M u s c u l o s k e l e t a l I m a g i n g • P i c t o r i a l E s s ay

Jamadar et al.
Pitfalls in Musculoskeletal Sonography

Musculoskeletal Imaging
Pictorial Essay
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved

Musculoskeletal Sonography:
Important Imaging Pitfalls
David A. Jamadar 1 OBJECTIVE. The purpose of this article is to describe the pitfalls that may be encoun-
Brian L. Robertson tered when performing musculoskeletal sonography.
Jon A. Jacobson CONCLUSION. Sonography of the musculoskeletal system is a useful diagnostic tech-
Gandikota Girish nique, but awareness and understanding of the pitfalls will minimize errors in diagnosis.
Brian J. Sabb
Yebin Jiang

Yoav Morag
Jamadar DA, Robertson BL, Jacobson JA, et al.
Keywords: musculoskeletal, pitfalls, sonography
DOI:10.2214/AJR.09.2712
Received March 9, 2009; accepted after revision
July 13, 2009.
1
All authors: Department of Radiology, University of
Michigan Hospitals, 1500 E Medical Center Dr., Ann
Arbor, MI 48109. Address correspondence to
D. A. Jamadar (djamadar@med.umich.edu).
AJR 2010; 194:216–225
0361–803X/10/1941–216
© American Roentgen Ray Society
L
earning musculoskeletal sonog-
raphy can be challenging. The
sonographic properties of soft
tissues, tendons, and ligaments
result in pitfalls that have to be recognized.
Pitfalls may occur for a variety of reasons.
The sonographic properties of different tis-
sues may be similar, as may be found in the
shoulder. Anatomy may be nonintuitive, as
in the elbow and hip, and anisotropy [1, 2]
may result in misleading appearances.
Anisotropy is the sonographic property of
linearly organized tissues, such as tendon
and ligaments (and to a lesser extent nerves)
in which sonographic appearance is deter-
mined in part by the angle of insonation of
the ultrasound beam [3] (Fig. 1). We present
important pitfalls that we have encountered
over the past decade of our practice.
General Considerations
Before starting the sonography examina-
tion, a short focused patient history is elic-
ited and the area of concern is examined.
A protocol-driven comprehensive sonog-
raphy examination is required when imag-
ing the shoulder and is advocated with other
joints as well to identify causes of referred
pain, develop an efficient examination, un-
derstand normal structures, and allow iden-
tification of subtle pathology; however, the
patient is asked to indicate a point of most
discomfort because this will often identify
the site of the abnormality [4]. The patient is
placed in a comfortable position that facili-
tates an easy examination as well as allow-
ing dynamic maneuvers. A high-frequency
linear transducer (10–12 MHz) is typically
used, although in the hands and feet, higher-
frequency transducers (15–17 MHz) provide
exquisite anatomic detail, and occasionally
in obese patients, particularly when exam-
ining the hip, a lower-frequency transducer
(5–9 MHz) may be necessary.
Shoulder
Lying between the supraspinatus and the
acromion and deltoid muscles is the subac-
romial–subdeltoid bursa, which extends be-
yond the distal insertion of the supraspinatus
tendon [5] (Fig. 2). Sometimes it is difficult
to differentiate tendon from bursa lying on
its superficial surface because the echoge-
nicity and contour may be similar. Overes-
timation of the thickness of the tendon as
well as misinterpretation of the echogenici-
ty of the tendon may result (Figs. 3 and 4).
By identifying the bursa that passes beyond
the supraspinatus insertion on the greater tu-
berosity and extrapolating proximally on a
long-axis image of the supraspinatus tendon,
one can get a reasonable idea of what consti-
tutes bursa and what is the actual supraspi-
natus tendon.
The long head of the biceps brachii tendon
is found in the bicipital groove of the proximal
humerus. When the tendon of the long head of
the biceps is dislocated or torn, there can be
linear echoes in the bicipital groove that may
simulate intact tendon fibers, although much
thinner compared with the normal biceps ten-
don [6, 7] (Fig. 5). Careful evaluation medial
to the proximal humerus for a dislocated ten-
don is necessary to exclude a dislocated bi-
ceps tendon. A long-standing tear of the long
head of the biceps brachii tendon may result

