Ultrasound in Med. & Biol., Vol. 37, No. 8, pp.

1204–1209, 2011
Copyright Ó 2011 World Federation for Ultrasound in Medicine & Biology
Printed in the USA. All rights reserved
0301-5629/$ - see front matter

doi:10.1016/j.ultrasmedbio.2011.05.008

d Original Contribution

SONOGRAPHIC APPEARANCES OF MORTON’S NEUROMA: DIFFERENCES

FROM OTHER INTERDIGITAL SOFT TISSUE MASSES

HEE-JIN PARK,*y SAM SOO KIM,y MYONG-HO RHO,* HYUN-PYO HONG,* and SO-YEON LEE*
* Department of Radiology, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Republic
of Korea; and y Department of Radiology, School of Medicine, Kangwon National University, Chuncheon, Republic of Korea

(Received 26 January 2011; revised 3 May 2011; in final form 9 May 2011)

Abstract—The purpose of the study was to evaluate the ultrasonographic characteristics of Morton’s neuroma
(MNs) and the usefulness of the ‘‘ginkgo leaf sign’’ for differentiating MNs from other interdigital soft tissue
masses. The inclusion criteria were 27 patients with the masses in the intertarsal region with surgical proof. Four-
teen masses in the 10 patients (mean age, 46) were MNs and nine cases of nine patients were ganglion cysts, seven
cases (seven patients) of epidermoid tumors and one case of fibroma were included. Ultrasonographic examina-
tions were performed by a musculoskeletal radiologist using the HDI 5000 (Philips, Bothell, WA, USA) or the Logiq
E9 (GE Medical Systems, Milwaukee, WI, USA) equipped with a linear 6–15 MHz probe, and the findings were
interpreted in consensus by two musculoskeletal radiologists. The ultrasonographic findings such as margin,
size, echogenicity and deepness of the MNs were compared with those for other interdigital soft tissue masses.
The ginkgo leaf sign was defined as the appearance of a biconcave shape of the mass from compression by adjacent
structures. The mean size of the MNs was 5.6 mm. There was a significant difference in incidence between males
and females (female dominant, p 5 0.003). There was no difference in incidence with regard to age (p 5 0.259). All
lesions were hypoechoic (100%, 14/14) and 10 cases exhibited the ginkgo leaf sign (71%, 10/14, p , 0.001). The
lesions were either well marginated (43%, 6/14) or poorly marginated (57%, 8/14, p 5 0.075). None of the lesions
abutted adjacent bony structures (p , 0.001). Interdigital MNs are primarily found in middle-aged women and
often demonstrate the ginkgo leaf sign. MNs are hypoechoic and do not abut adjacent bony structures. Based
on our findings, we believe ultrasound of interdigital soft tissue masses may provide useful information with respect
to their location to adjacent soft tissue structures. Detection of our Gingko leaf sign may be specific for Morton’s
neuromas and more studies are needed to confirm its value as a sonographic sign. (E-mail: parkhiji@kangwon.
ac.kr) Ó 2011 World Federation for Ultrasound in Medicine & Biology.
Key Words: Ultrasound (US), Morton, Neuroma, Soft tissue, Tumor.

INTRODUCTION The diagnosis of MN is suggested clinically by pain and

paresthesia radiating from the midfoot to the toes. Ultra-
Morton’s neuroma (MN) is characterized by non-
sound (US) is sensitive and specific in confirming the
neoplastic fusiform enlargement of a digital branch of
diagnosis of MN, with 95%–98% accuracy having been
the medial or lateral plantar nerve. It is not a true neuroma
reported in many studies (Hughes et al. 2007). However,
and is, therefore, characterized by perineural fibrosis,
in practice, some cases of false positive and false negative
vascular proliferation, edema of the endometrium and
diagnosis can be easily seen. We think the main cause of
axonal degeneration (Quinn et al. 2000). Females
the misdiagnosis is nonspecific morphologic characteris-
between the ages of 18 and 85 years and athletes are
tics and no vascular flow pattern. We encounter the MN as
most commonly affected (Murphy et al. 2004). The
nonspecific soft tissue mass. Neuromas are usually hypo-
condition is believed to be an entrapment neuropathy
echoic masses in the interdigital space; however, other
secondary to compression of the common interdigital
soft tissue masses can be found in this space. As such,
nerve under the overlying transverse metatarsal ligament.
the sonographic differentiation of MN from other lesions
is important for formulating focused therapeutic plans. In
Address correspondence to: So-Yeon Lee, Department of Radi- this article, we describe our experience with ultrasound
ology, Kangbuk Samsung hospital, Sungkyunkwan University School
of Medicine, #108 Pyung-dong, Jongno-gu, Seoul 110-102, Republic features of MNs that are useful for differentiating them
of Korea. E-mail: parkhiji@kangwon.ac.kr from other interdigital soft tissue tumors.

