Springer Science+Business Media, LLC 2012
Abdominal Published online: 15 April 2012
Imaging
Pelvic inflammatory disease: evaluation of
diagnostic accuracy with conventional MR with
added diffusion-weighted imaging
Wenhua Li,1 Yuzhen Zhang,1 Yanfen Cui,1 Ping Zhang,2 Xiangru Wu3
1
Department of Radiology, Xinhua Hospital Affiliated to Shanghai JiaoTong University School of Medicine, 1665 KongJiang
Road, Shanghai 200092, China
2
Department of Obstetrics and Gynecology, Xinhua Hospital Affiliated to Shanghai JiaoTong University School of Medicine, 1665
KongJiang Road, Shanghai 200092, China
3
Department of Pathology, Xinhua Hospital Affiliated to Shanghai JiaoTong University School of Medicine, 1665 KongJiang
Road, Shanghai 200092, China
Abstract
Purpose: To determine the incremental value of magnetic
resonance (MR) diffusion-weighted (DW) imaging for
the diagnosis of pelvic inflammatory diseases (PID).
Materials and methods: We added DW sequences to
conventional MR imaging in 187 patients with clinically
suspected PID. The imaging findings included shape,
signal intensity on T1-weighted, T2-weighted, and DW
imaging, shade in the peripheral lesions, free pelvic fluid,
and lymphadenopathy.
Results: Laparoscopic and pathological findings con-
firmed the diagnosis in all patients. Conventional MR
findings were consistent with a diagnosis of PID in 90.7%
(117/129) and of non-PID in 93.3% (28/30) of the 159
patients. The sensitivity, specificity, positive predictive
value, negative predictive value, and accuracy of con-
ventional MR imaging findings vs. the addition of DW
imaging to conventional MR protocols for predicting
PID were 90.7%, 93.3%, 98.3%, 70.0%, and 91.2% and
98.4%, 93.3%, 98.4%, 93.3%, and 97.5%, respectively.
Conclusion: The addition of DW sequences to conven-
tional MR imaging can improve the accuracy of
diagnosis in PID.
Key words: OvaryPelvic inammatory disease
Pelvic abscessDiffusion-weighted imaging
Correspondence to: Wenhua Li; email: wenhualimyj@sohu.com
Abdom Imaging (2013) 38:193200
DOI: 10.1007/s00261-012-9896-0
Magnetic resonance (MR) imaging of the pelvic cavity
has gained increasing attention as a powerful diagnostic
tool for the assessment of pelvic diseases [14]. Ultraso-
nography (US) is the first-line noninvasive technique
for detecting pelvic inflammatory disease (PID), but it is
less accurate for the characterization of complex pelvic
lesions, even when combined with color Doppler
Imaging. MR imaging has been shown to be more spe-
cific and accurate than US assessment for PID, with a
sensitivity of 95%, a specificity of 89%, and an accuracy
of 93%, compared with a sensitivity of 81%, a specificity
of 78%, and an accuracy of 80% for US [5].
Little data is available on the contribution of MR
diffusion-weighted (DW) imaging to the accuracy of
diagnosis in PID. DW imaging and the calculated
apparent diffusion coefcient (ADC) were originally
used to demonstrate early ischemic changes in brain tis-
sue and have also been applied for characterization of
brain, abdominal, and pelvic tumors [211].
The purpose of our study, therefore, was to pro-
spectively evaluate the potential of the addition of DW
imaging to a conventional MR imaging protocol for
accurate characterization of PID, using laparoscopy and
surgery as the reference standard.
Materials and methods
Patients
The study was approved by our institutional ethics com-
mittee and written informed consent was obtained from all
patients. Between January 2007 and October 2011, a
DW imaging sequence was added to conventional MR
imaging for each woman with clinically suspected PID.
