European Journal of Radiology 85 (2016) 1750–1756

Contents lists available at ScienceDirect

European Journal of Radiology

journal homepage: www.elsevier.com/locate/ejrad

Clinical evaluation vs magnetic resonance imaging of the skier’s

thumb: A prospective cohort of 30 patients
Mandhkani Mahajan a,∗ , Christine Tolman b , B. Würth c , Steven J. Rhemrev (Dr.) a
a
Medical Center Haaglanden, Department of Surgery, Lijnbaan 32, 2512 VA The Hague, The Netherlands
b
Medical Center Haaglanden, Department of Radiology, Lijnbaan 32, 2512 VA The Hague, The Netherlands
c
Medical Center Haaglanden, Department of Emergency Care, Lijnbaan 32, 2512 VA The Hague, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: Introduction: A skiers thumb, or a partial or complete rupture of the ulnar collateral ligament (UCL) is
Received 2 February 2016 a clinical diagnosis. Swelling, pain, natural left-right difference and inexperience of a young physician
Received in revised form 5 July 2016 can cause difficulty to correctly diagnose this injury. However, our theory is that any physician, given
Accepted 14 July 2016
the correct instructions, should be able to diagnose this injury solely on clinical findings, without the
necessity of additional imaging.
Keywords:
Material and methods: In a large Dutch teaching hospital, physicians (residents with working experience
Skier’s thumb
of 6 months–3 years) working at the ER received instructions for physical examination. Patients >18 years,
Ulnar collateral ligament
MR
with an injury <1 week old, suspected of a true skier’s thumb had an MRI reported by two independent
Sports injury radiologists to confirm the diagnosis.
Results: Thirty patients were included. Seven patients had no fixed endpoint (23%), all had a complete
ligamentous rupture of the UCL on MRI, of which three patients had a Stener lesion. Fifteen patients
(50%) met with the criteria >35◦ laxity in extension of MCP/ >20◦ laxity in 30◦ flexion of the MCP. Of
these, thirteen patients (81%) had a complete rupture (nine Stener lesions (56%)). One patient had a
partial injury and one patient had no UCL-injury. Eight patients (27%) had inconclusive results during
physical examination. Of these, two had a complete rupture (40%, 1 Stener). Three patients had a partial
rupture and three patients had an intact UCL.
Conclusion: A skier’s thumb can be diagnosed by any resident when correctly instructed. Additional imag-
ing when diagnosing a skier’s thumb should be reserved in cases when physical examination remains
inconclusive.
© 2016 Elsevier Ireland Ltd. All rights reserved.

1. Introduction treatment. A partial injury is treated conservatively with only

A skier’s thumb is an acute rupture of the ulnar collateral lig-
ament of the metacarpophalangeal joint of the thumb. It occurs
when a thumb, which is already in abduction, receives an extra
valgus stress. The classic trauma mechanism is when a skier falls
down while holding on to its skiing pole. However, this injury can
also occur with people playing sports involving a ball or stick, falling
from their bicycle when holding on the handlebars, or simply falling
on an outstretched thumb. The estimated incidence of this injury
in the United States is approximately 200.000 patients per year [1].
The injury can be a partial or a complete rupture, the differ-
ence of which is important when determining the appropriate
∗ Corresponding author.
E-mail addresses: mahajanmch@gmail.com
(M. Mahajan), c.tolman@mchaaglanden.nl (C. Tolman), b.wurth@mchaaglanden.nl
(B. Würth), s.rhemrev@mchaaglanden.nl (S.J. Rhemrev).
cast immobilization; a complete injury is usually treated by sur-
gical repair combined with cast immobilization. Therefore, the
difference between partial and complete rupture is an important
distinction.
The general thought is that the diagnosis of a (partial or com-
plete) skier’s thumb can be made using only patient history and
clinical examination. As aforementioned, patient history will usu-
ally involve a hyperabduction trauma. They will also present with
pain, swelling and hematoma on the ulnar side of the metacar-
pophalangeal (MCP) joint of the thumb. Sometimes a mass can be
felt, suggesting a Stener lesion, which is a subtype of complete
UCL-rupture with the ruptured ligament interpositioned under-
neath the adductor aponeurosis [Fig. 1]. However, this mass is not
a pathognomonic finding [2].

http://dx.doi.org/10.1016/j.ejrad.2016.07.007
0720-048X/© 2016 Elsevier Ireland Ltd. All rights reserved.
 M. Mahajan et al. / European Journal of Radiology 85 (2016) 1750–1756 1751

Fig. 1. Complete rupture of the ulnar collateral ligament, with the ruptured distal
end stuck underneath the adductor aponeurosis, the so-called ‘Stener lesion’.

