
of comparable echogenicity with the vaginal wall, is
separated from the latter by a thin echogenic line in
the posterior fornix. Defining the anterior margin of the
cervix is easier since it is adjacent to the probe. The margin
most difficult to accurately define was the inner one; the
anatomic landmark used was the angle of the bladder. An
imaginary line was drawn from the angle of the bladder
to the inner end of the endocervical canal, and this defined

Copyright 2012 ISUOG. Published by John Wiley & Sons, Ltd.

the inner margin of the cervix. Furthermore, the cervix has

slightly different echogenicity compared with the corpus
uteri. These criteria can be followed closely when assessing
normal cervices or cervices with precancerous lesions.
When assessing carcinomas, especially those at
advanced stage, the differences in echogenicity between
the tumor and the surrounding tissues, along with the previously described criteria, helped us to define the volume
of interest. In case of isoechoic tumors, compression of
the cervix during the examination could be used to define
the tumor borders, as the tumors were incompressible.
In cases with extensive infiltration of the vagina and the
paracervical tissues, the movement of the probe to the
anterior, posterior or lateral fornices helped us to better define the margins of the cervix. Finally, application
of power Doppler could also enhance our discriminative ability because the neoplastic tissue had increased
vascularity.

Statistical analysis
Quantitative variables were expressed as mean SD or
as median and interquartile range. Qualitative variables
were expressed as absolute and relative frequencies. For
comparison of 3D-PDU indices between two groups,
the MannWhitney U-test was performed, while for
comparison of indices between the three study groups,
the KruskalWallis test was used. Bonferroni correction
was applied in case of multiple comparisons. Spearmans
correlation coefficient was used to explore the association
of 3D-PDU indices with cervical volume. All reported
P-values are two-tailed. Statistical significance was set
at P < 0.05 and analyses were conducted using STATA
statistical software (STATA version 9.0; Stata Corp.,
College Station, TX, USA).

RESULTS
Patient characteristics of the study groups are presented
in Table 1. The mean age of the cervical cancer group
was 54.0 (range, 2684) years. Nearly 40% of them were
operable cases Stage IA to IIA and two-thirds were grade
III. Lymph node status was known only in the women
who were operated on (n = 27), and of these 63% had no
pelvic lymph node metastasis. Concerning the histological
type, a quarter of the cases were adenocarcinomas and
the remaining were squamous cell carcinomas. LVSI was
present in 18 patients (25.4%). Median cervical volume
was 35.7 (range, 10.116282.175) cm3 .
The mean age of the precancerous lesions group was
38.1 (range, 1862) years. In over half of these cases
the histology of the cone showed cervical intraepithelial
neoplasia grade 1. Median cervical volume was 25.3
(range, 9.29850.011) cm3 .
The control group had a mean age of 45.9 (range,
2480) years. Median cervical volume of this group was
19.6 (range, 6.56570.183) cm3 .
VI, FI and VFI values were significantly higher in the
cervical cancer group and precancerous lesions group than

Ultrasound Obstet Gynecol 2012; 40: 576581.

3D-PDU for cervical cancer

579

Table 1 Demographics and clinical characteristics of the study

groups
Characteristic

Value

Cervical cancer group (n = 71)

Age (years)
Menopause
No
Yes
Stage
IAIIA
IIBIV
Grade
I
II
III
Nodes
Negative
Positive
Histology
Adenocarcinoma
Squamous cell carcinoma
LVSI
No
Yes
Precancer lesions group (n = 61)
Age (years)
Menopause
No
Yes
Histology of cone
CIN 1
CIN 2
CIN 3
Control group (n = 33)
Age (years)
Menopause
No
Yes

54.0 12.8

32 (45.1)
39 (54.9)
27 (38.0)
44 (62.0)
1 (1.4)
23 (32.4)
47 (66.2)
17/27 (63.0)
10/27 (37.0)
18 (25.4)
53 (74.6)

according to grade, histology and presence of positive

lymph nodes or LVSI (P > 0.05). Significantly lower
median values of VI (P = 0.037) and VFI (P = 0.028) were
found in postmenopausal women with cervical cancer
than in premenopausal patients with cervical cancer.
VI, FI and VFI were not significantly different when
the two patient subgroups predefined according to stage
were compared (IIIA representing operable tumors and
IIBIV inoperable ones). However, a significant difference
in VI was observed when patients with more advanced
tumors, Stages IIIBIV, were compared with patients with
less advanced disease (Stages IIIIA) (P = 0.045).
A significant positive correlation was found between
cervical volume and VI ( = 0.30, P = 0.010), FI ( = 0.34,
P = 0.003) and VFI ( = 0.32, P = 0.006) in the cervical
cancer group, but not in the precancerous lesions or
control groups.

