CLINICAL ORTHOPAEDICS AND RELATED RESEARCH
Number 445, pp. 181–185
Comparison of Six Radiographic Projections to Assess
Femoral Head/Neck Asphericity
Dominik C. Meyer, MD*; Martin Beck, MD†; Tom Ellis, MD‡; Reinhold Ganz, MD†; and
Michael Leunig, MD†
Early radiographic detection of femoroacetabular impinge-
ment might prevent initiation and progression of osteoarthri-
Downloaded from http://journals.lww.com/clinorthop by BhDMf5ePHKbH4TTImqenVD7nNEO3Od0IdCalA4EhmYfYlJzfddAqJQ/o/9OBL5Ea on 06/25/2018
tis. The structural abnormality in femoral-induced femoro-
acetabular impingement (cam type) is frequently asphericity
at the anterosuperior head/neck contour. To determine
which of six radiographic projections (anteroposterior,
Dunn, Dunn/45° flexion, cross-table/15° internal rotation,
cross-table/neutral rotation, and cross-table/15° external ro-
tation) best identifies femoral head/neck asphericity, we
studied 21 desiccated femurs; 11 with an aspherical femoral
head/neck contour and 10 with a spherical femoral
head/neck contour. To radiographically quantify femoral
head asphericity, we measured the angle where the femoral
head/neck leaves sphericity (angle alpha). The aspherical
femoral head/neck contours had a greater maximum angle
alpha (70°) compared with the spherical head/neck contours
(50°). The angle alpha varied depending on the radiographic
projection: it was greatest in the Dunn view with 45° hip
flexion (71° ± 10°) and least in the cross-table view in 15°
external rotation (51° ± 7°). Diagnosis of a pathologic femo-
ral head/neck contour depends on the radiologic projection.
The Dunn view in 45° or 90° flexion or a cross-table projec-
tion in internal rotation best show femoral head/neck asphe-
ricity, whereas anteroposterior or externally rotated cross-
table views are likely to miss asphericity.
Level of Evidence: Prognostic study, level II-1 (retrospective
study). See the Guidelines for Authors for a complete de-
scription of levels of evidence.
Received: December 29, 2004
Revised: June 1, 2005; November 4, 2005
Accepted: November 14, 2005
From the *Department of Orthopaedic Surgery, Balgrist University Hospital,
Zurich, Switzerland; the †Department of Orthopaedic Surgery, Inselspital,
University of Berne, Berne, Switzerland; and the ‡Department of Orthopedic
Surgery, Oregon Health & Sciences University, Portland, Oregon.
Each author certifies that he or she has no commercial associations (eg,
consultancies, stock ownership, equity interest, patent/licensing arrange-
ments, etc.) that might pose a conflict of interest in connection with the
submitted article.
Correspondence to: Dominik Meyer, MD, Uniklinik Balgrist, Forchstr. 340,
CH-8008 Zürich, Switzerland. Phone: 0041-1-386-11-11; Fax: 0041-1-386-
16-09; E-mail: dominikmeyer@bluewin.ch.
DOI: 10.1097/01.blo.0000201168.72388.24
181
© 2006 Lippincott Williams & Wilkins
The exact mechanism for idiopathic osteoarthritis (OA) of
the hip is not known. There is evidence that motion-
induced femoroacetabular impingement might initiate a
progressive degenerative process leading to early OA of
the hip.8,9 Acetabular and femoral abnormalities leading to
contact between the proximal femur and the acetabular rim
can contribute to the development of femoroacetabular
impingement. 3 Reduced joint clearance is a well-
appreciated cause of repetitive contact between the pros-
thetic femoral neck and the edge of the acetabular com-
ponent in a malpositioned total hip arthroplasty.6 The na-
tive hip is under even tighter constraint, and any contact or
shearing forces arising from femoroacetabular impingment
can lead to labral, and even more harmful, chondral lesions
as the hip cannot escape these forces.3 In femoral-induced
femoroacetabular impingment (cam type),5 an aspherical
femoral head/neck contour caused by widening of the
femoral neck or a reduction in the head/neck offset reduces
joint clearance. This causes repetitive impingement be-
tween the proximal femur and the acetabular rim.6 The
shape and orientation of the acetabulum and the degree of
excursion of the femoral neck during hip flexion and in-
ternal rotation determines whether femoroacetabular con-
tact occurs at the acetabular rim.
The concept that femoral head/neck deviations lead to
OA of the hip is not new. Stulberg et al reported abnormal
head/neck configurations of the proximal femur on antero-
posterior (AP) radiographs (pistol grip deformity) in 40%
of patients with idiopathic OA.20 An abnormal anatomic
relationship between the femoral head and neck also is a
possible cause of OA.14,16,19,22 Subclinical displacement
of the femoral epiphysis has been reported as a risk factor
for OA, and the terms head-tilt or post-slip have been used
to describe the deformity resulting from a mild slipped
capital femoral epiphysis (SCFE).4,5,14,20 Some small se-
ries8,9 provide evidence to support femoral-induced (cam
type) femoroacetabular impingment in SCFE.16 Similar
deformities may occur from malunited femoral neck frac-
tures,1 from morphologic deviations including residual