1.1. Anatomy

The ulnar collateral ligament of the thumb consists of two parts,

the proper and the accessory (pUCL and aUCL). Both have their ori-
gin on the head of the first metacarpal bone and its insertion on the
lateral tubercle of the proximal phalanx. The aUCL runs more pal-
mar and more superficial compared to the pUCL. The pUCL is taut
in flexion of the MCP and the aUCL in extension.

1.2. Physical examination

Because the UCL comprises of two parts, the thumb must also be
tested in two positions. First, the patient’s thumb is held with the
MCP joint in extension, while applying valgus stress to the thumb.
Then, the test is repeated with the MCP joint in 30◦ of flexion. Laxity
of the joint is noted in these two positions. Care must be taken by
the investigator to place his/her thumb on the radial side of the
MCP joint to prevent rotational effects. Literature varies about what
degree of laxity warrants the diagnosis of a complete rupture, but
a standard of more than 35◦ during valgus stress with the MCP
in extension, more than 20◦ in extension and/or more than a 15◦
difference compared to the contralateral side [2–4] is commonly
agreed upon. An alternative measurement could be the presence
or absence of a firm endpoint during testing [2].
1.3. Limitations in physical examination
There are some problems in determining the correct diagnosis.
First, the knowledge of how to correctly perform and interpret the
physical examination often lacks in (young) physicians working at
Fig. 2. a–d: Complete UCL rupture without dislocation; the fibres of the UCL are
discontinuous, but the aponeurosis of the adductor pollicis longus appears normal.
a: T1W cor, b: PD TSE cor, c: T2 TSE cor, d: T2 FS cor.
the emergency department (ER). In non-skiing countries such as
the Netherlands, this injury is relatively uncommon, which could
explain why it often is not recognized in acute setting. In our hospi-
tal, we observed an incidence of approximately 20 patients a year
with a complete rupture of the UCL.
Lastly, there are limitations to the clinical investigation in acute
setting. A ‘firm endpoint’ seems logical however can still be unclear
to someone who has less experience with testing injuries of the
thumb. Also, measuring the degree of laxity of the thumb is diffi-
cult and prone to inter-observer variation, even with a goniometer.
A recent article has shown that when the examination is performed
incorrectly − with the MCP joint in slight pro- or supination − the
diagnosis can be misinterpreted [5]. Also, in approximately one
1752 M. Mahajan et al. / European Journal of Radiology 85 (2016) 1750–1756
Fig. 3. a–d: Complete UCL rupture with retraction of aponeurosis underneath the
UCL, thereby creating the typical yo-yo-on-a-string appearance, the so-called Stener
lesion. a: T1W cor, b: PD TSE cor, c: T2 TSE cor, d: T2 FS cor.
third of population, a natural variance between left and right is
present, with a difference in laxity of about 10 to sometimes 15◦ ,
which can incorrectly lead to the diagnosis of a UCL rupture [6].
Finally, clinical examination can be hindered due to swelling and
pain. The latter can be overcome by using Oberst anaesthesia, which
has shown in a study to enhance the clinical accuracy from 28% to
98% (one week after initial trauma) [7]. When swelling hinders the
physical exam, the patient can be given cast immobilization and the
thumb can be tested a few days later when most of the swelling has
subsided.
Fig. 4. a,b: Partial injury of the UCL, showing a subtle enhanced T2 signal in the T2
FS, indicating edema in the ligament. a: PD TSE cor, b: T2 FS cor.
All of the above leads to the question whether a skier’s thumb
really is a clinical diagnosis or that additional imaging is necessary
to come to the correct conclusion. The objective of this study was to
answer this question by determining the positive predictive value
of physical examination compared to the accuracy of MR imaging.
1.4. Additional imaging
A radiograph of the thumb is needed in all cases to diagnose
a bony skier’s thumb, an avulsion injury. If the radiograph does
not show a fracture, further imaging such as stress X-rays, CT (-
arthrogram), MRI or ultrasound can be used. Stress radiographs
have shown a 25% false negative result in previous studies [8],
which can partly be explained by a poorly cooperating patient with
a painful thumb, and the lack of clear cut-off points to determine
the diagnosis. MRI can be seen as the ‘gold’ standard with a reported
sensitivity of 96%-100% and a specificity of 95%-100% [8–11].
Both a tear and a displaced UCL can be identified on MRI. A
non-displaced torn UCL presents itself as a gap where the UCL
would normally be present, showing a complete discontinuity of
the ligament (Fig. 2a–d). Underneath the aponeurosis of the adduc-
tor pollicis muscle, the UCL is usually thickened due to associated
retraction of the torn fibres [12,13]. Both ligament and aponeurosis
adductor pollicis muscle follow the normal low signal intensity on
T2 weighted MR images. In acute setting bone marrow edema and
hematoma can be seen as subsequently high T2 signal intensity in
the bone, acute haemorrhage can be seen as focal high T1 signal
intensity. Sometimes a joint effusion is associated [13].
A completely ruptured UCL that is displaced underneath the
aponeurosis adductor muscle is called a Stener lesion. This Stener
lesion appears as a rounded or stump like area of low signal inten-
sity located more superficially than a normal ligament. This creates
the appearance of a ‘yo–yo on a string’ on the coronal images, in
which the yo–yo is the retracted, rolled up UCL, and the string is the
 M. Mahajan et al. / European Journal of Radiology 85 (2016) 1750–1756 1753

