Download Diffusion tensor imaging of the normal prostate at 3 Tesla

Eur Radiol (2008) 18: 716–721
DOI 10.1007/s00330-007-0795-7
Bengi Gürses
Neslihan Kabakci
Arzu Kovanlikaya
Zeynep Firat
Ali Bayram
Aziz Müfit Uluð
İlhami Kovanlikaya
Received: 10 April 2007
Revised: 13 September 2007
Accepted: 28 September 2007
Published online: 25 October 2007
# European Society of Radiology 2007
B. Gürses . N. Kabakci .
A. Kovanlikaya . Z. Firat . A. Bayram .
İ. Kovanlikaya
Department of Radiology,
Yeditepe University Hospital,
Kozyatagi, Istanbul, 34752, Turkey
A. M. Uluð
Department of Biomedical
Engineering, Yeditepe University,
School of Engineering,
Kayisdagi, Istanbul, 34755, Turkey
A. M. Uluð
Department of Radiology, Weill
Medical College of Cornell University,
New York, NY, 10021, USA
B. Gürses (*)
Yeditepe Üniversite Hastanesi,
Devlet Yolu Ankara Caddesi,
No: 102-104,
Kozyatagi, 34752, İstanbul, Turkey
e-mail: [email protected]
Tel.: +90-216-5784371
Fax: +90-216-4693796
UROGENI TAL
Diffusion tensor imaging of the normal
prostate at 3 Tesla
Abstract The aim of this study was to
assess the feasibility of diffusion
tensor imaging (DTI) of the prostate
and to determine normative fractional
anisotropy (FA) and apparent diffusion coefficient (ADC) values of
healthy prostate with a 3-Tesla magnetic resonance imaging (MRI)
system. Thirty volunteers with a mean
age of 28 (25–35) years were scanned
with a 3-Tesla MRI (Intera Achieva;
Philips, The Netherlands) system
using a six-channel phased array coil.
Initially, T2-weighted turbo spin-echo
(TSE) axial images of the prostate
were obtained. In two subjects, a
millimetric hypointense signal change
was detected in the peripheral zones
on T2-weighted TSE images. These
two subjects were excluded from the
study. DTI with single-shot echoplanar imaging (ssEPI) was performed
in the remaining 28 subjects. ADC and
FA values were measured using the
manufacturer supplied software by
positioning 9-pixel ROIs on each
zone. Differences between parameters
Introduction
Prostate cancer is one of the most common forms of
malignancy in men, with an increasing incidence. Magnetic
resonance imaging (MRI) has gained clinical acceptance in
the preoperative evaluation of prostatic carcinoma patients
in terms of extracapsular spread, seminal vesicle-neurovascular bundle invasion, lymph node metastases in
patients with biopsy proven prostatic carcinoma [1]. On
the other hand, the accuracy of T2-weighted sequences in
the diagnosis of prostate cancer is limited, due to lack of
sensitivity and specificity [2, 3]. Both prostatitis and
tumoral infiltration have overlapping features and are seen
of the central and peripheral zones
were assessed. Mean ADC value of the
central (1.220±0.271×10−3 mm2/s)
was found to be significantly lower
when compared with the peripheral
gland (1.610±0.347×10−3 mm2/s)
(P<0.01). Mean FA of the central
gland was significantly higher (0.26),
compared with the peripheral gland
(0.16) (P<0.01). This study shows the
feasibility of prostate DTI with a
3-Tesla MR system and the normative
FA and ADC values of peripheral
and central zones of the normal
prostate. The results are compatible
with the microstructural organization
of the gland.
Keywords Prostate . MRI . Diffusion
as decreased signal intensity foci on conventional T2weighted images [1, 4]. In order to increase the accuracy,
new functional techniques, such as dynamic contrast
enhanced MRI and MR spectroscopy, have been added to
the routine protocol in many centers [5].
Recently, a number of studies have shown the feasibility of
diffusion-weighted imaging of the prostate [6–9]. It is well
known that diffusion-weighted imaging provides valuable
data about the microstructural and pathophysiological
aspects of various structures. Fractional anisotropy (FA)
and apparent diffusion coefficient (ADC) values provided
from diffusion tensor imaging (DTI) data reflect the degree of
water diffusion restriction in different tissues. FA is obtained
717
by using at least six tensor applications, which provides
directional data about diffusion anisotropy [8].