1204
 Sonographic appearances of Morton’s neuroma d H.-J. PARK et al.
MATERIALS AND METHODS
Patients
Of the patients who had visited our hospital due to
interdigital masses between January 2007 and July
2010, 27 patients who underwent surgical intervention
and pathologically confirmed were included in the study.
The candidates for the surgery were patients who had the
prominent mass and combined unbearable symptom and
who had a mass larger than 2 cm. Fourteen masses in the
10 patients (mean age, 46) were MNs and nine cases of
nine patients were ganglion cysts, seven cases (seven
patients) of epidermoid tumors and one case of fibroma
were included. Of the MNs, nine of the lesions were in
the third interdigital space and five of the lesions were
in the second interdigital space. Four patients had two
MNs. Three of these patients had the lesions in the
same foot and one patient had one lesion in each foot.
All of the patients complained of pain in the metatarso-
phalangeal joint area and six patients described a tingling
sensation with walking. Five patients reported tenderness
on the plantar aspect of the metatarsal head area on phys-
ical examination. Two patients exhibited a positive
Mulder sign. Two of 14 cases were diagnosed as granu-
loma initially. The lesions other than the MNs were
located in either the second or third interdigital space
on the dorsal or plantar aspect of the foot. Four patients
complained growing mass, seven patients complained
of painful mass and one patient complained of painful
and growing mass. The others reported palpable mass.
This retrospective study was approved by the institu-
tional ethics review board and patient consents were
obtained.
US evaluation
Ultrasonographic examinations were performed by
a musculoskeletal radiologist using HDI 5000 (Philips,
Bothell, WA, USA) or the Logiq E9 (GE Medical
Systems, Milwaukee, WI, USA) devices equipped with
linear 6–15 MHz probes. Our current protocol for ultra-
sound evaluation of the interdigital mass consists of the
following standardized scanning planes and we retro-
spectively retrieved images from files that contained the
following image planes. The findings were interpreted
in consensus by two fellowship-trained academic muscu-
loskeletal radiologists who had 12 and 10 years of expe-
rience, respectively. All patients are scanned from the
plantar to the dorsal aspect of the foot while in a sitting
position. In some cases pressure is placed on the dorsal
aspect of the foot by the sonographer’s nonimaging finger
to assist in viewing the intermetatarsal space by splaying
the metatarsals. In each case, images are labeled by the
sonographer to identify the intermetatarsal space being
evaluated. Gray-scale images of the soft tissue masses
1205
are obtained in transverse and longitudinal planes.
Lesions with well-defined margins are measured in three
dimensions: length, width and thickness. Color Doppler
sonograms are acquired after modifying the color
window dimensions. Low-flow color Doppler settings
are used in all cases to permit optimal visualization of
small vessels and detection of low-velocity arterial and
venous flow. Pulse repetition frequency is set at 1000
Hz and Doppler gain is adjusted to the maximum level
that does not produce clutter. In all patients, radiography
of the involved segment, including both the anteroposte-
rior and lateral views, is also performed to exclude the
presence of calcification.
Image analysis
The determination of margin characteristics was
made on gray-scale images as either a well-circumscribed
margin or a poorly demarcated lesion. The margin of the
lesion was also classified as either good or poor. When the
distinction between the lesion and the adjacent tissue was
as prominent as a cystic wall and the margin of the lesion
can be drawn by the interpreter without border irregu-
larity, we described it as a good margin. The other cases
were classified as having poor margins. The largest trans-
verse diameter was also recorded. When a mass was iden-
tified, its echogenicity relative to the adjacent tissues was
determined and classified as hypoechoic, hyperechoic,
cystic or mixed echoic. Color Doppler flow images
were analyzed for the presence and locations of vessels.
The locations of the lesions were recorded as follows: if
the lesion was located in shallower than the half thickness
of the subcutaneous fat layer from the skin, the location
was termed ‘‘shallow’’ and if the lesion was located in
deeper than the half thickness of the subcutaneous fat
layer, the location was termed ‘‘deep’’. If the lesion was
in contact with the periosteum of the toe for more than
50% of its transverse length, we defined it as ‘‘abutting
bone’’. We also introduced a new morphologic sign
named ‘‘ginkgo-leaf sign’’. The ginkgo-leaf sign (G-
sign) was defined when the lesion took on a biconcave
shape secondary to compression by adjacent structures,
with or without Mulder’s maneuver, in the coronal plane
(Fig. 1). Mulder’s test was not routinely performed, espe-
cially when a discrete mass was observed. In the Mulder’s
test, a compression of the metatarsal heads against each
other provokes an increase in pain when MN exists and
click is felt. The ultrasonographic findings of the MNs
were compared with those of the other interdigital soft
tissue masses. The same imaging analysis parameters
were applied to the control group. Surgical excision
was performed in all cases and the diagnosis was
confirmed via histopathologic examination of the
specimen.
1206 Ultrasound in Medicine and Biology Volume 37, Number 8, 2011