182 Meyer et al
childhood diseases such as Legg-Calvé-Perthes disease,
and from surgical interventions such as femoral osteoto-
mies leading to reduction of joint clearance.2,18 The ma-
jority of patients with femoroacetabular impingement lack
a history of detected predisposing factors.3
Anteroposterior radiographs of the pelvis are the most
widely used for initial radiographic assessment of unex-
plained hip pain These radiographs often appear normal
during the early stages of OA according to classic radio-
logic criteria.10 However, closer examination may show
morphologic abnormalities. The most common abnormal-
ity is a bony prominence at the anterosuperior head/neck
junction that frequently is best seen on the lateral projec-
tion.1 Because the aspherical portion of the femoral
head/neck contour often is located anterosuperiorly, it
might be missed when using AP and lateral views.5 To
date, there is no conclusive information on the ideal posi-
tion for detecting femoral head/neck sphericity/asphericity
in suspected femoroacetabular impingement.
We questioned which of six standard radiographic pro-
jections would be best to evaluate femoral head/neck
asphericity in normal and pathologic femurs. We also
questioned whether the cross-sectional shape of the femo-
ral neck could explain possible variations in exposure of
the bony prominence on the femoral head/neck junction in
different radiographic projections.
MATERIALS AND METHODS
Twenty-one intact femurs were included from a collection of 30
desiccated skeletons with no signs of OA (osteophytes and cysts)
or sequelae of childhood disease (Legg-Calvé-Perthes disease,
proximal femoral focal deficiency, SCFE). Femurs were pro-
vided from the skeletal collection of the Institute of Anatomy,
University of Berne, Berne, Switzerland. Before making radio-
graphic measurements, the desiccated femurs were categorized
into two groups (spherical and aspherical femoral head/neck
junctions) by a spherical template that matched the size of femo-
ral heads.7 A spherical femoral head/neck junction was present
when there was normal transition of the head into the neck and
normal waist of the femoral neck, not leading to displacement of
the spherical template when approaching the head/neck junction
of the desiccated specimens. An aspherical femoral head/neck
junction showed lift-off of the spherical template7 when ap-
proaching the femoral head/neck junction, indicating anterosu-
perior asphericity from a locally decreased waist of the femoral
neck. There were 11 pathologic (aspherical) and 10 normal
(spherical) femoral head/neck junctions.
Radiographs were taken in six different standard radiographic
projections7: (1) plain AP view in neutral rotation of the femur
(AP); (2) cross-table lateral view in 15° internal rotation of the
femur (Lat IR); (3) cross-table lateral view in neutral (0°) rota-
tion of the femur (Lat NR); (4) cross-table lateral view in 15°
external rotation of the femur (Lat ER); (5) plain AP Dunn view
in 90° hip flexion, neutral rotation, 20° abduction (Dunn); and
Clinical Orthopaedics
and Related Research
(6) Dunn view in 45° hip flexion, neutral rotation, 20° abduction
(Dunn/45°). For the lateral cross-table view, the central xray
beam entered the pelvis of the patient, who was positioned su-
pine, horizontally at 45° from the contralateral side from distally,
with the examined leg extended and rotated. The contralateral
leg was held in flexion to avoid interference with the beam.
Rotation of the femur was determined by the angle between
a line parallel to the posterior aspect of the femoral condyles and
the radiograph table (table-top method).13 Flexion was deter-
mined by the angle between the mechanical long axis of the bone
and the radiograph table.
The anterior offset angle alpha11 was measured on each ra-
diograph to identify anterosuperior asphericity at the femoral
head/neck junction. Alpha represented the angle formed by a line
between the center of the femoral head and the center of the
femoral neck, and the line between the center of the femoral head
and Point A. Point A represented the point at the femoral
head/neck contour where the radius of the femoral head diverged
from the femoral neck (Fig 1).15 Such loss of femoral head
sphericity usually is found on the anterosuperior aspect of the
head/neck junction5 (Fig 2).
The cross-sectional shape of the femoral neck was flat and
oval (Fig 3).5 The radiographic projection that best showed the
smallest offset was obtained when the largest diameter of the
flattened femoral neck was parallel to the xray plate. We mea-
sured the rotational orientation of the greatest femoral neck (␳)
diameter around the long axis of the femoral neck relative to the
sagittal plane of the femur (Fig 3) using a caliper. Anteversion of
the femoral neck was measured in each femur relative to the
posterior aspect of the femoral condyles.5
All radiographic and anatomic measurements were done by
two examiners (DM, MB) blinded to the groups. The two ex-
aminers made measurements by consensus, not independently.
All measurements were repeated (after greater than 6 months) by
one of the two examiners (DM) and independently by a third
examiner (ML) to determine the intraobserver and interobserver
correlations of the measurements. Differences between the two
groups in each of the six projections were calculated with un-
paired two-tailed t tests. Differences between the six different
projections in the same group were calculated using paired two-
tailed t tests. Anteversion of the femoral head was measured and
correlated with the angle alpha using Pearson’s correlation. Ac-
cording to the Bonferroni correction, values of p < 0.01 are
considered statistically significant.
RESULTS
For all radiographic projections, the offset angle alpha was
larger in the aspherical group compared with the spherical
group (Dunn, p < 0.0005; Dunn/45° flexion, p < 0.0005;
cross-table/15° IR, p < 0.005; cross-table/NR, p < 0.005;
and cross-table/15° ER, p < 0.01). Because of substantial
standard deviation (± 16°), we found no difference in the
AP projection (Fig 4). The alpha angles for the aspherical
and spherical groups were the same.
The Dunn view with 45° hip flexion was the most sen-
sitive projection for detecting a large angle alpha. The
Number 445
April 2006 Femoroacetabular Impingement 183