216 AJR:194, January 2010


in increased echogenicity of part of the biceps
muscle, a combination of retracted tendon and
fatty change, which is a clue that the tendon
may have been torn (Fig. 6).
Additional pitfalls that can be encountered
when scanning the shoulder include mistak-
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved
ing anisotropy at the base of the supraspina-
tus tendon for a tear (Fig. 7); the anisotro-
py resulting in a hypoechoic biceps tendon in
the long axis, simulating tendonosis or mak-
ing the tendon difficult to visualize (Fig. 8);
and the normal anatomy of the multipennate
subscapularis tendon simulating clefts in the
tendon or tendonosis (Fig. 9).
Elbow
The deep branch of the radial nerve, the
posterior interosseous nerve, becomes flat-
tened as it passes through the supinator, and
in long-axis imaging it appears to narrow af-
ter entering this muscle (Figs. 10 and 11).
Awareness of this normal anatomic feature
of this specific nerve will prevent a misdi-
agnosis of entrapment syndrome due to the
apparent change in caliber of the nerve as it
passes into the supinator muscle.
Wrist
The extensor retinaculum at the wrist is
a fibrocartilaginous sling that holds the ex-
tensor tendons close to the dorsal wrist and
prevents bowstringing. This normally hyper-
echoic structure may appear hypoechoic be-
cause of anisotropy and may simulate a com-
plex fluid collection or tenosynovitis [3] (Fig.
12). Profound thickening of the extensor reti-
naculum without much underlying tenosyno-
vitis may be a presentation of de Quervain’s
tenosynovitis.
At the wrist, it is possible to confuse the
median nerve with one of the long flexor ten-
dons and vice versa. This can be avoided by
recognizing that the median nerve is the only
long linear structure at the wrist that changes
anatomic planes more proximally in the fore-
arm. This nerve is initially superficial at the
wrist but lies between the flexor digitorum
superficialis and flexor digitorum profundus
more proximally in the forearm [8] (Fig. 13).
The long flexor tendons may also be distin-
guished from the median nerve because they
merge with their respective muscles more
proximally and display anisotropy. The me-
dian nerve does not show as much anisotro-
py as the adjacent flexor tendons, which is
useful in differentiating between these struc-
tures (Fig. 14). These changes also affect a
bifid median nerve (Fig. 15).
AJR:194, January 2010 217
Pitfalls in Musculoskeletal Sonography
Hip and Thigh
The rectus femoris muscle originates prox-
imally from two tendons: a direct (straight)
head from the anterior inferior iliac spine that
courses approximately parallel to the over-
lying skin and an indirect (reflected) head
that originates from the lateral acetabulum
and courses at an angle to the overlying skin.
This orientation of the tendons results in the
straight head being easily identified by its
echogenic parallel linear appearance, where-
as the indirect head usually shows marked
anisotropy (Fig. 16), which may be misin-
terpreted as acoustic shadowing from calci-
fication [9] or ossification or the shadowing
caused by refraction of sound at the edge of
a tendon tear. By scanning more laterally and
obliquely angling the transducer, the reflected
head may be identified (Fig. 17).
At the anterior recess of the hip, the anteri-
or joint capsule reflects back superiorly along
the femoral neck. In the absence of an effu-
sion, the two layers of the capsule lying on
each other may give the impression of com-
plex fluid or synovitis in the hip joint (Fig.
18). Joint fluid may be identified separating
these two layers (Fig. 19). It has been shown
that a measurement of the thickness of the
tissue adjacent to the anterior femoral neck
should measure more than 7 mm. Asymme-
try of this measurement of more than 1 mm
between both sides suggests the diagnosis of
fluid in the hip joint [10].
The sciatic nerve is proximate to the ham-
string insertion at the hip. Knowledge of the an-
atomic relationship of the sciatic nerve to the
hamstring tendon is necessary to avoid confus-
ing these linear fibrillar structures [11] (Fig.
20A). More proximally, the structures are even
closer (Fig. 20B), emphasizing that care should
be taken with needle placement at the hamstring
origin when a medial approach is prudent.
Knee
The semimembranosus tendon curves an-
teriorly to attach to the posterior tibia where
anisotropy may make it hypoechoic, simulat-
ing a Baker cyst (Fig. 21A). By angling the
ultrasound beam along the long axis of the
semimembranosus tendon, anisotropy [12]
can be visualized confirming that this hy-
poechoic structure is a tendon (Fig. 21B).
When examining the lateral knee in a su-
pine patient, the knee may be slightly flexed
and in valgus so that the fibular collateral lig-
ament may have a wavy contour (Fig. 22).
This is a normal appearance and does not in-
dicate a tear or laxity of the ligament.
Foot and Ankle
The insertion of the tibialis posterior ten-
don is complex, fanning out and inserting di-
rectly onto the navicular and also possibly an
accessory ossicle. The tendon also sends a
slip (deeper lateral division) to insert more
distally at the middle cuneiform and the
bases of the second, third, and fourth meta-
tarsals. This complicated insertion (distal
1.5–2.0 cm) results in the tendon appearing
thicker than it does along its more proximal
length, and the different orientation of its fi-
bers often shows anisotropy [13] (Figs. 23
and 24), which should not be misinterpreted
as tendinosis or tear.
Conclusion
There are many pitfalls in musculoskeletal
sonography due to misinterpretation of nor-
mal anatomic structures as well as misinter-
pretation of pathologic conditions. Knowl-
edge of such pitfalls will improve accuracy
in the sonographic diagnosis of musculo­
skeletal disorders.
Acknowledgments
We thank Tracy Boon, Rita Lewis, Alli-
son Berry, Heidi Ehrich, and Yinghui Bian
for their help and support during the prepa-
ration of the manuscript.
References
1. Garcia T, Hornof WJ, Insana MF. On the ultra-
sound properties of tendon. Ultrasound Med Biol
2003; 29:1787–1797
2. Crass JR, van de Vegte GL, Harkavy LA. Tendon
echogenicity: ex vivo study. Radiology 1988;
167:499–501
3. Robertson BL, Jamadar DA, Jacobson JA, et al.
Extensor retinaculum of the wrist: sonographic
characterization and pseudotenosynovitis appear-
ance. AJR 2007; 188:198–202
4. Jamadar DA, Jacobson JA, Caoili EM, et al. Mus-
culoskeletal sonography technique: focused versus
comprehensive evaluation. AJR 2008; 190:5–9
5. van Holsbeeck M, Strouse PJ. Sonography of the
shoulder: evaluation of the subacromial–subdel-
toid bursa. AJR 1993; 160:561–564
6. Farin PU. Sonography of the biceps tendon of the
shoulder: normal and pathologic findings. J Clin
Ultrasound 1996; 24:309–316
7. Armstrong A, Teefey SA, Wu T, et al. The effica-
cy of ultrasound in the diagnosis of long head of
the biceps tendon pathology. J Shoulder Elbow
Surg 2006; 15:7–11
8. Jamadar DA, Jacobson JA, Hayes CW. Sono-
graphic evaluation of the median nerve at the
wrist. J Ultrasound Med 2001; 20:1011–1014
 Jamadar et al.

