AJR 2004; 183:24-28
© American Roentgen Ray Society
Small-Field-of-View MRI of the Knee and Ankle
Gregory Ernest Antonio1,
James Francis Griffith1 and
David K. W. Yeung2
1 Department of Diagnostic Radiology and Organ Imaging, The Chinese University
of Hong Kong, Prince of Wales Hospital, Shatin, NT, Hong Kong.
2 Department of Clinical Oncology, The Chinese University of Hong Kong, Prince
of Wales Hospital, Shatin, NT, Hong Kong.
Received October 7, 2003;
accepted after revision December 23, 2003.
Address correspondence to G. E. Antonio.
Introduction
Imaging of joints has developed appreciably since the introduction of MRI,
and image quality continues to improve with new technical developments.
Improved visibility can be achieved by increasing the contrast or spatial
resolution of the area of interest.
Although small-field-of-view imaging in other areas such as the
temporomandibular joints, fingers, and neck has been described, its use in MRI
of large joints has not, to our knowledge, been reported. Through five
illustrative cases, we present our experience with MRI of large
joints—namely, the knee and ankle—using a small-field-of-view
surface coil to provide high-resolution images of predetermined areas of
interest.
Materials and Methods
Patients
Five patients with clinical suspicion of internal joint derangement were
examined on MRI. Three patients underwent knee examination, and two underwent
ankle examination.
Imaging Methods
All examinations were performed on the same 1.5-T whole-body MRI scanner
(Magnetom Sonata, Siemens). For the standard knee or ankle examination, a
23-cm circularly polarized extremity no-tune transmit–receive coil was
used. For the knee, the standard sequences were sagittal proton density
(TR/TE, 3,500/45) and fat-saturated T2-weighted (2,220/83), coronal
T1-weighted (450/14) and fat-saturated T2-weighted, and axial fat-saturated
proton density images. For the ankle, the standard sequences for suspected
osteochondral defects were sagittal proton density (3,500/45) and
fat-saturated T2-weighted (2,220/83), coronal T1-weighted (450/14) and
fat-saturated T2-weighted, and axial fat-saturated proton density images. The
parameters were as follows: field of view, 13.3 x 16.0 cm; matrix, 256
x 512 (displayed at 512 x 512); slice thickness, 3 mm with a
0.3-mm interslice gap; and number of averages, 3. Turbo spin-echo sequences
were used for all acquisitions except the T1-weighted sequence, for which a
spin-echo sequence was used.
The images were reviewed by the attending musculoskeletal radiologist, and
if a subtle lesion or a suspicious area was identified, additional images were
obtained using a surface coil provided the area of interest was superficial
enough to be encompassed by the small-field-of-view surface coil. For
additional targeted imaging, a commercially available 4-cm no-tune loop
receive surface coil approved by the United States Food and Drug
Administration was placed over the area of interest. The optimal site of
placement, with the shortest depth from surface to lesion, was chosen using
the standard images as a guide. Parameters for small-field-of-view imaging
were as follows: field of view, 10.0 x 10.0 cm; matrix, 256 x 512
(displayed at 512 x 512); slice thickness, 3 mm with a 0.3-mm interslice
gap; and number of averages, 3. Imaging planes and sequences for the
examination with the small-field-of-view coil were selected on the basis of
the planes and sequences that were most informative with the standard coil.
The overall scanning time was increased by an amount dependent on the number
of additional sequences used. On average, only two sequences were performed,
adding less than 15 min of additional preparation and scanning time.
Image Evaluation
Spatial resolution.—Because the scanning sequences were
identical, direct comparison of the spatial resolution was possible using the
field of view and pixel matrix (Appendix 1).
Contrast resolution.—The signal-to-noise ratio was measured
in patient 1 as a reference. The same area (in this patient, the lateral
aspect of the knee centered on the lateral meniscus) was examined using an
extremity coil with a 16-cm field of view, an extremity coil with a 10-cm
field of view, and a superficial coil with a 10-cm field of view. An identical
coronal proton density sequence was used. An identical area (i.e., region of
interest) was chosen at the lateral tibial plateau on all three examinations,
and the signal intensity was measured. To estimate the degree of background
noise, we used an area in the scanning plane and outside the patient to assess
the standard deviation of signal intensity. The ratio of the two values was
used as an estimation of the signal-to-noise ratio (Appendix 1).
Results
Image Resolution
Spatial resolution.—In-plane resolution for the superficial
coil (10-cm field of view) was 0.076 mm2 (pixel size) compared with
0.200 mm2 (pixel size) for the standard knee coil (16-cm field of
view) (Appendix 1), which is a 2.6-fold improvement.