Fig. 1. Drawing shows a ginkgo leaf sign. Arrows indicate
extrinsic compression from the adjacent soft tissue.
Statistical analysis
Statistical analyses were performed using SPSS
software, version 10.1 (SPSS Inc., Chicago, IL, USA).
Demographic data (age, sex) were compared between
patients with MNs and other interdigital soft tissue
tumors using Fisher’s exact test. Ultrasonographic find-
ings of the MNs were compared with those of other inter-
digital soft tissue masses. We used Student’s t-test in
age and size correlation and used the Wilcoxon-Mann-
Whitney test and Fisher’s exact test in sex correlation,
G-sign, depth, margin, bone abutting and echogenicity
analysis. A p value less than or equal to 0.05 was consid-
ered statistically significant.
RESULTS
A summary of the clinical and imaging findings for
each Morton’s neuroma is shown in Table 1 and
a summary of the control group is shown in Table 2.
The comparison of the MNs and the controls is summa-
rized in Table 3. The MNs exhibited an obvious female
predominance (100%, 10/10, p 5 0.003). The mean
size of the MNs was 5.6 mm, which was significantly
smaller than the mean sizes of other interdigital soft tissue
masses (p 5 0.001). The MNs ranged in size from 3–10
mm. None of the MNs abutted adjacent bony structures
(100%, 14/14, p , 0.001) and all of them were hypoe-
choic (100%, 14/14, p 5 0.001). The G-sign was
observed in 71% of the MNs (10/14, p , 0.001)
(Figs. 2 and 3). The presence of the G-sign was not
affected by Mulder’s maneuver. Of the four patients
with multiple MNs, one patient with lesions in both feet
exhibited a discrepancy in the G-sign. The G-sign was
observed in the right foot, but in the left foot, the mass
had a more rounded appearance and did not demonstrate
the G-sign (Fig. 4). With regard to Mulder’s test (1), in
one case the size of the lesion was largest (10 mm) using
Mulder’s maneuver but in the other case the lesion was
equivalent to the mean size. There were no differences
in incidence related to patient age (p 5 0.259). Some
lesions were well marginated (43%, 6/14) while others
were poorly marginated (57%, 8/14, p 5 0.075). There
were no significant differences with respect to lesion
depth (p 5 0.232). On color Doppler imaging, no signif-
icant vascular flow was seen in the MNs. In the control
group, the epidermoid tumors exhibited variable echoge-
nicities ranging from hypoechoic to hyperechoic and four
cases abutted bony structures (Fig. 5A). The ganglion
cysts demonstrated cystic echogenicities and good
margins; six cases exhibited close contact with bony
structures (Fig. 5B). In the one case of interdigital