DISCUSSION

53 (74.6)
18 (25.4)
38.1 10.5

53 (86.9)
8 (13.1)
34 (55.7)
16 (26.2)
11 (18.0)
45.9 15.5

22 (66.7)
11 (33.3)

Data given as mean SD or n (%). CIN, cervical intraepithelial

neoplasia; LVSI, lymphovascular space involvement.

in controls (P < 0.001) (Table 2 and Figure 2). In addition,

significantly higher values of all three indices were found
in the cervical cancer group than in the precancerous
lesions group (P < 0.001).
In the cervical cancer group, further analysis showed
that VI, FI and VFI were not significantly different

There have been only a few studies of cervical cancer

using 3D-PDU and, to the best of our knowledge, no
study that has systematically compared 3D-PDU indices
of the cervix between women with cervical precancerous
lesions and those with cervical cancer.
In one study, vascular indices (median relative color,
median average color and median flow measure) were
studied by 3D color power angiography in 44 patients
and found to be statistically significantly higher in cases
of cervical cancer than in normal cervices17 . In another
study, 97 patients with measurable cervical tumors were
evaluated by 3D-PDU and VI, FI and VFI of patients
with detectable tumors were found to be significantly
higher than those from normal cervices or cervices that
had undergone prior conization18 . This is consistent with
our findings, although the investigators used women
with prior conization in their control group, whereas
we excluded all women with prior conization because
this procedure could alter the vasculature and blood-flow
characteristics in both normal and pathological cervices.
A few studies have focused on changes in these indices
before and during the course of chemotherapy. In one,
30 patients with cervical cancer were compared with 35

Table 2 Comparison of three-dimensional power Doppler indices between the three study groups
Vascularization index (%)
Group
Cervical cancer
Precancer lesion
Control

Flow index

Vascularization flow index

Median
(IQR)

5th 95th
percentile

Median
(IQR)

5th 95th
percentile

Median
(IQR)

5th 95th
percentile

4.036
(1.54913.305)
0.720
(0.1582.975)
0.041
(0.0010.212)

0.06926.748

36.041
(31.55238.753)
31.374
(28.02334.434)
23.232
(17.60825.901)

25.71844.756

1.496
(0.4845.045)
0.234
(0.0751.178)
0.010
(0.0000.057)

0.02311.966

0.00117.700
0.0000.381

22.96944.336
0.00031.959

0.0157.064
0.0000.116

P < 0.001 after Bonferroni correction for cervical cancer vs precancerous lesions, cervical cancer vs control group and precancerous lesions
vs control group for all indices. IQR, interquartile range.

Copyright 2012 ISUOG. Published by John Wiley & Sons, Ltd.

Ultrasound Obstet Gynecol 2012; 40: 576581.

Belitsos et al.

580
(a)

50

VI (%)

40
30
20
10
0
Controls

Precancer lesion

Cervical cancer

Controls

Precancer lesion

Cervical cancer

Controls

Precancer lesion

Cervical cancer

(b) 60

FI

40

20

(c)

30

VFI

20

10

Figure 2 Box plots of vascularization index (VI) (a), flow index (FI)
(b) and vascularization flow index (VFI) (c) obtained from the
cervices of controls, women with precancerous lesions and women
with cervical cancer. Boxes and internal lines represent 25th 75th
percentiles and median, whiskers show the range excluding outliers
and small circles represent outliers.

normal women19 . Significantly higher VI and VFI were

found in the cervical cancer group than in controls, and
patients showed a significant decrease after neoadjuvant
therapy (chemotherapy, radiation or chemoradiation).
The use of VI has also been assessed in the prediction
of the response of cervical carcinoma to neoadjuvant
chemotherapy and it was found that women who did not
respond to neoadjuvant chemotherapy had significantly

Copyright 2012 ISUOG. Published by John Wiley & Sons, Ltd.