Contrast resolution.—The signal-to-noise ratio for the
superficial coil was approximately double that of the standard knee coil
(Appendix 1). Although there was a decrease in signal intensity with the
smaller superficial coil, this decrease was more than compensated for by the
larger decrease in noise.
Illustrative Examples
Patient 1.—An 18-year-old man who underwent anterior
cruciate ligament reconstruction 1 year earlier presented with lateral joint
pain. Imaging with the superficial coil showed a thin line of high signal
extending to the inferior meniscal surface consistent with an undisplaced tear
(Fig. 1C). This finding was not
apparent on MR images obtained using the standard extremity coil (Figs.
1A and
B). These three images were
used to calculate the change in the signal-to-noise ratio as discussed.
View larger version (153K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —MR images of 18-year-old man who presented with lateral joint
pain in right knee; he had undergone reconstruction of anterior cruciate
ligament 1 year earlier. Coronal proton density image (3,050/39) of same area
as A and B obtained using 4-cm surface coil positioned over
lateral joint line, 10-cm field of view, and 512 x 512 matrix has
improved signal-to-noise ratio and spatial resolution compared with A
and B. Tear (arrowhead) can be seen extending to inferior
articular surface of lateral meniscus.
|
|
View larger version (174K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —MR images of 18-year-old man who presented with lateral joint
pain in right knee; he had undergone reconstruction of anterior cruciate
ligament 1 year earlier. Coronal proton density image (TR/TE, 3,500/45)
obtained using standard extremity coil, 16-cm field of view, and 512 x
512 matrix shows high-signal lesion (arrow) in peripheral portion of
lateral meniscus. Extension to articular surface is not apparent.
|
|
View larger version (187K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —MR images of 18-year-old man who presented with lateral joint
pain in right knee; he had undergone reconstruction of anterior cruciate
ligament 1 year earlier. Coronal proton density image (3,050/39) obtained
using standard extremity coil, 10-cm field of view, and 512 x 512 matrix
shows lateral meniscal lesion is less distinct on this image than on A
due to poor signal-to-noise ratio.
|
|
Studies have shown that a small proportion of arthroscopically apparent
meniscal tears were not seen not to extend to articular surface on MRI
[1,
2]. This is particularly common
in the lateral meniscus, either peripheral or posterior when an associated
anterior cruciate ligament injury is present
[3]. The improved spatial
resolution and signal-to-noise ratio achieved using a superficial coil may
improve one's diagnostic accuracy in this situation.
Patient 2.—A 27-year-old man presented with persistent pain
from an ankle injury that occurred 3 years earlier that had been treated
conservatively. Standard MRI revealed an osteochondral lesion of the talar
dome laterally (Fig. 2A). We
were not able to delineate the tibial or talar articular cartilage contour
despite magnification. Images obtained using the superficial coil showed a
flap of cartilage was separated from the subchondral bone in the region of the
deep radial zone, depression of the subchondral bone (Figs.
2B and
2C), and subchondral cysts
(Fig. 2B). The configuration
and thickness of this lesion were confirmed at arthroscopy.
View larger version (160K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. —MR images of 27-year-old man who injured ankle 3 years
earlier presented with persistent pain despite conservative treatment. Coronal
proton density image (TR/TE, 3,500/42) obtained using standard extremity coil
shows mildly displaced osteochondral lesion (arrow) involving lateral
corner of talar dome. Contour of articular cartilage of distal tibia and talar
dome is difficult to discern. Tibial and talar articular surfaces cannot be
differentiated.
|
|
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —MR images of 27-year-old man who injured ankle 3 years
earlier presented with persistent pain despite conservative treatment. Coronal
proton density image (3,500/39) of lesion shown in A obtained using
4-cm surface coil positioned over lateral malleolus shows separation of talar
cartilage (straight arrow) in deep radial zone. Cartilage surface
contour of distal tibia and of talar dome is more easily discerned. Small
cystlike lesions (curved arrow) are present in underlying bone.
|
|
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C. —MR images of 27-year-old man who injured ankle 3 years
earlier presented with persistent pain despite conservative treatment.