Table 1. Summary of the clinical and imaging findings for each Morton’s neuroma
Patient no. Sex Age (y) Location Size (mm) Echogenicity G-sign Abut bone Depth Margin

1 F 52 3rd 3 Hypoechoic N N Deep Good

2 F 52 2nd 5 Hypoechoic N N Deep Good
3 F 42 2nd 7 Hypoechoic P N Deep Good
4 F 65 3rd 6 Hypoechoic P N Deep Good
5 F 28 3rd 10 Hypoechoic N N Deep Poor
6 F 28 3rd 4 Hypoechoic P N Deep Poor
7 F 31 2nd 7 Hypoechoic P N Deep Poor
8 F 35 3rd 6 Hypoechoic P N Deep Poor
9 F 35 3rd 5 Hypoechoic P N Deep Poor
10 F 48 3rd 5 Hypoechoic P N Deep Good
11 F 66 3rd 4 Hypoechoic N N Deep Poor
12 F 48 2nd 5 Hypoechoic P N Deep Good
13 F 61 2nd 6 Hypoechoic P N Deep Poor
14 F 61 3rd 5 Hypoechoic P N Deep Poor

G-sign5 Ginkgo-leaf sign.

 Sonographic appearances of Morton’s neuroma d H.-J. PARK et al. 1207

Table 2. Summary of the clinical and imaging findings for controls

Patient No. sex age (y) diagnosis size (mm) echogenicity G-sign abut bone depth margin

1 F 54 epidermid tumor 5 hypoechoic N N shallow poor

2 M 3 epidermid tumor 5 hypoechoic N N shallow good
3 M 52 epidermid tumor 10 hypoechoic N P deep poor
4 M 46 epidermid tumor 8 hyperechoic N P deep poor
5 F 39 epidermid tumor 6 hypoechoic N N shallow poor
6 M 13 epidermid tumor 22 mixed N P deep good
7 F 34 epidermid tumor 13 mixed N P deep good
8 M 76 fibroma 28 hypoechoic N P deep good
9 F 25 ganglion cyst 8 anechoic N P deep good
10 M 49 ganglion cyst 12 anechoic N N deep good
11 F 45 ganglion cyst 10 anechoic N P deep good
12 M 15 ganglion cyst 14 anechoic N P deep good
13 F 68 ganglion cyst 14 anechoic N P deep good
14 F 51 ganglion cyst 15 anechoic N P deep good
15 F 37 ganglion cyst 9 anechoic N N deep good
16 M 25 ganglion cyst 24 anechoic N N deep good
17 F 40 ganglion cyst 29 anechoic N P deep good

G-sign5 Ginkgo-leaf sign.

fibroma, a large mixed-echoic mass was noted and arte-
rial flow was observed in the center of the lesion on
pulsed Doppler analysis. None of the control cases re-
vealed G-signs. No discernible calcifications were noted
in the interdigital space of the patients.
DISCUSSION
Morton’s neuroma was first described in 1876 by
Thomas Morton. MN is a disorder of the plantar digital
nerve; affected patients are typically middle-aged women
who present with pain and numbness in the forefoot that
is exacerbated by walking and relieved by rest. Women
are more frequently affected than men (Redd et al.
1989). Our study included only female patients with
a mean age at presentation of 46 years. Lassman et al.
(1979) attributed female predominance to the more deli-
cate and pliable nature of women’s feet and the fact that
women tend to wear narrow-toed shoes. Ninety percent of
MNs occur between the heads of the second and third or
third and fourth metatarsals (Beggs et al. 1999). We found
slightly more cases in the third interdigital space and did
not observe any cases in the first or fourth interdigital
spaces.
Quinn et al. (2000) reported the mean width of MNs
to be 6 mm. In our study, the mean width of the lesions
was 5.6 mm. Redd et al. (1989) proposed 5 mm as
a threshold value for symptomatology; however, Pollack
et al. (1992) reported that size and symptomatology are
not necessarily related. In this study, we encountered
three cases of lesions smaller than 5 mm in patients
who endorsed symptoms of pain and paresthesias. The
average size of the control interdigital soft tissue masses
was 14 mm. The mean diameter of the epidermoid tumors
was 9 mm and that of the ganglion cysts was 15 mm. The
average size of the MNs in this study was significantly
smaller compared with these other interdigital soft tissue
masses (p 5 0.001). Lee et al. (2007) suggested that if
a mass in the interdigital space is greater than 20 mm in
length, it is likely an abnormality other than a neuroma,
such as a ganglion cyst, a synovial cyst or a giant cell
tumor of the adjacent tendon sheath. It has also been
hypothesized that MNs may be detected earlier than other
interdigital lesions because of their characteristic neuro-
logic symptoms.
Many reports have described the echogenicity of
MNs (Quinn et al. 2000; Redd et al. 1989). The majority
of MNs is hypoechoic, although some cases exhibit an
echogenicity and mixed echogenicity. No previous
reports have noted any isoechoic MNs and acoustic
enhancement has not been detected. Additionally, the
previous reports suggested well-defined margins in