9. Sarkar JS, Haddad FS, Crean SV, Brooks P. Acute
calcific tendonitis of the rectus femoris. J Bone
Joint Surg Br 1996; 78:814–816
10. Koski JM, Anttila PJ, Isomaki HA. Ultrasonogra-
phy of the adult hip joint. Scand J Rheumatol
1989; 18:113–117
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved
11. Miller SL, Gill J, Webb GR. The proximal origin
of the hamstrings and surrounding anatomy en-
countered during repair: a cadaveric study. J Bone
Joint Surg Am 2007; 89:44–48
12. Connolly DJA, Berman L, McNally EG. The use
of beam angulation to overcome anisotropy when
viewing human tendon with high frequency linear
array ultrasound. Br J Radiol 2001; 74:183–185
13. Patel S, Fessell DP, Jacobson JA, Hayes CW, van
Holsbeeck MT. Artifacts, anatomic variants, and
pitfalls in sonography of the foot and ankle. AJR
2002; 178:1247–1254

A B
Fig. 1—Sonography in 54-year-old woman with normal long head of biceps tendon (arrow) showing anisotropy. GT = greater tuberosity, LT = lesser tuberosity,
Del = deltoid muscle, Sub = subscapularis muscle.
A, In sonogram obtained with tendon fibers at right angles to beam, tendon appears echoic.
B, In sonogram obtained with tendon fibers at angle to beam, tendon appears hypoechoic.