194
The consecutive population included 187 women, who
satised the following inclusion criteria: (a) a history of
acute pelvic or lower abdominal pain of <3 weeks duration
with or without fever; (b) the presence of lower abdominal
tenderness; (c) bilateral adnexal tenderness and cervical
motion tenderness; and (d) elevation of C-reactive protein
(>10 mg/L). The MR imaging database was reviewed,
and subjects who subsequently underwent laparoscopy
and surgery were selected. Patients with endometriomas
(n = 17), mature teratomas (n = 6), free pelvic fluid only
(n = 2), and DW imaging artifacts (n = 3) were excluded
to limit selection bias. For endometriomas, mature terato-
mas, and free pelvic fluid, conventional MR imaging can
provide an appropriate diagnosis.
The nal cohort included 159 patients (premenopau-
sal 103 and postmenopausal 56; mean age, 43.5 years;
age range, 1872 years). Of the 159 patients, 81.1% (129/
159) had laparoscopic and pathologically proven PID.
Thirty subjects (18.9%) were shown not to have PID and
were randomly selected to serve as control subjects.
MR image acquisition
In all patients, imaging was performed with a 1.5-T clin-
ical MR imaging system (Twinspeed, GE Medical Sys-
tems, Milwaukee, WI, USA), using a pelvic phased-array
coil. All sequences were acquired with saturation bands
placed anteriorly and posteriorly to eliminate the high
signal from subcutaneous fat. The imaging protocol was
comprised of axial non-contrast T1-weighted (TR/TE
range, 400600/1014 ms) and axial T2-weighted (TR/TE
range, 40006000/100120 ms) imaging performed with a
chemical shift-selective fat saturation pulse using the fol-
lowing parameters: slice thickness, 6 mm; gap, 1 mm; eld
of view (FOV), 3242 cm; matrix, 256 9 256; and exci-
tation, 2. Sagittal T1-weighted and T2-weighted (TR/TE
range, 30006000/100110 ms) fast spin-echo imaging
without chemical shift-selective fat saturation pulse was
also performed, as well as post-contrast enhanced-axial
and sagittal T1-weighted imaging using the same param-
eters described above except that a slice thickness of
57 mm was used. DW MRI was acquired in the axial
plane before administration of contrast medium by using
a single-shot echo-planar imaging sequence (TR/TE
effective range, 800010,000/70100; slice thickness,
6 mm; gap, 1 mm; FOV, 3242 cm; matrix, 128 9 128;
excitation, 2). B values of 0 and 1000 s/mm2 were also
applied in three orthogonal (Z, Y, and X) directions.
MR image analysis
Conventional MR data was analyzed on an Advantage
Windows workstation 4.2 (GE Healthcare, Milwaukee,
WI) by two radiologists (who had 11 and 12 years of
experience in pelvic MR imaging, respectively), without
knowledge of clinical presentation and laparoscopic and
W. Li et al.: Pelvic inflammatory disease
pathological ndings, in two steps. In the rst step, the
radiologists reviewed the conventional MR images. In
the second step, they reviewed the conventional MR and
DW images. For both steps, the observers characterized
lesions as PID or non-PID. Discrepancies in interpreta-
tion were resolved by consensus.
Conventional MR ndings, including signal intensity
characteristics and morphological appearance of lesions,
were evaluated. The normal outer myometrium was used
as a signal intensity (grade 3: low, iso- or intermediate,
and high) reference on T1-weighted, T2-weighted and
DW images. The diagnosis of PID can be made in the
presence of the following ndings: a uid-lled tube
(FFT) in acute salpingitis was diagnosed when a sausage-
like or tortuous, slightly dilated FFT with wall thickening
and enhancement, with the tubal content demonstrating
low signal intensity on T1-weighted images and high
signal intensity on T2-weighted images. A pyosalpinx was
diagnosed when a dilated tubular structure with thick
enhancing walls was identied with lack of internal
enhancement, and with signal intensity characteristics
similar to those of the FFT. A hydrosalpinx has a similar
appearance to the pyosalpinx, with C-, S-, or cystic
shading with or without plica, but with thin walls. A
tubo-ovarian or ovarian abscess was dened as an
ill-dened adnexal mass with thick regular or irregular
enhanced walls containing uid. It usually had low signal
intensity on T1-weighted images, high signal intensity on
T2-weighted and on DW images, but the signal intensity
can vary from low signal intensity on T1-weighted images
to intermediate or high signal intensity on T2-weighted
images. An adnexal mass was dened as one that could
not be condently categorized into either an abscess or a
site of origin. A diagnosis of hematosalpinx or fallopian
tubal torsion was made on the basis of a blood-containing
dilated tube with high signal intensity on T1-weighted
images and high or low signal intensity on T2-weighted
images with possible shading or layering [58].