Fig 2A–B. A drawing of the anterior view of two femora shows

one with a (A) normal transition of the femoral head to the neck
and one with an (B) asphericity anterolaterally on the transition
zone of the femoral head and neck.

DISCUSSION
A key for prevention of orthopaedic diseases is early ap-
preciation and eventual treatment of predisposing morpho-
logic features. For hip dysplasia, where the insufficient
acetabular coverage is a radiographically well-appreciated
predisposing factor, ascertaining and correcting this under-
coverage can lead to substantial improvement of the

Fig 1A–F. The contour of the same proximal femur is shown

in six different projections: (A) standard AP view, (B) Dunn
view in 45° hip flexion, (C) standard Dunn view, (D) cross-table
view in 15° internal rotation, (E) in neutral rotation, and (F) in
15° external rotation. The anterior offset angle alpha is defined
as the angle between: (1) the connecting line of the center of
the femoral head and center of the femoral neck, and (2) the
connecting line of the center of the femoral head and Point A.
Point A represents the point at the femoral head/neck junction
where the radius of the femoral head diverges from the femoral
neck.

alpha angle in the externally rotated cross-table view was

smaller (p < 0.05) in all projections except the AP view
(Fig 4).
Orientation of the greatest diameter of the femoral neck
rho around the long axis of the femoral neck was similar
in the spherical group (21° ± 9°) and the aspherical group
(25° ± 8°). Antetorsion of the femoral neck did not cor-
relate with the measured offset angle alpha.
The intraobserver and interobserver correlations for all
measurements combined were R ⳱ 0.95 and R ⳱ 0.88,
respectively. The best correlation was with the cross-table Fig 3. A drawing shows the orientation of the greatest diam-
view in internal rotation (R ⳱ 0.97 for intraobserver and eter of the femoral neck around its own axis (␳) relative to the
interobserver correlations). mechanical long axis of the femur in the sagittal plane.