DTI has found wide clinical applications, especially in
neuro-and musculoskeletal imaging. In neuroradiology,
diffusion MRI have been mainly used for stroke, multiple
sclerosis and tumor patients. Pathological processes may
cause change in normative FA values and disruption of
fibers in tractography [9, 10].
There are only a few clinical studies about diffusionweighted imaging of the normal prostate and prostatic
carcinoma. The majority of these studies have been carried
out with 1.5-Tesla MRI systems [6–9]. The main limitations of DTI application in the abdomen have been the
presence of motion and poor signal-to-noise ratio values
obtained, especially in low-Tesla systems [10]. With the
advent of 3-Tesla MRI, parallel imaging technology, and
ultrafast single-shot echo-planar sequence, diffusion and
diffusion tensor imaging has been more feasible for
abdominal and pelvic viscera, including the prostate
gland [10].
Three-Tesla MR scanners, providing a twofold increase
in signal-to-noise ratio, provide better spatial and temporal
resolution, and significantly shorter acqusition times [2]. In
3-Tesla systems, a major disadvantage is increased susceptibility artifacts, but the use of parallel imaging
technique, by reducing susceptibility artifacts, has been
shown to provide less distorted diffusion images of the
prostate [10].
The aim of this prospective study is to show the
feasibility of prostate DTI with a 3-Tesla scanner, using the
parallel imaging technique with optimized parameters in
healthy volunteers, to determine the normative FA and
ADC values in peripheral and central zones of the gland,
and to correlate these values with zonal microstructural
organization.
Materials and methods
All MRI examinations were performed with a 3-Tesla
scanner (Intera Achieva; Philips, The Netherlands)
equipped with high-performance gradients of a maximum
strength of 80 mT/m and a slew rate of 200 mT/m/ms,
using six-channel phased array SENSE Torso coil.
Thirty volunteers with a mean age of 28 (range between
25 and 35 years) were scanned in this study. The subjects
did not mention any history of prostatic disease with
established diagnosis. The examination protocol included
axial T2-weighted TSE and DTI with EPI. Initially, T2weighted turbo spin echo images (TR: 2,569 ms; TE:
80 ms; FOV: 430 mm; matrix: 243×512; slice thickness:
2.7 mm; gap: 0.0 mm; flip angle: 90°) were obtained in the
axial plane for all subjects. In two patients, there were
millimetric hypointense foci in the peripheral zone on T2weighted images and these two patients were excluded
from the study and referred to the urology clinic.
Signal intensity of the prostate was normal on T2weighted images in the remaining 28 subjects. DTI was
performed in this group. DTI was obtained using a singleshot echo-planar-imaging technique (TR: 10,000 ms; TE:
53 ms; FOV: 430 mm; matrix: 128×126; SENSE factor: 2;
slice thickness: 2.7 mm; gap: 0.0 mm; flip angle: 90°; voxel
size: 1.68×1.68×2.70 mm). The diffusion gradients were
applied in 32 directions, images were acquired at b values
of 0 and 700 s/mm2.
The optimal b value was determined considering the
ADC value of the prostate tissue. In a diffusion experiment,
the optimal b value for any tissue should be such that
b×ADC is equal to 1. The ADC of interest in our study was
1.220×10−3 mm2/s for the central zone and 1.610×10−3 mm2/s
for the peripheral zone.
The average ADC is 1.415×10−3 mm2/s. The optimal b
value is calculated as: 1/(1.415×10−3)=706.7 s/mm2. We
chose the b_value of 700, which is within 1% of the
theorotical optimal value.
Slice locations, thickness, FOV and gap values were
identical on T2 and DTI sequences in order to maintain
accurate zonal and anatomical correlation. The total scan
duration was 8 min. DTI image acquisition time was 6 min.
Analysis of DTI data
After data acqusition, all images were transferred to the
Workstation for analysis with the special manufacturersupplied software (“PRIDE”). ROIs were drawn on b=0
images, using axial T2-weighted images as anatomical
guide. Each ROI consisted of 9 pixels (1.68×1.68×
2.70 mm each). For detection of FA and ADC values,
four ROIs were used, two on each side. On each side, one
ROI was placed on the central, the other ROI on the
periperal zone.
ROI placements were performed on the level of midgland in each subject where peripheral zone is largest on
axial images. In the young people with no prostatic disease,
the peripheral zone of the prostate is rather thin, that’s why
the ROIs could only be placed in the mid-gland where the
peripheral zone is large enough. Larger ROIs could not be
used because of age-related anatomical reasons.