All ADC measurements of tubal uid and high signal
intensity areas of adnexal masses on DW images were
made by one radiologist, at an Advantage Windows
workstation 4.2 (GE Healthcare, Milwaukee, WI), by
placing freehand regions of interest (ROIs) on DW
images. ROIs were placed carefully within the center of the
tubal uid and the high signal intensity areas. To minimize
variability, the largest possible ROIs, which varied from 18
to 120 mm2, were manually placed in the target area. When
the lesion exhibited multiple high signal intensity areas,
ROIs were drawn within the targeted components and the
mean ADC value was used for the results.
Statistical analysis
Laparoscopic and pathological ndings were used as the
reference standard. Mean ADC values were compared
using a one way analysis of variance and an unpaired t test
W. Li et al.: Pelvic inflammatory disease 195

with a P < 0.05 considered statistically significant. Table 2. Performance of the two blinded observers for the detection of
PID
Quadratic K coefficients were calculated to assess the
interobserver agreement between conventional MR imag- Parameter cMR imaging alone cMR and DW imaging
ing and the combination of conventional MR and DW No. of TP findings 117 127
imaging readers with regard to lesion characterization. A k No. of TN findings 28 28
value of 0.610.80 was indicative of substantial agreement No. of FP findings 2 2
No. of FN findings 12 2
and a value of 0.80 reflected almost perfect agreement. Sensitivity (%) 90.7 98.4
Specificity (%) 93.3 93.3
Results PPV (%) 98.3 98.4
NPV (%) 70.0 93.3
Clinical demographics Accuracy (%) 91.2 97.5

One hundred and fty-nine patients met our inclusion
criteria. Laparoscopy and surgery were performed within
2 weeks of MR imaging, and histological ndings are
shown in Table 1.
Comparison of conventional MR
and conventional MR with added DW imaging
The results in all 159 patients with suspected PID
revealed considerable variability in MR imaging ndings.
However, there was almost complete agreement between
the radiologists in lesion characterization between con-
ventional MR imaging and the combination of conven-
tional MR with DW imaging (k = 0.893). Conventional
MR findings were consistent with a diagnosis of PID in
90.7% (117/129) and of non-PID in 93.3% (28/30) of the
159 patients. The sensitivity, specificity, positive predic-
tive value (PPV), negative predictive value (NPV), and
accuracy of conventional MR imaging findings for pre-
dicting the nature of suspected PID were 90.7%, 93.3%,
98.3%, 70.0%, and 91.2%, respectively. When DW
imaging was added to conventional MR protocols, the
sensitivity, specificity, PPV, NPV, and accuracy of MR
imaging findings for predicting PID were 98.4%, 93.3%,
98.4%, 93.3%, and 97.5%, respectively. Thus, the com-
bined imaging strategy may increase the accuracy, sen-
sitivity, and NPV of MR imaging for distinguishing PID
from non-PID (Table 2).
TP true positive, TN true negative, FP false positive, FN false negative,
cMR conventional MR, PPV positive predictive value, NPV negative
predictive value
The signal characteristics of an abscess can demon-
strate considerable variability in MR imaging ndings on
T1-weighted and T2-weighted images. Among 40 tubo-
ovarian and ovarian abscess patients, 52.5% (21/40) had
low signal intensity on T1-weighted images and high
signal intensity on T2-weighted images. Of these patients,
37.5% (15/40) had intermediate or high signal intensity
on T1-weighted images and high signal intensity on
T2-weighted images, and 35.0% (14/40) had intermediate
or high signal intensity on T1-weighted images and inter-
mediate signal intensity (pseudosolid) on T2-weighted
images. Associated pelvic abscess formation was observed
in 27.5% (11/40). Conventional MR imaging also revealed
thickening of the uterosacral ligaments in 68.2% (88/129),
periovarian stranding and enhancement of the peritoneum
in 29.5% (38/129), and free pelvic uid in 100% (129/129)
of patients with PID. Pelvic lymphadenopathy was
observed in 35.7% (5/14) of those with fallopian tube and
ovarian cancers.