higher VI before, during and after chemotherapy when

compared to those who did respond20 . However, in a
recent study21 , 61 patients with locally advanced cervical
carcinoma were studied using 3D-PDU indices in order to
predict response to chemotherapy before surgery, and
it was found that VI, FI and VFI were significantly
higher in clinical responders than non-responders. FI was
also significantly higher in histological responders. On
multiple logistic regression analysis, FI was the only factor
significantly associated with both clinical and histological
responses, and it was therefore concluded that it could be
a possible predictive factor for response to chemotherapy.
In our study we found statistically significant differences
in all 3D-PDU indices among all groups of women studied.
It appears that as the pathological process progresses from
the normal cervix to the premalignant and malignant
status, the vasculature of the cervix changes in parallel to
support this. This change can be quantitatively assessed
using 3D-PDU angiography. The differences in VI, FI and
VFI were significant (P < 0.001) for the transition from
normal cervix to precancerous lesions and to cancer. VI, FI
and VFI were significantly higher in patients with cancer
than in normal controls. This finding is in agreement with
those reported by other investigators17 19,22 .
It is also interesting that the presence of malignancy
or precancerous lesions per se is the determining factor
that alters these indices and not the histology of the
tumor, grade, metastases to pelvic lymph nodes or local
invasion of the lymphovascular space. This has also been
reported in other studies that have shown that tumor
diameter17 , histological type17,22 , LVSI22 and lymph
node metastasis22 do not significantly affect the vascular
indices.
Our data also show that advanced stage tumors
(IIIBIV) have significantly greater values of VI (P < 0.05),
which reflects the presence of increased vasculature in
higher-stage tumors. However, this is not accompanied
by significantly greater blood flow inside the tumor
(FI, P > 0.05). This could be explained by the possible
compression of vessels by bulky or advanced tumors, or by
the presence of necrotic areas inside the tumor. In contrast,
another study, of 56 patients with cervical cancer, showed
significantly higher indices in advanced-stage tumors but
also in poorly differentiated tumors22 . We could not find
any significant correlation between either FI and tumor
stage or vascular indices and histological grade.
In our study we did not perform any evaluation of
interobserver or intraobserver reproducibility. However,
high inter- and intraobserver reproducibility in the
measurement of all three 3D-PDU indices, as well as
volume measurements (intraclass correlation coefficient
> 0.900), have been found in numerous studies14,22 25 .
In our study we recruited women consecutively,
as long as they met the criteria set. Postmenopausal
cancer patients slightly outnumbered premenopausal
ones, a finding consistent with the prevalence of cancer
in the postmenopausal-age subgroup. Premenopausal
women with precancerous lesions outnumbered by
far postmenopausal women. This is expected because

Ultrasound Obstet Gynecol 2012; 40: 576581.

3D-PDU for cervical cancer

precancerous lesions usually affect younger women.
Finally, two-thirds of women comprising the control
group were premenopausal. This is mainly attributable
to the fact that this subgroup participates more actively in
the annual screening programs that were the basis for the
recruitment of healthy women. If we had chosen to have
equal populations of pre- and postmenopausal women
in all age subgroups then this would have inevitably
led to selection bias. Furthermore, it seems that the
alterations in blood circulation caused by the menopause
are minimal compared with the changes caused by the
tumors themselves.
In conclusion, our study assessed women with both
precancerous lesions and cervical cancer and found
significant differences in all 3D-PDU indices studied (VI,
FI and VFI). More studies, with a larger number of
participants, could help in establishing reference values
for 3D-PDU indices in cervical lesions, as it appears that
3D-PDU could be a promising tool for the assessment of
the vascularity of cervical lesions and their response to
treatment.

RE FERENCES
1. Scappaticci FA. Mechanisms and future directions for
angiogenesis-based cancer therapies. J Clin Oncol 2002; 20:
39063927.
2. Folkman J. Seminars in Medicine of the Beth Israel Hospital,
Boston. Clinical applications of research on angiogenesis.
N Engl J Med 1995; 333: 17571763.
3. Hanahan D, Folkman J. Patterns and emerging mechanisms of
the angiogenic switch during tumorigenesis. Cell 1996; 86:
353364.
4. Weidner N, Semple JP, Welch WR, Folkman J. Tumor angiogenesis and metastasis correlation in invasive breast carcinoma. N Engl J Med 1991; 324: 18.
5. Fitzgibbons PL, Page DL, Weaver D, Thor AD, Allred DC,
Clark GM, Ruby SG, OMalley F, Simpson JF, Connolly JL,
Hayes DF, Edge SB, Lichter A, Schnitt SJ. Prognostic factors
in breast cancer. College of American Pathologists Consensus
Statement 1999. Arch Pathol Lab Med 2000; 124: 966978.
6. Park Y, Moon E, Park S, Roh J, Park S. Correlation between
vascularization index (VI) and histologic tumor angiogenesis
in cervical cancer: contrast enhanced three dimensional (3D)
power Doppler vs. immunohistochemistry (IHC). Proc Am Soc
Clin Oncol 2002; 21: 2529.
7. Pecorelli S, Benedet JL, Creasman WT, Shepherd JH. FIGO
staging of gynecologic cancer. 19941997 FIGO Committee on
Gynecologic Oncology. International Federation of Gynecology
and Obstetrics. Int J Gynaecol Obstet 1999; 64: 510.
8. Broders AC. Squamous cell epithelioma of the lip: a study of
five hundred and thirty-seven cases. JAMA 1920; 74: 656664.
9. Wright CW, Ferenczy A, Kurman RJ. Carcinoma and other
tumors of the cervix. In Blausteins Pathology of the Female
Genital Tract (5th edn), Kurman RJ (ed). Springer: New York,
NY, 2002.
10. Kurman JR, Norris HJ, Wilkinson E. Tumors of the cervix,
vagina, and vulva. In Atlas of tumor pathology, Rosai J,
Sobin LH (eds). Armed Forces Institute of Pathology: Washington, DC, 1992.