The average of both values on each sides were obtained
for both zones as the final value. During ROI selection in
the central zone, great care was taken to exclude the
urethra. The neurovascular bundle was avoided while
putting ROIs on the peripheral gland. FA and ADC values
of the selected ROIs were calculated by the manufacturersupplied software system (Fig. 1).
The significance of the difference in ADC and FA values
of the central and peripheral zones were determined using
Student’s t-test. P values <0.05 were regarded as statistically significant.
Tractography was performed using manufacturer supplied software with streamlines fiber tracking algorithm.
718
Fig. 1 Axial T2 TSE (a), b=0 (b), b=700 (c) images, color-coded FA map (d) of a subject. ROI selection is shown on b=0 and FA map
image (d) in the central and peripheral zones
Anisotropy and angular thresholds were selected in the
range of 0.12–0.20 and 30–40°, respectively. The orientation of prostate fibers in different zones were assessed
using tractography images. Blue color represented craniocaudal orientation of diffusion; red color, right-left; green
color, antero-posterior. Changes in intensity of the color
represented different strengths of anisotropy.
Results
In all the subjects, there was statistically significant
difference in ADC and FA values between central and
peripheral zones. The mean ADC value of the central zone
(1.220±0.271×10−3 mm2/s) was significantly lower than the
mean ADC of the peripheral gland (1.610±0.347×10−3 mm2/s)
(P<0.01).
719
Table 1 ADC and FA values for central and peripheral prostate
zones
n=28
Central zone
Peripheral zone
ADC (×10−3 mm2/s)
FA
Mean
SD
Mean
SD
1.220
1.610
0.271
0.347
0.260
0.160
0.060
0.030
The mean FA value of the peripheral gland (0.16) was
significantly lower than the mean FA value of the central
zone (0.26) (P<0.01) (Table 1).
Tractography revealed dark blue dominancy on the
central, and light blue on the peripheral gland. Blue color
was consistent with supero-inferior orientation of fiber
structure within the gland. Moreover, darker color
indicated increased anisotropy in the central zone compatible with higher FA values obtained from statistical analysis
(Fig. 2). On the tractography of the central gland, the
urethra could be seen in the center with no detectable fiber
structure compatible with the free fluid content (Fig. 3).
Discussion
DTI is known to provide valuable data about the
microstructural and pathophysiological aspects of many
tissues which influence diffusion of water molecules. The
main ongoing clinical applications of DTI are in neuro- and
musculoskeletal imaging [11–14]. In neuro-imaging, DTI
is mainly used for cerebral ischemia, multple sclerosis, and
differential diagnosis of tumoral processes.
Fig. 3 Tractography of a subject on the coronal plane (a) showing
the central zone with urethra in the center devoid of fibers. Sagital
plane tractography (b), each zone shown is displayed with a
different color
Fig. 2 Tractography image of the prostate with dark blue
dominancy in the center, and light blue in the periphery
Tissues with a random and less-organized microstructure, such as CSF and gray matter, exhibit isotropic
diffusion characterized with lower FA values. On the other
hand, tissues such as muscle and white matter with
organized microstructural features are anisotropic with
higher FA values [8, 9, 11]. The reason why DTI had been
mostly limited to neuro- and musculoskeletal imaging so
far is probably motion, pulsation artifacts and magnetic
susceptibility from air in the bowel in body imaging.
720
Recently, diffusion imaging has also been applied for
some of the abdominal organs. With the advent of higher
magnetic field systems and phased-array coils with parallel
imaging technology, DTI of some of the intraabominal
organs has been feasible in a clinically acceptable scan time
[9, 12–14].
MRI has been accepted as a clinical tool in the preoperative evaluation of patients with prostatic carcinoma. Due to
the lack of sensitivity and specificity of MRI in the detection
of prostatic carcinoma, clinical MRI has some important
limitations in the differential diagnosis [2, 3]. Both prostatic
carcinoma and prostatitis are seen as hypointense foci on T2weighted images, and the sole criteria for tumoral infiltration
is nodular appearance and invasive behaviour [1, 4, 5]. In the
last few years, attempts are made to increase the sensitivity
and specificity by MR spectroscopy and contrast enhanced
dynamic studies, that are added in some centers to the routine
protocol [6, 15, 16].
DTI is a relatively new technique for prostate imaging.