Among 12 misclassied patients using conventional
MR imaging, 1 pyosalpinx was misclassied as acute
salpingitis with a FFT correctly reclassied by taking
into account the high signal intensity of luminal uid
on DW images with a low ADC value (Fig. 1). Two

Table 1. Summary of patient characteristics (n = 159)

Characteristic Patient no. Distribution Mean ADC value

Patients with PID (n = 129)

FFT with salpingitis 35 Unilateral 21; bilateral 14 2.49 0.24a
Hydrosalpinx 41 Unilateral 9; bilateral 32 2.57 0.21a
Pyosalpinx 13 Unilateral 5; bilateral 8 0.78 0.26b
TOA 19 Unilateral 15; bilateral 6 0.73 0.29b
Ovarian abscess 21 Unilateral 11; bilateral 10 0.74 0.25b
Patients with non-PID (n = 30)
Hematosalpinx 7 Unilateral 7 0.79 0.11b
Tubal torsion 6 Unilateral 6 0.91 0.17b
Tube ligation 2 Bilateral 2 2.57 0.08a
FTC 2 Unilateral 2 2.51 0.06a
Ovarian cancer 12 Unilateral 12 2.41 0.22a

Numbers are expressed as mean SD, or range of ADC values. There is a significant difference between a and b group (P < 0.05). ADC is unit
910-3 mm2/s. The ADC value in the ovarian cancer group is that of the cystic component
PID pelvic inflammatory disease, FFT fluid-filled tube, TOA tubo-ovarian abscess, FTC fallopian tube carcinoma
196
Fig. 1. A 46-year-old woman with a right pyosalpinx. A Axial
T1-weighted image shows a sausage-like tubal structure with
low signal intensity (arrow). B Axial T2-weighted image
reveals that the dilated tube has high signal intensity. C Axial
contrast-enhanced MR image shows the tube with wall
fallopian tube ligations with hydrosalpinx were misclas-
sified as chronic salpingitis with hydrosalpinx, and were
not correctly reclassified by the addition of DW imaging
to conventional MR protocols. Both of these patients
had similar low signal intensity on T1-weighted and high
signal intensity on T2-weighted images, and low signal
intensity on DW images as PID patients with hydrosal-
pinx and high ADC values. Among nine conventional
MR imaging findings misclassified as adnexal tumors, all
areas detected were low and intermediate signal intensity
on T1-weighted images, with heterogeneous (intermedi-
ate and high mixed) signal intensity on T2-weighted
images within the masses. The area exhibiting interme-
diate signal intensity on T1-weighted images and
T2-weighted images were regarded as soft-tissue
(pseudosolid) within the mass. These 9 patients were
W. Li et al.: Pelvic inflammatory disease
thickening and enhancement similar to that in acute salpigitis.
D Axial DW images shows a tubal content of high signal
intensity, representing pus with a low ADC value
(ADC = 0.763 9 10-3 mm2/s).
correctly reclassified by taking into account the fact that
all intermediate and high signal intensity areas on
T2-weighted images were high signal intensity on DW
imaging with low ADC values (Fig. 2). Of these 40 tubo-
ovarian and ovarian abscess patients, 72 abscesses were
revealed by DW imaging, but only 46 abscesses were
confidently found as the missed abscesses were of inter-
mediate signal intensity (misclassified as solid components
of ovarian tumors) on T2-weighted images (Fig. 3).
ADC analysis
The mean ADC values of contained uid in PID and
non-PID patients are presented in Table 1. There are
many overlaps between the ADC values for both groups.