Fig. 2—63-year-old woman with shoulder pain and bursal thickening. Sonogram
in long axis of supraspinatus tendon (SST) shows subacromial–subdeltoid
bursal thickening (all arrowheads) extending beyond insertion of supraspinatus
tendon on greater tuberosity (GT) of humerus. Bursa is easily differentiated
from supraspinatus tendon. Note outer layer of supraspinatus tendon (black
arrowheads) is echogenic but less so than peribursal fat layer (white arrowheads).
Bursa lies between these two echogenic interfaces. D = deltoid muscle.
Fig. 3—43-year-old man with right shoulder pain and bursal thickening.
Sonogram in long axis of supraspinatus tendon (SST) shows bursal thickening
(all arrowheads) extending beyond insertion of supraspinatus tendon on greater
tuberosity (GT) of humerus. Bursa in this patient is not as easily differentiated from
supraspinatus tendon. Note visualized outer layer of supraspinatus tendon (white
arrowheads) is echogenic but less so than peribursal fat layer (black arrowhead).
D = deltoid muscle.

Fig. 4—39-year-old woman with right shoulder injury, pain, and bursal thickening.
Sonogram in long axis of supraspinatus tendon (SST) shows bursal thickening
(arrowheads) extending beyond insertion of supraspinatus tendon on greater
tuberosity (GT) of humerus. Bursa and peribursal fat in this patient are difficult to
differentiate from supraspinatus tendon because echogenic interface between
supraspinatus tendon and bursa is not appreciated. D = deltoid muscle.

218 AJR:194, January 2010

 Pitfalls in Musculoskeletal Sonography
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved

A B
Fig. 5—73-year-old man with right shoulder injury and dislocated biceps tendon.
A, Long-axis sonogram through bicipital groove shows linear echoes (arrows) not
consistent with tendon fibers. Hum = humeral diaphysis.
B, Short-axis sonogram through bicipital groove (curved arrow) shows echoes
(straight arrows) not consistent with normal tendon. Long head of biceps brachii
tendon (arrowhead) lies over lesser tuberosity (LT).
C, Long-axis sonogram through dislocated biceps tendon (arrows) shows normal
tightly packed linear appearance lying over lesser tuberosity proximally.

A B
Fig. 6—85-year-old woman with right shoulder pain after fall, with long-standing complete tear of long head of biceps brachii tendon.
A, Long-axis sonogram through bicipital groove shows linear echoes (arrows) but no visualization of normal biceps tendon. Hum = humeral diaphysis.
B, Axial sonogram through biceps muscle shows echogenicity of long head (LH arrow) secondary to fatty change. Normal-appearing short head of muscle (SH arrow) is
seen medially representing normal short head of biceps brachii. Curved arrow indicates humeral diaphysis.

AJR:194, January 2010 219

 Jamadar et al.
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved

A B
Fig. 7—54-year-old woman with normal supraspinatus tendon. SST = supraspinatus tendon, Delt = deltoid, Gt Tub = greater tuberosity.
A, Inserting supraspinatus tendon fibers curve away from ultrasound beam and appear hypoechoic (arrows), which should not be interpreted as tear.
B, By angling transducer to bring orientation of these fibers at right angles to ultrasound beam, hypoechoic area fills in with normal-appearing tendon (asterisk).

A B
Fig. 8—54-year-old woman with normal long head of biceps tendon showing anisotropy.
A, When long head biceps tendon (arrows) is at right angle to ultrasound beam, normal echogenic fibrillar pattern is seen.
B, When long head biceps tendon (arrows) is oriented at angle to ultrasound beam, echogenic fibrillar pattern is not as well seen or is lost, simulating tear or tendonosis.