Fig. 5. pocket card with instructions on how to diagnose a skier’s thumb.

adductor aponeurosis [Fig. 3a–d] [12]. The aponeurosis, which nor- In this study, we chose to compare the clinical diagnosis of a
mally appears as a thin hypointense structure, usually also shows skier’s thumb to the diagnosis on MRI, the last being our ‘gold
surrounding hyperintense effusion on T2-weighted MR images standard’.
[12].
Sometimes, a small injury without a complete rupture of the UCL
can be seen. However, this is very subtle on MRI and can only be
seen as a focally enhanced T2 signal on the T2 FS, indicating oedema
2. Objectives
in the ligament [Fig. 4ab]. Combined with the clinical finding of an
acute injury with a painful UCL, this could be classified as a partial
To determine the positive predictive value of the clinical diag-
injury.
nosis of a skier’s thumb compared to the accuracy of MRI of the
UCL.
1754 M. Mahajan et al. / European Journal of Radiology 85 (2016) 1750–1756

Table 1
Diagnosis of physical examination compared to diagnosis on MRI.

Findings at physical examination No. of patients Complete rupture/Stener lesion Partial injury No UCL injury

No firm end point 7 (23%) 7 (100%)/3 (43%) 0 (0%) 0 (0%)

>35◦ in extension >20◦ in flexion 15 (50%) 13 (86%)/9 (56%) 1 (7%) 1 (7%)
Inconclusive 8 (27%) 2 (25%)/1 (13%) 3 (38%) 3 (38%)
Total 30 22/13 4 4

Table 2
Inter-rater agreement of the radiologists.

Radiologist B

Radiologist A Complete UCL rupture Partial UCL rupture No UCL injury

Complete UCL 21 1 0 22
rupture
Partial UCL rupture 0 3 1 4
No UCL injury 0 0 4 4
Total 21 4 5 30