The mean ADC values did not differ significantly
W. Li et al.: Pelvic inflammatory disease
Fig. 2. A 49-year-old woman with a right TOA. A Axial
T1-weighted image shows an iso- and mildly high mixed sig-
nal intensity mass in the right side of the pelvic cavity. B Axial
T2-weighted image reveals that the mass has heterogeneous
high signal intensity. C Axial contrast-enhanced MR image
shows complex cystic and solid mass with thick, irregular
between the FFT group, the hydrosalpinx group, the
fallopian tube ligation with hydrosalpinx group, the
fallopian tube carcinoma (FTC) group, and the cystic
component in the ovarian cancer group (P > 0.05). The
mean ADC values of contained fluid did not differ sig-
nificantly between the pyosalpinx, tubo-ovarian abscess
(TOA), ovarian abscess, hematosalpinx, and tubal tor-
sion groups (P > 0.05), but there was a statistically
significant difference in mean ADC values between
groups a and b (P < 0.05) (Table 1).
Discussion
Our results demonstrate that dilated fallopian tubes can
be identied on MR imaging, and that DW imaging in
197
enhancing walls and septa. D Axial DW images show an
S-like structure and high signal intensity, and an ovoid high
signal intensity area within a complex mass, findings that are
compatible with TOA (circle 1 ADC = 0.636 9 10-3 mm2/s,
circle 2 ADC = 1.28 9 10-3 mm2/s, circle 3 ADC = 1.35 9
10-3 mm2/s in the outer myometrium).
combination with conventional MR imaging is not only
sensitive and specic for establishing the diagnosis of
PID, but also for distinguishing PID from other patho-
logic processes. With the use of ADC measurements and
signal characteristics, MR imaging can lead to the cor-
rect assessment of PID with a higher degree of accuracy.
DW imaging has already been used to improve con-
ventional MR imaging by allowing the identication of
specic criteria for diagnosis according to tissue speci-
cities of various brain, breast, prostate, and abdominal
tumors [1218]. However, until now no such data has
been available for PID patients. To the best of our
knowledge, this is the first study to evaluate the role of
DW imaging in conjunction with signal characteristics to
further characterize PID. Acute tubo-ovarian or ovarian
198
Fig. 3. A 54-year-old woman with right ovarian abscess. A
Axial T1-weighted image shows a low heterogeneous signal
intensity mass in the right side of the pelvic cavity. B Axial T2-
weighted image reveals that the mass has intermediate signal
intensity and a small cystic component (high signal intensity
area). C Axial contrast-enhanced MR image shows a complex
cystic and solid mass with thick enhancing walls and a low
abscesses may produce very complex masses with wall
thickening and pseudosolid areas. The pseudosolid areas
can demonstrate low, intermediate, or high signal
intensity on T1-weighted images and low or intermediate
signal intensity on T2-weighted images, and are similar
to those of soft-tissue or mucinous components of ad-
nexal masses. The signal characteristics of pseudosolid
areas may be related to the presence of high inflamma-
tory cellularity, a matrix of proteins, cellular debris, and
bacteria in high viscosity purulent fluids, as well as large
molecules playing a key role in restricting the diffusion
of protons in pus. These can demonstrate marked
enhancement, thereby mimicking malignancy. The clini-
cal features of acute pelvic infection and malignancy
usually differ, though some chronic infections such as
TOA may exactly mimic malignancy. In our patients, all
abscess cavities initially had high signal intensity on DW
W. Li et al.: Pelvic inflammatory disease
signal intensity area that mimics a soft-tissue (pseudosolid)
signal (long arrow). D Axial DW images show that the
pseudosolid and high signal intensity area on the T2-weighted
image are all high signal intensity, compatible with purulent
cavities with low ADC values (circle 1 ADC = 0.555 9 10-3
mm2/s, circle 2 ADC = 0.519 9 10-3 mm2/s).
images with low ADC values. These results confirm that
the addition of DW imaging enabled us to increase the
diagnostic confidence of conventional MR imaging in up
to 22.5% (9/40) of tubo-ovarian and ovarian abscesses.