Fig. 9—42-year-old man with normal subscapularis tendon imaged in short

axis. Multipennate tendon shows echogenic tendon slips (T) with intervening
hypoechogenic slips of muscle that may simulate cleft or tear. Delt = deltoid
muscle, Les Tub = lesser tuberosity of humerus.

220 AJR:194, January 2010

 Pitfalls in Musculoskeletal Sonography

Fig. 10—20-year-old woman with medial right forearm subcutaneous nodule.

Sonogram in long axis shows normal posterior interosseous nerve before (curved
arrow) and after (straight arrow) entering supinator muscle (Sup). Apparent
change in caliber is normal because nerve becomes flattened in supinator muscle.
Rad = radius, C = capitellum.
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved

A B

Fig. 11—61-year-old man with left medial elbow pain. Sup = supinator muscle.
A, Sonogram along long axis of normal posterior interosseous nerve shows
apparent narrowing of nerve as it enters supinator muscle. Curved arrow indicates
nerve before entering supinator, and straight arrow indicates nerve after entering
supinator.
B, Sonogram along short axis of normal posterior interosseous nerve (curved
arrow) before it enters supinator muscle shows rounded appearance. Arrowheads
indicate radius.
C, Sonogram along short axis of normal posterior interosseous nerve (arrow)
within supinator muscle shows more flattened appearance. Rad = radius.
C

AJR:194, January 2010 221

 Jamadar et al.

Fig. 12—78-year-old woman with normal extensor retinaculum. Sonogram

through long axis of extensor digitorum tendons (ED) shows extensor retinaculum
as hypoechoic structure (arrows). This should not be misinterpreted as fluid
collection or tenosynovitis. Rad = distal radius, Carp = carpal bones.
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved

A B
Fig. 13—51-year-old asymptomatic man with normal sonographic anatomy
at wrist. Fl Dig = flexor digitorum superficialis and flexor digitorum profundus
tendons, PL = flexor pollicis longus, A = radial artery, Pro Quad = pronator
quadratus muscle, Rad = radius, U = ulna, Fl Dig Sup = flexor digitorum superficialis
muscle, Fl Dig Prof = flexor digitorum profundus muscle, MN = median nerve.
A, Axial sonogram just proximal to distal volar wrist crease shows median nerve
(curved arrow) lying superficial (anterior) to flexor digitorum superficialis and
flexor digitorum profundus tendons, lateral to palmaris longus tendon (straight
arrow), and medial to flexor carpi radialis (arrowhead) and flexor pollicis longus.
B, Axial sonogram obtained more proximally shows median nerve (curved arrow)
lying between flexor digitorum superficialis muscle and flexor digitorum profundus
muscle, having moved proximally to deeper anatomic plane. Palmaris longus
tendon (straight arrow) and flexor carpi radialis tendon remain in same superficial
plane. Flexor carpi radialis tendon has now merged with its muscle (arrowhead).
C, Sonogram along long axis of median nerve shows median nerve passing from
superficial at wrist (right of image) to lie between flexor digitorum superficialis and
flexor digitorum profundus more proximally. C

A B
Fig. 14—51-year-old man with normal median nerve and flexor tendons showing effect of anisotropy. T = flexor tendons, Uln = ulna, Rad = radius.
A, Sonogram with ultrasound beam at right angles to median nerve (curved arrow) and long flexor tendons (T) including flexor carpi radialis (arrowhead) shows normal
cross-section appearances. Straight arrow indicates radial artery.
B, By angling transducer, flexor tendons (T, arrowhead) show marked anisotropy, becoming hypoechoic. Median nerve (curved arrow) shows much less anisotropy.
Straight arrow indicates radial artery.