3. Methods reproducibility and interobserver variation in our results. As the

operation of a Stener lesion according to our protocol and referring
In a large teaching hospital in the Netherlands, a prospective to previous literature had to be performed within 5 days, perform-
cohort study was founded. From the 1st of October 2013 till the 1st ing MRI did not lead to unnecessary patient or doctor’s delay.
of October 2015, all patients presenting in both the ER and outpa-
tients clinic of Surgery and Orthopedics with a clinical suggestion
of a complete UCL tear were included, after giving informed con-
4. Results
sent. Up till now, thirty patients have been included. Participating
physicians were defined as residents in Emergency medicine, Gen-
The five-step plan was clear and comprehensible after careful
eral Surgery or Orthopaedic Surgery, varying from 1st to 5th year
instructions to the performing physicians and was easily repeated
residents, who worked in the ER and outpatients clinic during their
and taught to colleagues by using hand cards and posters.
shifts. To ensure sufficient knowledge of the relatively uncommon
Twenty-two patients (73%) had a clear diagnosis of a com-
UCL injury, a clear 5-step plan to clinically diagnose a skier’s thumb
plete UCL rupture at clinical examination. Of this group, only seven
with additional background information was presented in advance.
patients were found not to have a firm end point. All these patients
During this year, new residents and ER doctors were continuously
were shown to have a total rupture of the ligament on MR imaging,
trained in diagnosing UCL tears, by repeated lectures and hands-
three of them (43%) had a Stener lesion [Table 1].
on sessions. Pocket cards and posters with clear instructions were
Fifteen of the 22 patients (50%) fulfilled the criteria of a skier’s
available in the ER and outpatient clinics during the whole course
thumb as defined by the rule >35◦ laxity in extension and >20◦ in
of the study [Fig. 5]. First, a radiograph of the thumb was made to
flexion. Of this group, 13 patients (86%) had a completely ruptured
exclude osseous avulsion fractures. When the radiograph was neg-
UCL on MRI, 9 of those patients had a Stener lesion. One patient (7%)
ative, the thumb was tested using the 5-step plan. When the patient
was diagnosed on MRI as having a partial rupture, implying injury
was suspected of having a skier’s thumb, i.e. a complete rupture of
of the UCL but with an intact distal part attaching to the proximal
the UCL, the patient would receive a cast and an MRI was scheduled
phalanx of the first digit. One patient had an intact UCL on MRI.
within 3 days.
A smaller group of patients (8 patients, 27%) were found to have
MR imaging of the finger was performed in both axial and coro-
inconclusive results, either because of too much swelling during
nal planes on a Siemens Avanto 1.5 T MRI (Software MR B19), with
clinical investigation, or because of the results were ambiguous due
a small flex coil. The protocol we used was distracted from former
to a bilateral (mild) degree of laxity. All patients had pain on the
studies [8–13]. It consisted of an axial T1 W TSE (repetition time
ulnar side of the MCP joint after a hyperabduction trauma. Of these
msec/echo time; msec = 414/11, slice thickness 2.5 mm), coronal
patients, 2 patients (40%) had a complete UCL rupture and 1 had a
T1 TSE (592/11, slice thickness 2.5 mm), coronal PD TSE (2500/39,
Stener lesion. Three patients only had a partial injury on MRI and 3
slice thickness 2.5 mm), coronal T2 TSE (2750/65, slice thickness
patients had an intact UCL.
2.5 mm), and coronal T2 FS sequences (2750/65, slice thickness
The majority of patients that were clinically and radiologically
2.5 mm). The images were obtained with an 11 cm field of view
diagnosed with a complete rupture of the UCL, received an oper-
and in all sequences using a 256/320 × 192/256 acquisition matrix.
ative repair of the ligament. During the operation, in all patients
Total scanning time was less than 16 min.
the diagnosis of a complete rupture was confirmed. Two patients
Two specialised musculoskeletal radiologists, both with over 20
refused surgery. The last patients that were included also partici-
years of radiology experience, and who were responsible for all MR
pated in a currently running randomised controlled trial (UCL-trial),
imaging in this hospital, blindly assessed the MRI’s within two days.
in which conservative treatment is compared to operative repair of
At the time of viewing the case, they were both unaware neither of
a complete ligament rupture of the UCL. Four patients that were
the diagnosis yielded by the other reader nor of the clinical diag-
included were randomly assigned to the conservative group, and
nosis. This was secured by the radiology resident responsible for
therefore were not operated.