This positive result is mainly due to the fact that the
addition of DW imaging to conventional MR imaging
can confidently differentiate pseudosolids (pus) from
other conditions based on signal characteristics. When a
high signal intensity area is depicted on DW imaging
with low ADC values, our results demonstrate that high
or intermediate signal intensity on T2-weighted images of
this area without enhancement is the best criteria for
predicting the presence of an abscess. These findings are
similar to previous reports that demonstrated the rela-
tionship between high signal intensity with low ADC
values in the central part of a brain abscess and the
presence of pus [12, 13].
W. Li et al.: Pelvic inflammatory disease
Our results also showed that the signal intensity
of fallopian tube uid is associated with pathological
conditions, and the combination of T1-weighted,
T2-weighted, DW, as well as morphology, allowed us to
further characterize lesions that were clinically suspected
to be PID. In acute salpingitis, the fallopian tubes are
often revealed as sausage-like, slightly dilated FFTs with
wall thickening and enhancement. The tubal content has
low signal intensity on T1-weighted, DW, and high signal
intensity on T2-weighted images. On MR images, both
hydrosalpinx and pyosalpinx appear as dilated, uid-l-
led, and tubal structures. With pyosalpinx, the wall of a
dilated fallopian tube may be thickened, and it has var-
iable signal intensity on T1-weighted images and heter-
ogeneous signal intensity on T2-weighted images. In
general, a pyosalpinx cannot be reliably differentiated
from a hydrosalpinx on conventional MR images. The
combination of T1-weighted, T2-weighted, and DW
images allowed us to further characterize and differen-
tiate both entities, with hydrosalpinx demonstrating low
signal intensity and pyosalpinx high signal intensity on
DW images. This is explained by the fact that the vis-
cosity or protein concentrations of the uid contained in
a hydrosalpinx and that of pus in a pyosalpinx are
essentially different.
In our study, the signal intensity of the tubal uid
correlated with the presence of hemorrhage or blood. A
fallopian tubal torsion and hematosalpinx may have a
similar appearance, which is high signal intensity on
T1-weighted, T2-weighted, and DW images with mod-
erate to marked T2 shortening in the form of shading.
DW imaging alone is not useful for differentiating tubal
torsion and hematosalpinx from TOA and ovarian
abscesses.
Our study also demonstrated that the solid compo-
nents of ovarian tumors and FTC are generally high
signal intensity on DW images with low ADC values. It
is difcult to separate solid parts of ovarian tumors and
FTC from abscesses only depending on signal charac-
teristics on DW images and ADC values, but the solid
parts of ovarian tumors and FTC reveal enhancement is
different the pseudosolid component of abscess without
enhancement after administration of contrast medium.
In addition, necrotic and cystic components within
ovarian cancers usually contain cellular debris, serous or
mucinous uid, and fewer inammatory cells. Water
molecules, therefore, have a greater freedom of motion.
The necrotic and cystic components often show mildly
increased diffusion with low signal intensity on DW
images and high ADC values.
Some limitations of this study need to be highlighted.
First, all patients were highly selected and the sample
size was relatively small. Therefore, our results will need
to be conrmed in further clinical studies that involve
larger numbers of patients. Second, this study included
199
non-PID patients. Finally, the variable concentrations
of inammatory cells and bacteria with different path-
ogenic organisms, the age of an abscess and blood
parameters might inuence the uid viscosity, resulting
in variations of signal intensity and ADC values. As a
result, conclusions drawn from the results observed in
this cohort might not be generally applicable. A con-
tinued evaluation of the clinical utility and indications of
combined DW and MRI, therefore, still needs to be
investigated, especially as there are few reports of the
use of ADC values for the differential diagnosis of pa-
tients with PID.
In conclusion, the ndings in this study demonstrate
that DW imaging in combination with conventional MR
imaging is a useful tool for improving the characteriza-
tion of clinically suspected PID. With the use of ADC
measurements and signal characteristics, MR imaging
can lead to a more accurate assessment of adnexal
lesions.