184 Meyer et al
Fig 4. The bar diagram shows the results of the anterior offset
angle measurement for the groups with spherical and aspheri-
cal head/neck junctions. Probability values less than 0.01 are
significant.
course of the disease.11 Predisposing morphologic alter-
ations have not been as well defined for primary (idio-
pathic) OA. With the new concept of femoroacetabular
impingment, acetabular and femoral alterations such as
asphericity of the head/neck junction are thought to dam-
age cartilage with subsequent OA.3,5,12,21 To prevent such
early changes, timely and reliable diagnosis of routine ra-
diographs is desirable. Our study focused on the radio-
graphic determination of femoral head/neck asphericity as
a predisposing factor in femoral-induced femoroacetabular
impingement (cam type). Based on the assumption that
anterosuperior femoral head asphericity may be hidden in
some radiographic projections, we evaluated the optimal
radiographic exposure of the femur to identify this disorder.
A limitation of this study is that a small number of
desiccated specimens was analyzed. However, differences
between the groups were obvious in macroscopic inspec-
tion. On conventional radiographs, only the bony structure
is evaluated to determine the shape of the head/neck junc-
tion, which was preserved in these desiccated specimens.
However, in patients, the shape of the bone may be less
clearly visible than in the desiccated specimens used in
this study; small and hidden irregularities even more likely
may be missed. This emphasizes the need for an optimal
exposure of the hip on routine radiographs.
There is a continuum from spherical to aspherical femo-
ral head/neck junctions. However, we were interested in
identifying differences between radiographic projections
rather than the absolute values of the angle alpha of the
specimens; the specimens we used seemed adequate for
this purpose. When we compared our data for the angle
alpha with that of Notzli et al,15 we found similar values
(71° versus 50°) as with magnetic resonance imaging
Clinical Orthopaedics
and Related Research
(MRI) (74° versus 42°) for abnormal and normal hips.
Notzli et al15 found a 12% decrease in the asymptomatic
population of 30-year-old volunteers. Although the carti-
lage layer thickness at the femoral head/neck junction
might differ from other parts of the femoral head, this does
not seem to alter the data they obtained. We focused on
femoral alterations detectable on radiographs, but clini-
cally, the femur must be evaluated in association with the
acetabulum (entire hip), as both parts determine the degree
of femoroacetabular impingement.
We evaluated six radiographic projections to observe
femoral head asphericity. In the same femur, the measured
angles of the head/neck offset alpha varied by greater than
30° depending on the radiographic projection. This is
partly because of the flat, oval-shaped femoral neck. In
cross-section, the oval shape of the femoral neck is rotated
anteriorly by a mean rho angle of 23°. This directs the
largest diameter of the femoral neck anterosuperiorly,
making this the region with the least femoral head/neck
offset. In the cross-table view, the anterosuperior ridge of
the neck is hidden behind the abnormal part of the femoral
neck with external rotation of the femur. With internal
rotation, the anterosuperior ridge of the neck appears on
the projection (Fig 1D,E). In the AP view, the critical zone
may be hidden behind the normal parts of the femoral
neck. The best of the six exposures tested was obtained
with the Dunn view where the femur is flexed 45° and the
flat geometric shape of the femoral neck is nearly parallel
to the xray plate.
Based on the bony structure of the femoral neck, the
ideal radiographic position to identify the critical zone
most likely would be obtained with the hip flexed 25°,
neutrally rotated, and abducted 20°. This position ensures
maximum parallel orientation of the flat femoral neck with
the xray plate. However, this would require establishing a
new standard view. We think this is not indicated as the
Dunn view in 45° flexion, which is similar to the
Schneider femoral head contour projection,17 provides
sufficient information. The superiority of the Dunn view in
45° flexion supports the abnormal femoral head/neck con-