this study. The resident would only show the MRI series with the
The inter-rater agreement of the two radiologists that reviewed
clinical question “UCL rupture? Stener lesion?” to the viewing radi-
the MRIs was calculated by Cohen’s kappa coefficient and was 85%
ologists. Both musculoskeletal radiologists reliably established the
(very good) [Table 2]. This suggests that the assessment of MRI
diagnosis by using their experience and, if necessary, the available
scans can easily be reproduced, meaning this injury is not difficult
literature. The inter-rater observer ratio was calculated to show the
to identify.
 M. Mahajan et al. / European Journal of Radiology 85 (2016) 1750–1756
5. Discussion
Most patients could correctly be diagnosed to have a true skiers
thumb, i.e. a complete rupture of the UCL, by patient history and
clinical examination alone. These patients did not benefit from
additional MR imaging. A Stener lesion could not be diagnosed
in advance by the physicians; however, treatment for all unstable
MCP joints, including both complete, non-displaced UCL rupture
and complete, displaced UCL rupture (Stener) was the same. One
could however argue that because a non-displaced rupture could
also be treated conservatively, an MRI would be valuable in identi-
fying these patients in order to avoid unnecessary operations. We
hope that the aforementioned UCL trial will further aid clinicians
in making this decision.
A small part of the examined patients (27%) had an inconclusive
clinical examination, of which 5 (63%) had a positive MRI, showing
either partial or complete UCL injury and of 3 (38%) had no ligament
tear. In these cases, additional imaging could be of value.
The question does arise why some patients have inconclusive
findings at clinical examination. Do the physicians need more expe-
rience; is this part of a learning curve? However, in the acute phase,
swelling can obscure conclusive results. One option would be to
re-examine the patient after 24–48 h after the oedema and haem-
orrhage has subsided. Also, some people have a uni- or bilateral
natural laxity that makes it hard to distinguish from an acute injury.
To find an answer to the learning curve question, more patients
need to be examined. This study will continue for a longer period
of time and in more hospitals to see whether our results will differ.
Previous literature also notes the difficulty in physical examina-
tion due to above-mentioned reasons. Some experts have therefore
suggested using the criterion of the loss of a firm endpoint to diag-
nose a complete rupture of the UCL [2]. In our study however, only
approximately one fourth of all patients that had a total rupture of
the UCL exhibited a lack of a firm end point during clinical examina-
tion. Therefore, it stands to wonder whether in practise this could
be a useful criterion.
Of importance as well is that all patients were systematically
examined in the same way, contrary to other studies that did not
describe by whom and how the patients were examined. Also, some
studies describe outdated techniques such as stress radiographs
and arthrography, which are no longer up to date and are consid-
ered by many to be too invasive.
For this study, physicians were chosen who had less experience
with identifying and treatment of this injury than fully trained hand
surgeons. This was done for two reasons. First, because we think
that every intern or resident in surgery, orthopaedics or emergence
care should be able to diagnose this injury. Secondly, we wanted
to mimic a real life situation. Normally, a patient with this type
of injury will go to the emergency department and is investigated
by a junior physician with less experience. This is also the reason
why the patients were not re-examined by an experienced surgeon,
because this in reality this rarely happens. This makes our study a
true reflection of the current situation and therefore its results are
useful to implement in daily practise.
In this study, we used MR imaging, which has previously shown
excellent sensitivity and specificity. Another type of imaging that is
gaining more popularity is ultrasound. In the last years, ultrasound
(US) has proved itself as a cost-effective, cheap, fast and widely
available alternative; besides, US can produce the same results
as MRI. A recent study has shown that the ultrasound finding of
absence of normal UCL fibres and presence of a heterogeneous mass
proximal to the MCP joint is 100% accurate when diagnosing a com-
plete UCL rupture [14]. When diagnosing a Stener lesion, the same
radiological features as seen on MR imaging (‘yo–yo effect’) can be
identified with ultrasonography [15,16].
1755
However, some downsides exist when using ultrasound. Most
importantly, the quality of the equipment should be adequate and
the radiologist performing the ultrasound should have sufficient