Table 3. Comparison of the Morton’s neuroma and the controls

Echogenicity
Sex (female) Age Size (mm) (hypoechoic) G-sign (1) Bone abut (1) Depth (shallow) Margin (poor)

MN 10 (10) 46.57 6 13.64 5.57 6 1.69 14 (14) 14 (10) 14 (0) 14 (0) 14 (8)
Control 17 (8) 39.53 6 19.20 13.65 6 7.68 17 (5) 17 (0) 17 (11) 17 (3) 17 (4)
p value 0.003 0.259 0.001 0.001 ,0.001 ,0.001 0.232 0.075

G-sign 5 Ginkgo-leaf; MN 5 Morton’s neuroma.

1208 Ultrasound in Medicine and Biology
Fig. 2. Images from a 61-year-old woman complaining of pain
and a tingling sensation in her foot. (A) Sonogram of the third
intermetatarsal space in the coronal plane (transverse to meta-
tarsal shaft) revealing a hypoechoic mass exhibiting the ginkgo
leaf sign between the hyperechoic metatarsal heads (arrow). The
biconcave portion corresponds to the gingko leaf and the thin
hypoechoic portion connecting to the nerve bundle corresponds
to the stalk of the leaf (small arrow). (B) Photograph of the gross
specimen showing a smooth neuroma (arrow).
MNs. All of the MNs in this study demonstrated hypoe-
chogenicity and similar echotextures. We did not observe
any anechoic or mixed echoic lesions. However, the
marginal natures we observed were different from those
of previous reports. Poor demarcations were noted in
nearly half of the cases. This was not significantly
different from the other interdigital soft tissue masses
(p 5 0.075). As such, we propose that clear demarcation,
such as that seen at the wall of a ganglion cyst, cannot be
used to differentiate MNs from other interdigital soft
tissue masses. However, we do believe that the presence
of a homogeneous hypoechoic echotexture can be helpful
in establishing a diagnosis.
Previous papers have described the shapes of MNs
on ultrasound (Redd et al. 1989; Beggs et al. 1999) and
magnetic resonance imaging (MRI) (Weishaupt et al.
2003). These ultrasound reports suggested ovoid or round
shapes of the lesions while the MRI reports described
pear, dumbbell or inverted pear-shaped lesions. These
differences may have resulted from differences in the
penetration abilities of the study methods. Ultrasound
cannot fully evaluate deeply seated structures such as in-
terdigital bursas, transverse metatarsal ligaments and
opposite surface (plantar or dorsal) structures. We identi-
fied coronal plane constant image patterns of MNs. In the
coronal plane, MNs appear to have a biconcave shape on
Volume 37, Number 8, 2011
Fig. 3. Images from a 65-year-old woman with foot pain. (A)
Sonogram of the third intermetatarsal space in the coronal plane
showing a hypoechoic mass shaped like a ginkgo leaf (arrow).
Note the absence of direct contact with the bone. (B) Photo-
graph of the gross specimen showing a smooth neuroma
(arrow).
the plantar surface that mimics the appearance of a ginkgo
leaf (Fig. 1). We named this image pattern the ‘‘ginkgo-
leaf sign’’ (G-sign). The concave portions were created
by extrinsic compression from the adjacent soft tissue,
making the upper portion of the lesion appear convex.
Thin stalk-like portions may be connected to an interdigi-
tal nerve. This unique pattern was noted in 10 of the 14
MN cases (71%) and was not observed in any of the other
interdigital soft tissue masses (p , 0.001). Additionally,
this pattern was seen regardless of Mulder’s maneuver.
Beggs et al. (1999) reported observing MNs with irreg-
ular shapes that ‘‘spill out’’ in a plantar direction. We
propose that the G-sign is a very specific feature of
MNs; however, because this sign was not noted in all