222 AJR:194, January 2010

 Pitfalls in Musculoskeletal Sonography

Fig. 15—Sonogram in 62-year-old woman shows bifid median nerve (N) lying
superficial to flexor digitorum tendons. Note small persistent median artery
(arrow).
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved

A B
Fig. 16—28-year-old man with right inguinal pain with normal rectus femoris origin.
A, Axial sonogram just distal to origin of rectus femoris muscle shows echogenic tendon (S) of direct (straight) head lying adjacent to area of intense anisotropy (A) of
indirect head of same muscle. Sar = sartorius muscle.
B, Sagittal sonogram at origin of rectus femoris muscle shows tendon of direct (straight) head (S) and area of intense anisotropy (A) of indirect head. AIIS = anterior
inferior iliac spine.

Fig. 17—Sonogram in 18-year-old woman shows normal tendon of reflected head

(arrowheads) of rectus femoris muscle (RF). Curved arrow indicates acetabular
labrum. Acet = acetabulum, Fem = femoral head.

AJR:194, January 2010 223

 Jamadar et al.
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved

Fig. 18—28-year-old man with right inguinal pain. Sonogram of normal anterior hip Fig. 19—3-year-old girl with hip joint effusion. Anterior hip joint capsule is
joint shows capsule (arrowheads) over anterior femoral neck (Fem) with trace of composed of two layers. Large arrowheads indicate anterior layer and small
fluid (arrow). H = femoral head. arrowheads indicate posterior layer, shown separated from each other by effusion
(Eff). A = acetabulum, Fem = femoral neck.

A B
Fig. 20—23-year-old man with normal proximal posterior thigh. Axial images show relationship between sciatic nerve and adjacent structures. F = femur, AM = adductor
magnus muscle, BF = biceps femoris muscle, Glut = gluteus maximus muscle.
A, Note similarity in echogenicity between sciatic nerve (curved arrow) and semimembranosus tendon (straight arrow) and conjoint tendon of biceps femoris long head
and semitendinosus (arrowhead).
B, More proximally, sciatic nerve (curved arrow) is very close to semimembranosus tendon (straight arrow) and conjoint tendon (arrowhead).

A B
Fig. 21—23-year-old man evaluated to rule out Baker cyst. G = medial head of gastrocnemius muscle, Fem = medial femoral condyle, A = popliteal artery, SM =
hypoechoic structure.
A, Sonogram over popliteal fossa reveals hypoechoic structure that may be interpreted as complex fluid-filled Baker cyst lying adjacent to tendon (arrow) of medial head
of gastrocnemius muscle.
B, In axial plane, by angling transducer in cranial caudal direction, hypoechoic structure becomes hyperechoic and is shown to be tendon of semimembranosus muscle.

224 AJR:194, January 2010

 Pitfalls in Musculoskeletal Sonography
Downloaded from www.ajronline.org by 187.131.160.44 on 07/20/23 from IP address 187.131.160.44. Copyright ARRS. For personal use only; all rights reserved

Fig. 22—75-year-old woman with normal fibular collateral ligament. Sonogram
along long axis of fibular collateral ligament (straight arrows) shows normal wavy
appearance, which should not suggest tear. Ligament merges with tendon of
biceps femoris (asterisks) to insert into fibula (out of field of view). Curved arrow
indicates popliteus tendon. Fem = femur, Tib = tibia.
Fig. 23—51-year-old man with fifth-toe pain. As normal tibialis posterior tendon
approaches its insertion into navicular bone (N), it widens (arrowheads and
arrows), with deepest fibers showing anisotropy (arrows). These deeper fibers
(arrows) are composed of lateral division of tendon insertion and run at oblique
angle to larger and more superficial fibers (arrowheads). This thickening and
hypoechoic appearance should not be routinely misinterpreted as tendinosis.
T = talus.

Fig. 24—Sonogram with extended field of view of normal tibialis posterior tendon
in 56-year-old woman shows echogenic linear fibrillar echotexture of tendon (large
arrowheads) and effect of mild anisotropy on tendon (curved arrows).
Distally, tendon appears hypoechogenic from more marked anisotropy of complex
insertion of tendon (small arrowheads) into navicular bone (Nav). Tal = talus, Sp =
superomedial calcaneonavicular component of spring ligament.

AJR:194, January 2010 225