experience. Besides that, in subacute injuries older than a week,
when swelling of the ligament has subsided and naturally occurring
retraction due to the formation of fibrosis can be misinterpreted as
a Stener lesion. Other conditions such as a dorsal hood tear, tendon
tear, joint effusion or hematoma can also mimic findings of an UCL
injury [17].
We did not use ultrasound in our study because reproducibility
of the results was essential and could not be guaranteed when using
ultrasound.
The last issue that needs to be addressed is that the design of
this study could lead to a selection bias, because only those patients
suspected to have a true skiers thumb at clinical examination were
included in the study. Therefore, we were only able to calculate
the positive predictive value and not the sensitivity and speci-
ficity, as patients without any laxity on clinical examination were
not included, therefore a skiers thumb due to insufficient clinical
examination could have been missed. However, chances of having
an acute UCL injury with a normal range of motion are very slim,
and do not outweigh the time and costs of unnecessary and costly
additional imaging.
6. Conclusion
Our study suggests that patient history combined with clin-
ical examination of a skier’s thumb performed by a correctly
instructed resident should be sufficient to diagnose a true skiers
thumb. Additional imaging such as Magnetic Resonance Imaging is
very accurate with excellent inter-rater agreement but should be
reserved for cases in which clinical examination remains inconclu-
sive.
References
[1] D.H. Hinke, S.J. Erickson, L. Chamoy, et al., Ulnar collateral ligament of the
thumb: MR findings in cadavers volunteers, and patients with ligamentous
injury (gamekeeper’s thumb), AJR Am. J. Roentgenol. 163 (1994) 1431–1434.
[2] A.W. Ritting, P.C. Baldwin, C.M. Rodner, Ulnar collateral ligament injury of the
thumb metacarpophalangeal joint, Clin. J. Sport Med. 20 (2010) 106–112.
[3] P.C. Rhee, D.B. Jones, S. Kakar, Management of thumb metacarpophalangeal
ulnar collateral ligament injuries, J. Bone Joint Surg. Am. 94 (2012)
2005–2012.
[4] P. Heyman, R.H. Gelberman, K. Duncan, et al., Injuries of the ulnar collateral
ligament of the thumb metacarpophalangeal joint: biomechanical and
prospective clinical studies on the usefulness of valgus stress testing, Clin.
Orthop. Relat. Res. 292 (1993) 165–171.
[5] S.W. Mayer, D.S. Ruch, F.J. Leversedge, The influence of thumb
metacarpophalangeal joint rotation on the evaluation of ulnar collateral
ligament injuries: a biomechanical study in a cadaver model, J. Hand Surg.
Am. 39 (2014) 474–479.
[6] A.K. Malik, T. Morris, D. Chou, et al., Clinical testing of ulnar collateral
ligament injuries of the thumb, J. Hand Surg. Eur. 34 (2009) 363–366.
[7] J.G. Cooper, A.J. Johnstone, P. Hider, et al., Local anesthetic infiltration
increases the accuracy of assessment of ulnar collateral ligament injuries,
Emerg. Med. Australas. 17 (2005) 132–136.
[8] M.T. Harper, V.P. Chandnani, J. Spaeth, et al., Gamekeeper thumb: diagnosis of
ulnar collateral ligament injury using magnetic resonance imaging, magnetic
resonance arthrography and stress radiography, J. Magn. Reson. Imaging 6
(1996) 322–328.
[9] N. Ahn, D.J. Sartoris, Gamekeeper thumb: comparison of MR arthrography
with conventional arthrography and MR imaging in cadavers, Radiology 206
(1998) 737–744.
[10] K. Hergan, C. Mittler, W. Oser, Ulnar collateral ligament: differentiation of
displaced and nondisplaced tears with US and MR imaging, Radiology 194
(1995) 65–71.
[11] K.D. Plancher, Role of MR imaging in the management of skier’s thumb
injuries, Magn. Reson. Imaging Clin. N. Am. 7 (1999) 73–84.
[12] J.A. Clavero, MR imaging of ligament and tendon injuries of the fingers,
Radiographics 22 (2002) 237–256.
[13] J.J. Peterson, Evaluation of collateral ligament injuries of the
metacarpophalangeal joints with magnetic resonance imaging and magnetic
resonance arthrography, Curr. Probl. Diagn. Radiol. 36 (2007) 11–20.
1756 M. Mahajan et al. / European Journal of Radiology 85 (2016) 1750–1756

[14] D. Melville, J.A. Jacobson, S. Haase, C. Brandon, M.K. Brigido, D. Fessell,
Ultrasound of displaced ulnar collateral ligament tears of the thumb: the
Stener lesion revisited, Skeletal Radiol. 42 (2013) 667–673.
[15] F.S. Ebrahim, US diagnosis of UCL tears of the thumb and stener lesions:
technique: pattern-based approach and differential diagnosis, Radiographics
26 (2006) 1007–1020.
[16] D.M. Melville, J.A. Jacobson, D.P. Fessell, Ultrasound of the thumb ulnar
collateral ligament: technique and pathology, AJR Am. J. Roentgenol. 202
(2014) W168.
[17] C.F. Arend, T.R. da Silva, The role of US in the evaluation of clinically suspected
ulnar collateral ligament injuries of the thumb: spectrum of findings and
differential diagnosis, Acta Radiol. 55 (2014) 814–823.