In the recent literature, there are conflicting FA and ADC
values reported for normal gland and tumoral foci [6, 8, 16,
17]. The aim of the current study was to establish the
normative values for the central zone and peripheral gland,
and to display the tractography in healthy volunteers using
3-Tesla MRI system and parallel imaging technology with
optimized sequence parameters. The optimization included
choosing the optimal b value for prostate and other
sequence parameters such as slice thickness, matrix and
FOV values. All of these values were adapted for the
prostate. There are different studies regarding diffusionweighted imaging of the prostate that have used different b
values. Yet, there is no consensus about the parameters and
we believe that prospective studies with large groups
should be performed to define the optimal parameters in a
3-Tesla system.
Most of the previous literature included diffusionweighted imaging of the prostate using the single parameter of ADC without any direction data [8, 18]. On the
other hand, recently, DTI of the prostate has been carried
out in only a small number of clinical studies. In 2004,
Sinha et al. [9] performed DTI of the prostate in six
volunteers with a 1.5-Tesla system. The b values used in
this study were 0–349.76 mm/s. They have reported that
the mean FA value of the peripheral gland (0.46±0.04) was
slightly higher than the central zone (0.40±0.08) (P<0.01).
They have also mentioned that their results were not in
concordance with the microstructural organization of the
zones. Desouza et al. [18] performed diffusion-weighted
MR imaging in patients with prostate cancer using an
endorectal coil with a 1.5-Tesla system. They reported that
ADC values decreased in tumoral areas. Using an
endorectal coil is an advantage due to improved signalto-noise ratio, but in this study an endorectal coil could not
be used since it was not available at the time of the study.
In 2006, Gibbs et al. [10] performed diffusion and
diffusion tensor imaging in patients with prostatic carci-
noma using 3-Tesla system. They used DTI sequences with
b=0-700 mm/s. Tumoral foci were found to have higher
FA values (0.24±0.05) when compared with the normal
appearing peripheral gland (0.16±0.06).
The FA values they have found in the peripheral gland is
similar to our study. Haker et al. [19] in 2005 performed
DTI of the prostate in eight patients with the suspicion of
cancer using 1.5-Tesla MRI equipment. The normative
mean FA value on benign peripheral zone was reported to
be 0.15±0.06, which is similar to the results of the current
study. The mean FA values of the peripheral gland (0.18±
0.06) obtained in the study by Chen et al. [20] is again
compatible with our results. To our knowledge, there is no
published prospective large series regarding the normal
diffusion tensor parameters of prostate zones.
On the other hand, although there is a limited number of
DTI studies performed on healthy and diseased prostates,
there is no concensus on normative values [6, 10, 16, 21].
Different parameters used in these studies, such as
different signal-to-noise ratio values, imaging protocols,
technical limitations, and b values, may be the reason of
this conflict.
This study establishes the normative values within given
parameters of the central and peripheral zones of prostate.
We used the state-of-the-art MR hardware to optimize the
calculations in a large group of volunteers. The slice
thickness, gap and location were similar for T2-weighted
and DTI sequence for accurate anatomical localization. Our
results are in concordance with the microstructural organization of central and peripheral gland. The prostate gland
is a partly glandular and partly muscular organ enclosed by
a firm fibrous capsule. This capsule is firmly adherent to
the gland and is structurally continuous with the stroma of
the gland. The stroma of the gland is mainly composed of
smooth muscle and fibrous tissue. The central gland with
compact and organized smooth muscle fibers and accompanying glands create more restricton against the diffusion
of water molecules in certain directions with resultant
anisotropy. On the other hand, peripheral gland with rather
loose microstructural organization exhibits less restriction
to diffusion with resultant anisotropy. The results of our
study and the tractography images reflect these facts with
higher FA values in the central zone compared with the
peripheral gland.
Meanwhile, there is an ongoing study on DTI of
prostate cancer in our center. The preliminary data show
increased FA and decreased ADC values at the tumoral
foci. This finding correlates well with the recent literature
[17, 19, 20, 22].
The current study shows the feasibility of prostate DTI
using the state-of-the-art MR hardware and optimized
sequence parameters in a large group of volunteers. We
believe that further studies with high magnetic field systems
and optimized protocols including prostate cancer and prostatitis patients should be undertaken, in order to validate the
role of DTI in the evaluation of prostate pathologies.
721
References
1. Akin O, Hricak H (2007) Imaging of
prostate cancer. Radiol Clin North Am
45:207–222
2. Rouviere O, Hartman RP, Lyonnet D
(2006) Prostate MR imaging at highfield strength: evolution or revolution?