cases and the size of the patient population was small,
we cannot be overly confident that this sign is pathogno-
monic for MN. Continuing evaluation is required to
further assess the usefulness of this finding.
A large number of the other control interdigital soft
tissue masses that we observed abutted adjacent bony
structures. Specifically, four of the six cases of epider-
moid tumors, six of the nine cases of ganglion cysts,
 Sonographic appearances of Morton’s neuroma d H.-J. PARK et al.
Fig. 4. Images from a 52-year-old woman complaining of foot
pain and a tingling sensation. (A) Sonogram of the second and
third intermetatarsal spaces in the coronal plane reveals multiple
hypoechoic masses (arrow and small arrow). The ginkgo leaf
sign is not visible. (B) Photograph of the gross specimen
showing a prominent neuroma (arrow).
and the sole fibroma case abutted bone; in contrast, none
of the MNs abutted bony structures (p , 0.001). The
cause for this is not certain. MNs arise from digital
branches of the medial or lateral plantar nerves, whereas
other soft tissue tumors, especially ganglion cysts, may
arise from the periosteum and, therefore, closely contact
bony structures. We believe this finding can be helpful in
excluding a diagnosis of MN.
The limitations of our study include the small
number of cases, the retrospective analytic nature, and
the non-blinded study design. And we did not applied
power Doppler, which could increase the sensitivity for
slower flow. Despite these limitations, we believe these
results can help in correctly diagnosing MNs and differ-
entiating them from other interdigital soft tissue masses.
In conclusion, US is a very useful imaging modality
in patients with MNs. Interdigital MNs are found
predominantly in middle-aged women and often exhibit
a G-sign on ultrasonography. All MNs are hypoechoic
and they do not abut adjacent bony structures. These
1209
Fig. 5. Other interdigital soft tissue masses. (A) An epidermoid
tumor in a 34-year-old woman. Note the close contact with the
bone (arrow) and the absence of the ginkgo-leaf sign. (B) A
ganglion cyst in a 25-year-old man. A round anechoic lesion
is observed (arrow).
characteristics can be helpful in differentiating MNs
from other interdigital soft tissue masses.
REFERENCES
Beggs I. Sonographic appearances of nerve tumors. J Clin Ultrasound
1999;27:363–368.
Hughes RJ, Ali K, Jones H, Kendall S, Connell DA. Treatment of Mor-
ton’s neuroma with alcohol injection under sonographic guidance:
Follow-up of 101 cases. AJR 2007;188:1535–1539.
Lassman G. Morton’s toe: Clinical, light and electron microscopic
investigations in 133 cases. Clin Orthop 1979;142:73–84.
Lee MJ, Kim SJ, Huh YM, Song HT, Lee SA, Lee JW, Suh JS. Morton
neuroma: Evaluation with ultrasonography and MR imaging.
Korean J Radiol 2007;8:148–155.
Murphy P. Ultrasound diagnosis of Morton’s neuroma. ASUM Ultra-
sound bulletin 2004;2:21–25.
Pollack R, Bellacosa R, Dornbluth NC, Strach WW, Devall JM. Sono-
graphic analysis of Morton’s neuroma. J Foot Surg 1992;31:
534–537.
Quinn TJ, Jacobson JA, Crig JG, van Hosbeeck T. Sonography of Mor-
ton’s neuromas. AJR 2000;174:1723–1728.
Redd RA, Peters VJ, Emery SF, Branch HM, Rifkin MD. Morton
neuroma: Sonographic evaluation. Radiology 1989;171:415–417.
Weishaupt D, Treiber K, Kundert HP, Zollinger H, Vienne P, Holder J,
Willmann JK, Marincek B, Zenetti M. Morton neuroma: MR
imaging in prone, supine, and upright weight-bearing body posi-
tions. Radiology 2003;226:849–856.