Sagittal proton density image (3,500/39) of lesion shown in A obtained
using 4-cm surface coil positioned over lateral malleolus shows 1-cm (anterior
to posterior) flap of cartilage separation in deep radial zone. This finding
is at "tidemark" zone (arrow), boundary between calcified
and noncalcified cartilage that is a relatively weak plane subject to shearing
injuries.
|
|
This case indicated separation at the deepest layer of the articular
cartilage, probably representing a "tidemark" separation. The
tidemark, the junction between the calcified and noncalcified cartilage, is a
vulnerable zone for cartilage, subject to injury caused by shearing forces
[4].
Using a superficial coil allows improvement in spatial resolution. This
alternative method could be used to complement efforts to improve contrast
resolution with sequences such as driven equilibrium Fourier transformation,
refocused steady-state free precession, or diffusion imaging.
Patient 3.—An 18-year-old male squash player presented with
an 18-month history of ankle pain aggravated by exercise. The contour of the
cartilage surface of an osteochondral lesion in the medial aspect of the talar
dome was difficult to discern on standard MRI
(Fig. 3A). The lesion—a
focal bulge suggesting hypertrophic, reparative fibrocartilage
(Fig. 3B)— was better
delineated with the superficial coil. High-signal foci within this bulge
suggested that the lesion was not simply a flap of displaced hyaline
cartilage.
View larger version (155K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. —MR images of ankle in 18-year-old male squash player with
18-month history of ankle pain aggravated by exercise. Sagittal proton density
image (TR/TE, 3,500/42) obtained using standard extremity coil shows
osteochondral lesion (arrow) of medial corner of talar dome.
Cartilage surface contour is difficult to discern.
|
|
View larger version (132K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —MR images of ankle in 18-year-old male squash player with
18-month history of ankle pain aggravated by exercise. Sagittal proton density
image (3,050/39) of lesion shown in A obtained using 4-cm surface coil
positioned over medial malleolus more clearly reveals thickened articular
cartilage resulting in bulging contour (arrow) than A. Signal
intensity of this thickened articular tissue differs from normal articular
hyaline cartilage, which suggests it may be fibrocartilage.
|
|
Cartilage surface contour is difficult to visualize when little joint fluid
is present [5]. The contour of
the talar dome cartilage is particularly difficult to visualize in the absence
of sufficient joint fluid because the cartilage is relatively thin and is
closely applied to the distal tibial cartilage. The thin cartilage is also
prone to MRI artifacts (namely truncation, chemical shift, and partial
volume). The use of a surface coil improved contrast and spatial resolution
because it allowed a small amount of joint fluid to delineate clearly the
contour of the cartilage that was not seen on images obtained using a standard
extremity coil. Signal differences within the cartilage are also easier to
detect using the surface coil.
Patient 4.—A 36-year-old woman presented with anterior knee
pain from a tennis injury that occurred 6 months earlier. MRI with a
superficial coil revealed a cartilage abnormality in the lateral facet of the
patella with subchondral bone change (Fig.
4). An area of slightly thickened articular cartilage was visible
at the medial aspect of the lateral patellar facet. The thickened cartilage
shows loss of the trilaminar pattern (as seen on the medial facet). The
cartilaginous signal is heterogeneous with high signal in the region of the
radial (deep) zone and loss of the superficial dark line (truncation artifact)
(Fig. 4). This detail was not
apparent on similar sequences using a standard knee coil.
View larger version (111K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4. —MR image of patellar region in 36-year-old woman with
anterior knee pain from tennis injury 6 months earlier. Axial fat-saturated
proton density image (TR/TE, 3,500/39) obtained using 4-cm surface coil
positioned over patella reveals abnormality in lateral facet articular
cartilage of patella with subchondral bone changes. Mild focal swelling of
articular cartilage (arrow) is visible. Swollen cartilage shows loss
of trilaminar pattern (as seen on medial facet). Cartilage signal is
heterogeneous with high signal in region of radial (deep) zone and loss of
superficial dark line (truncation artifact). This detail was not apparent on
standard extremity coil imaging (not shown).
|
|
Patient 5.—A 13-year-old female long-jump athlete presented
with knee pain from a knee sprain that occurred during training 5 days
earlier. A moderate-sized joint effusion helps outline the meniscofemoral
ligament, which is avulsed from its femoral attachment
(Fig. 5A). The tear extended
to involve the overlying fat. The rest of the medial collateral ligament
complex was intact. This detail was not apparent on similar sequences using a
standard knee coil. This area is a blind spot for arthroscopy, making clear
delineation of an MRI-detectable abnormality
valuable.
View larger version (134K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A. —MR images of knee in 13-year-old female long-jump athlete who
was injured during training 5 days earlier. Coronal proton density image
(TR/TE, 3,500/39) using 4-cm surface coil positioned over medial joint line
shows moderatesized joint effusion outlines meniscofemoral ligament
(straight arrow), which has been avulsed from its femoral attachment.