tour being located anterosuperiorly.
We considered it reasonable to use a combination of an
AP view of the pelvis and a cross-table view in internal
rotation. These projections do not need additional leg
holders for the examined extremity. Femoral head/neck
asphericity was best detected with the Dunn view in 45° or
90° hip flexion, neutral rotation, and 20° abduction. The
cross-table lateral view had a comparable sensitivity, but
should be obtained with the leg in approximately 15° in-
ternal rotation. These radiographs will minimize false-
negative results in patients with suspected femoroacetabu-
lar impingement.
Number 445
April 2006
References
1. Eijer H, Leunig M, Mahomed MN, Ganz R. Anterior femoral head-
neck off-set: a method for measurement. Hip Int. 2001;11:37–41.
2. Eijer H, Myers SR, Ganz R. Anterior femoroacetabular impinge-
ment after femoral neck fractures. J Orthop Trauma. 2001;15:475–
481.
3. Ganz R, Parvizi J, Beck M, Leunig M, Notzli H, Siebenrock KA.
Femoroacetabular impingement: a cause for osteoarthritis of the hip.
Clin Orthop Relat Res. 2003;417:112–120.
4. Goodman DA, Feighan JE, Smith AD, Latimer B, Buly RL,
Cooperman DR. Subclinical slipped capital femoral epiphysis: re-
lationship to osteoarthrosis of the hip. J Bone Joint Surg. 1997;79:
1489–1497.
5. Ito K, Minka MA II, Leunig M, Werlen S, Ganz R. Femoroace-
tabular impingement and the cam-effect: a MRI-based quantitative
anatomical study of the femoral head-neck offset. J Bone Joint
Surg. 2001;83:171–176.
6. Kobayashi S, Takaoka K, Tsukada A, Ueno M. Polyethylene wear
from femoral bipolar neck-cup impingement as a cause of femoral
prosthetic loosening. Arch Orthop Trauma Surg. 1998;117:390–391.
7. Leunig M, Beck M, Woo A, Dora C, Kerboull M, Ganz R. Ace-
tabular rim degeneration: a constant finding in the aged hip. Clin
Orthop Relat Res. 2003;413:201–207.
8. Leunig M, Casillas MM, Hamlet M, Hersche O, Notzli H, Slongo T,
Ganz R. Slipped capital femoral epiphysis: early mechanical dam-
age to the acetabular cartilage by a prominent femoral metaphysis.
Acta Orthop Scand. 2000;71:370–375.
9. Leunig M, Fraitzl CR, Ganz R. Early damage to the acetabular
cartilage in slipped capital femoral epiphysis: therapeutic conse-
quences. Orthopade. 2002;31:894–899.
10. Locher S, Werlen S, Leunig M, Ganz R. Inadequate detectability of
early stages of coxarthrosis with conventional roentgen images. Z
Orthop Ihre Grenzgeb. 2001;139:70–74.
11. Millis MB, Kim YJ. Rationale of osteotomy and related procedures
Femoroacetabular Impingement 185
for hip preservation: a review. Clin Orthop Relat Res. 2002;405:
108–121.
12. Murphy S, Tannast M, Kim YJ, Buly R, Millis MB. Debridement of
the adult hip for femoroacetabular impingement: indications and
preliminary clinical results. Clin Orthop Relat Res. 2004;429:178–
181.
13. Murphy SB, Simon SR, Kijewski PK, Wilkinson RH, Griscom NT.
Femoral anteversion. J Bone Joint Surg. 1987;69:1169–1176.
14. Murray RO. The aetiology of primary osteoarthritis of the hip. Br J
Radiol. 1965;38:810–824.
15. Notzli HP, Wyss TF, Stoecklin CH, Schmid MR, Treiber K, Hodler
J. The contour of the femoral head-neck junction as a predictor for
the risk of anterior impingement. J Bone Joint Surg. 2002;84:556–560.
16. Rab GT. The geometry of slipped capital femoral epiphysis: impli-
cations for movement, impingement, and corrective osteotomy. J
Pediatr Orthop. 1999;19:419–424.
17. Schneider PG, Bick JA. Spontaneous necrosis of the femoral head
and alcoholism. Med Klin. 1971;66:1694–1697.
18. Snow SW, Keret D, Scarangella S, Bowen JR. Anterior impinge-
ment of the femoral head: a late phenomenon of Legg-Calve-
Perthes disease. J Pediatr Orthop. 1993;13:286–289.
19. Solomon L. Patterns of osteoarthritis of the hip. J Bone Joint Surg.
1976;58:176–183.
20. Stulberg SD, Cordell LD, Harris WH, Ramsey PL, MacEwen GD.
Unrecognized childhood hip disease: a major cause of idiopathic
osteoarthritis of the hip. The Hip: Proceedings of the Third Open
Scientific Meeting of the Hip Society. St. Louis, MO: CV Mosby;
1975:212–228.
21. Tanzer M, Noiseux N. Osseous abnormalities and early osteoarthri-
tis: the role of hip impingement. Clin Orthop Relat Res. 2004;429:
170–177.
22. Tonnis D, Heinecke A. Acetabular and femoral anteversion: rela-
tionship with osteoarthritis of the hip. J Bone Joint Surg. 1999;81:
1747–1770.