Eur Radiol 16:276–284
3. Ikonen S, Kärkkäinen P, Kivisaari L, Salo
JO, Taari K, Vehmas T, Tervahartiala P,
Rannikko S (2001) Endorectal magnetic
resonance imaging of prostatic cancer:
comparison between fat-suppressed
T2-weighted fast spin echo and threedimensional dual-echo, steady-state
sequences. Eur Radiol 11:236–241
4. Ikonen S, Kivisaari L, Tervahartiala P,
Vehmas T, Taari K, Rannikko S
(2001) Prostatic MR imaging.
Accuracy in differentiating cancer
from other prostatic disorders. Acta
Radiol 42:348–354
5. Hricak H (2006) New horizons in
genitourinary oncologic imaging.
Abdom Imaging 31:182–187
6. Shimofusa R, Fujimoto H, Akamata H,
Motoori K, Yamamoto S, Ueda T, Ito H
(2005) Diffusion-weighted imaging of
prostate cancer. J Comput Assist Tomogr
29:149–153
7. Reinsberg SA, Payne GS, Riches SF,
Ashley S, Brewster JM, Morgan VA,
deSouza NM (2007) Combined
use of diffusion-weighted MRI
and 1H MR spectroscopy to
increase accuracy in prostate cancer
detection. AJR Am J Roentgenol
188:91–98
8. Issa B (2002) In vivo measurement of
the apparent diffusion coefficient in
normal and malignant prostatic tissues
using echo-planar imaging. J Magn
Reson Imaging 16:196–200
9. Sinha S, Sinha U (2004) In vivo
diffusion tensor imaging
of the human prostate. Magn
Reson Med 52:530–537
10. Gibbs P, Pickles MD, Turnbull LW
(2006) Diffusion imaging of the
prostate at 3.0 Tesla. Invest Radiol
41:185–188
11. Mukherjee P (2005) Diffusion tensor
imaging and fiber tractography in acute
stroke. Neuroimaging Clin N Am 15:655–
665, xii
12. Sinha S, Sinha U, Edgerton VR (2006)
In vivo diffusion tensor imaging of the
human calf muscle. J Magn Reson
Imaging 24:182–190
13. Cova M, Squillaci E, Stacul F,
Manenti G, Gava S, Simonetti G, PozziMucelli R (2004) Diffusion-weighted
MRI in the evaluation of renal lesions:
preliminary results. Br J Radiol 77:
851–857
14. Thoeny HC, De Keyzer F (2007)
Extracranial applications of diffusionweighted magnetic resonance imaging.
Eur Radiol 17:1385–1393
15. Padhani AR, Gapinski CJ, Macvicar
DA, Parker GJ, Suckling J, Revell PB,
Leach MO, Dearnaley DP, Husband JE
(2000) Dynamic contrast enhanced
MRI of prostate cancer: correlation
with morphology and tumour stage,
histological grade and PSA. Clin
Radiol 55:99–109
16. Sato C, Naganawa S, Nakamura T,
Kumada H, Miura S, Takizawa O, Ishigaki
T (2005) Differentiation of noncancerous
tissue and cancer lesions by apparent
diffusion coefficient values in transition
and peripheral zones of the prostate. J
Magn Reson Imaging 21:258–262
17. Pickles MD, Gibbs P, Sreenivas M,
Turnbull LW (2006) Diffusionweighted imaging of normal and malignant prostate tissue at 3.0T. J Magn
Reson Imaging 23:130–134
18. Desouza NM, Reinsberg SA, Scurr ED,
Brewster JM, Payne GS (2007) Magnetic resonance imaging in prostate
cancer: the value of apparent diffusion
coefficients for identifying malignant
nodules. Br J Radiol 80:90–95
19. Haker SJ, Barnes AS, Maier SE et al
(2005) Diffusion tensor imaging for
prostate cancer detection. In: Proc 13th
Annual Meeting ISMRM, Miami
20. Chen AP, Xu D, Henry R et al (2003)
Diffusion tensor imaging of the prostate
following therapy. In: Proc 11th Annual
Meeting ISMRM, Toronto
21. Manenti G, Squillaci E, Di Roma M,
Carlani M, Mancino S, Simonetti G
(2006) In vivo measurement of the
apparent diffusion coefficient in normal
and malignant prostatic tissue using
thin-slice echo-planar imaging. Radiol
Med (Torino) 111:1124–1133
22. Miao H, Fukatsu H, Ishigaki T (2007)
Prostate cancer detection with 3-T
MRI: comparison of diffusionweighted and T2-weighted imaging.
Eur J Radiol 61:297–302