References
1. Medeiros LR, Freitas LB, Rosa DD, et al. (2011) Accuracy of
magnetic resonance imaging in ovarian tumor: a systematic quan-
titative review. Am J Obstet Gynecol 204:67.e167.e10
2. Qayyum A (2009) Diffusion-weighted imaging in the abdomen and
pelvis: concepts and applications. Radiographics 29:17971810
3. Spencer JA, Ghattamaneni S (2010) MR imaging of sonographi-
cally indeterminate adnexal mass. Radiology 256:677694
4. Li WH, Chu CT, Cui YF, Zhang P, Zhu MJ (2011) Diffusion-
weighted MRI: a useful technique to discriminate benign versus
malignant ovarian surface epithelial tumors with solid and cystic
components. Abdominal Imaging . doi:10.1007/s00261-011-9814-x
5. Tukeva TA, Aronen HJ, Karjalainen PT, et al. (1999) MR imaging
in pelvic inflammatory disease: comparison with laparoscopy and
US. Radiology 210:209216
6. Rezvani M, Shaaban AM (2011) Fallopian tube disease in the
nonpregnant patient. Radiographics 31:527548
7. Kim MY, Rha SE, Oh SN, et al. (2009) MR imaging findings of
hydrosalpinx: a comprehensive review. Radiographics 29:495507
8. Outwater EK, Siegelman ES, Kilger AM, Dunton CJ, Talerman A
(1998) Dilated fallopian tubes: MR imaging characteristics. Radi-
ology 208:463469
9. Thomassin-Naggara I, Darai E, Cuenod CA, et al. (2009) Contri-
bution of diffusion-weighted MR imaging for predicting benignity
of complex adnexal masses. Eur Radiol 19:15441552
10. Umemoto M, Shiota M, Shimono T, et al. (2006) Preoperative
diagnosis of ovarian tumors, focusing on the solid area based on
diagnostic imaging. J Obstet Gynaecol Res 32:195201
11. Thomassin-Naggara I, Toussaint I, Perrot N, et al. (2011) Char-
acterization of complex adnexal masses: value of adding perfusion-
and diffusion-weighted MR imaging to conventional MR imaging.
Radiology 258:793803
12. Luthra G, Parihar A, Nath K, et al. (2007) Comparative evaluation
of fungal, tubercular, and pyogenic brain abscesses with conven-
tional and diffusion MR imaging and proton MR spectroscopy.
AJNR Am J Neuroradiol 28:13321338
13. Cartes-Zumelzu FW, Stavrou I, Castillo M, et al. (2004) Diffusion-
weighted imaging in the assessment of brain abscesses therapy.
AJNR Am J Neuroradiol 25:13101317
14. Erturk SM, Ichikawa T, Sano K, et al. (2008) Diffusion-weighted
magnetic resonance imaging for characterization of focal liver
masses: impact of parallel imaging (SENSE) and b value. J Comput
Assist Tomogr 32:865871
200
15. Nakai G, Matsuki M, Inada Y, et al. (2008) Detection and evalu-
ation of pelvic lymph nodes in patients with gynecologic malig-
nancies using body diffusion-weighted magnetic resonance imaging.
J Comput Assist Tomogr 32:764768
16. Fujii S, Katite S, Nishihara K, et al. (2008) Diagnostic accuracy of
diffusion-weighted imaging in differentiating benign from malig-
nant ovarian lesions. J Magn Reson Imaging 28:11491156
W. Li et al.: Pelvic inflammatory disease
17. Khouli RH, Jacobs MA, Mezban SD, et al. (2010) Diffusion-
weighted imaging improves the diagnostic accuracy of conventional
3.0-T breast MR imaging. Radiology 256:6473
18. Oto A, Kayhan A, Jiang YL, et al. (2010) Prostate cancer: differ-
entiation of central gland cancer from benign prostatic hyperplasia
by using diffusion-weighted and dynamic contrast-enhanced MR
imaging. Radiology 257:715723