Small blood clot (curved arrow) is present. Tear (arrowhead)
extends to involve overlying fat.
|
|
View larger version (136K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B. —MR images of knee in 13-year-old female long-jump athlete who
was injured during training 5 days earlier. Coronal proton density image
(3,500/39) of contralateral normal knee was obtained using 4-cm surface coil
positioned over medial joint line for comparison with injured knee. Intact
meniscofemoral ligament (arrow) is clearly visible.
|
|
Discussion
MRI of large joints has improved over the past decade with better
examination technique and higher image resolution, thus increasing diagnostic
accuracy. Various techniques have been developed to improve images, such as MR
arthrography, tailored cartilage sequences, positional and kinematic imaging,
and oblique imaging planes to better visualize particular structures. This
study illustrates how the resolution of images of large joints can be further
improved by applying a superficial coil on the area of interest identified on
a standard examination.
In our experience over the past 18 months, a superficial coil can be used
with good effect for examining superficial structures in large joints on
imaging. The standard examination can be used to identify areas of
abnormality, and a superficial coil can subsequently be used to provide better
resolution when required. This is provided the area of interest is around the
optimal signal depth encompassed by the smaller field of view of the surface
coil, which is, as a rule, half the diameter of the coil. In practical terms,
the 4-cm superficial coil described here is useful for examining structures at
more peripheral areas of the knee and ankle joints but not in central areas.
Changing coils during an examination is not part of standard MRI. In most
clinical applications, applying a superficial coil helped clarify the final
diagnosis and allowed clearer depiction of subtleties apparent on standard
imaging in most cases. Changing coils during MRI examination is akin to using
transducers of different frequencies during sonography to look at deeper or
superficial structures. In a small or remote radiology practice, the attending
radiologist may not be available to supervise the MRI examination during
scanning. In such situations, patients could be called back at a later date
for additional imaging with a superficial coil. Recalling patients for more
detailed imaging when the routine protocol is insufficient is practiced in
many centers.
Not all the cases presented had surgical or arthroscopic correlation, the
traditional gold standard; therefore, the sensitivity of imaging to detect
lesions using a superficial coil compared with using a standard coil was not
determined. On the other hand, visualization of subtle abnormalities on
imaging may reduce the need an invasive procedure such as arthroscopy.
There are limitations to using a superficial coil, the most important of
which is the limited usable field of view. Because the superficial coil is of
4-cm diameter, optimal signal is obtained in the 2-cm range from the coil
position. Signal drop-off is progressive for deeper structures, and signal
increase is progressive for superficial structures. This limits the usable
field of view. Fortunately, most structures of clinical interest lie within
this usable field of view. The coil should be placed immediately over the area
of interest and repositioning may occasionally, although infrequently, be
required.
In conclusion, the use of a surface coil for imaging the knee and ankle
joints provides superior image resolution for some lesions compared with the
use of a standard extremity coil; a surface coil is a useful adjunct to
standard MRI.
View this table:
[in this window]
[in a new window]
|
APPENDIX 1. Spatial Resolution, Contrast Resolution, and Signal-to-Noise Ratio for
MR Images Obtained Using a Superficial Coil Versus a Standard Extremity
Coil
|
|
References
- Kaplan PA, Nelson NL, Garvin KL, Brown DE. MR of the knee: the
significance of high signal in the meniscus that does not clearly extend to
the surface. AJR1991; 156:333
–336[Abstract/Free Full Text]
- De Smet AA, Norris MA, Yandow DR, Quintana FA, Graf BK, Keene JS.
MR diagnosis of meniscal tears of the knee: importance of high signal in the
meniscus that extends to the surface. AJR1993; 161:101
–107[Abstract/Free Full Text]
- De Smet AA, Graf BK. Meniscal tears missed on MR imaging:
relationship to meniscal tear patterns and anterior cruciate ligament tears.
AJR 1994;162:905
–911[Abstract/Free Full Text]
- Kumar P, Oka M, Nakamura T, Yamamuro T, Delecrin J. Mechanical
strength of osteochondral junction. Nippon Seikeigeka Gakkai
Zasshi 1991;65:1070
–1077[Medline]
- Reeder SB, Pelc NJ, Alley MT, Gold GE. Rapid MR imaging of
articular cartilage with steady-state free precession and multipoint
fat–water separation. AJR2003; 180:357
–362[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Introduction
Materials and Methods