Copyright 2012 ISUOG. Published by John Wiley & Sons, Ltd.

581
11. Berek JS, Hacker NF, Fu YS, Sokale JR, Leuchter RC, Lagasse
LD. Adenocarcinoma of the uterine cervix: histologic variables
associated with lymph node metastasis and survival. Obstet
Gynecol 1985; 65: 4652.
12. Raine-Fenning N, Campbell B, Collier J, Brincat M, Johnson I.
The reproducibility of endometrial volume acquisition and
measurement with the VOCAL-imaging program. Ultrasound
Obstet Gynecol 2002; 19: 6975.
13. Pairleitner H, Steiner H, Hasenoehrl G, Staudach A. Threedimensional power Doppler sonography: imaging and quantifying blood flow and vascularization. Ultrasound Obstet Gynecol
1999; 14: 139143.
14. Raine-Fenning NJ, Campbell BK, Clewes JS, Kendall NR, Johnson IR. The reliability of virtual organ computer-aided analysis
(VOCAL) for the semiquantification of ovarian, endometrial
and subendometrial perfusion. Ultrasound Obstet Gynecol
2003; 22: 633639.

15. Alcazar
JL. Three-dimensional power Doppler derived vascular
indices: what are we measuring and how are we doing it?
Ultrasound Obstet Gynecol 2008; 32: 485487.
16. Raine-Fenning NJ, Nordin NM, Ramnarine KV, Campbell BK,
Clewes JS, Perkins A, Johnson IR. Determining the relationship
between three-dimensional power Doppler data and true blood
flow characteristics: an in-vitro flow phantom experiment.
Ultrasound Obstet Gynecol 2008; 32: 540550.
17. Testa AC, Ferrandina G, Distefano M, Fruscella E, Mansueto D, Basso D, Salutari V, Scambia G. Color Doppler
velocimetry and three-dimensional color power angiography
of cervical carcinoma. Ultrasound Obstet Gynecol 2004; 24:
445452.
18. Hsu KF, Su JM, Huang SC, Cheng YM, Kang CY, Shen MR,
Chang FM, Chou CY. Three-dimensional power Doppler
imaging of early-stage cervical cancer. Ultrasound Obstet
Gynecol 2004; 24: 664671.
19. Tanaka K, Umesaki N. Impact of three-dimensional (3D)
ultrasonography and power Doppler angiography in the
management of cervical cancer. Eur J Gynaecol Oncol 2010;
31: 1017.
20. Chen CA, Cheng WF, Lee CN, Su YN, Hsieh CY, Hsieh FJ.
Power Doppler vascularity index for predicting the response of
neoadjuvant chemotherapy in cervical carcinoma. Acta Obstet
Gynecol Scand 2004; 83: 591597.
21. Qin J, Cheng X, Chen X, Zhang X, Lu W, Xie X. Value of
three-dimensional power Doppler to predict clinical and
histological response to neoadjuvant chemotherapy in locally
advanced cervical carcinoma. Ultrasound Obstet Gynecol 2012;
39: 226234.

22. Alcazar
JL, Jurado M, Lopez-Garc
a G. Tumor vascularization
in cervical cancer by 3-dimensional power Doppler angiography:
correlation with tumor characteristics. Int J Gynecol Cancer
2010; 20: 393397.
R, Martinez-Monge R, Mazaira J, Alcazar
23. Jurado M, Galvan
JL. Neoangiogenesis in early cervical cancer: correlation
between color Doppler findings and risk factors. A prospective
observational study. World J Surg Oncol 2008; 6: 126.
24. Testa AC, Mansueto D, Lorusso D, Fruscella E, Basso D,
Scambia G, Ferrandina G. Angiographic power 3-dimensional
quantitative analysis in gynecologic solid tumors: feasibility and
reproducibility. J Ultrasound Med 2004; 23: 821828.
25. Qin J, Wang J, Lu W, Li Y, Yao W, Qian Y. Comparison of
reproducibility of manual and sphere contour methods for the
measurement of vascularization in cervical carcinoma using the
virtual organ computer-aided analysis II system. J Ultrasound
Med 2010; 29: 18011810.

Ultrasound Obstet Gynecol 2012; 40: 576581.