Radiology of the Hip


The pelvis is tilted anteriorly by about 45 degrees and normally does not have any rotation or obliquity. That is why an AP pelvis xray is an intermediate view of the pelvis with the outlet view being a true AP of the sacrum and the inlet view showing the entire pelvic ring in the axial plane. The ilium is oriented obliquely and a 45 degree external or iliac oblique view shows the full profile of the ilium, anterior wall of the acetabulum, and the posterior column well and the internal oblique or obturator view shows an orthogonal view of the ilium, posterior wall, and the anterior column.

Normal morphology is widely variable. We are better off by calling it ideal morphology when we describe each parameter so that we can recognize deviations from it.  Morphological variation is often not pathological because several factors are dependent on other factors in producing pathology. For example, Coxa valga is a variation that can cause increased joint reaction forces and contribute to instability, but can be tolerated well if there is no acetabular deficiency or increased anteversion of the femur.

There are certain abnormalities like cysts in the bone, fracture lines, joint space narrowing, slipped epiphysis, and fragmentation in early Perthes that are always abnormal and pathological. Certain findings like bone islands are incidental findings. Then the other variations like coxa vara, valga, over coverage, cam morphology, and dysplasia which are deviations from the ideal morphology that are abnormal only when they seem to cause pathology and correlate with symptoms or influence the outcome of treatment.

Plain radiograps

Radiograph gives a global two dimensional view of the bony anatomy of the hip. With experience, interpretation of the three-dimensional structure becomes adequate. A full pelvis radiograph allows interpretation of the orientation of each acetabulum and even determines its version. Anatomy can be seen in any AP view but standing AP is better in showing subluxation in dysplastic hips and functional position of the hips while supine xray is easier to control to get a perfect view.

Perfect Ap view

Beam centered on the center of the pelvis showing both walls of the acetabulum with no rotation or obliquity and normal tilt of the pelvis.

Obliquity – The ideal pelvis X-ray should not have any obliquity. That means both hips should be at the same level.

Ideal rotation is when the spinous process line passes through the pubic symphysis. Obturator foramina are symmetrical.

Pelvic tilt – Ideal AP view should have 2-3 cm between sacro-coccgeal junction and the top of the pubic symphysis. Ganz accepts up to 6 cm in a female pelvis. More distance is going towards inlet and less is going towards the outlet view.

The ideal inlet view for trauma purposes is when the anterior borders of S1 and S2 overlap.

The ideal outlet view is when the top of the symphysis is just below S1 body.

Inlet view falsely exaggerates anterior wall coverage of the femoral head and falsely decreases posterior wall coverage. Outlet view has the opposite effect.

Right rotation of the pelvis exaggerates anterior coverage and falsely decreases posterior coverage of the right femoral head and does the opposite to the left femoral head. 

Inlet and right rotation have additive effect on exaggerating anterior coverage and decreasing posterior coverage on the right and negating the effects of each other on the left.

AP view findings

Both hemipelves are usually symmetrical in size and shape of all structures. Radiographic asymmetry and abnormalities are not uncommon and do not have much importance in the absence of symptoms or high risk signs of damage like subluxation, joint space narrowing, and changes in bone density. Discussion of synovial diseases, infections, bone tumors, and marrow diseases like avascular necrosis will not be the focus of further discussion. Some abnormalities are often associated with pathology while others are considered normal variants when they do not correlate with symptoms or clinical exam.

Pubic symphysis

Widening and irregularities of the pubic symphysis are commonly associated with hip impingement.

Prominent AIIS

Healthy AIIS should stay above the sourcil. Prominent AIIS extends inferiorly overlapping the acetabulum causing a false cross over sign.

Ilio-inguinal line represents the anterior column and is used to note proper rotation of an acetabular osteotomy involving a pubic osteotomy.

The ilio-ischial line, also called Kohler’s line represents the posterior column and is a pelvic parameter to define the position of the hip joint. The acetabular fossa is lateral to Kohler’s line. Tear drop is a ‘U’ shaped radiographic appearance formed by the Kohler’s line medially and the acetabular fossa line laterally. Narrow tear drop or reversed ‘U’ are considered signs of a deep acetabulum or medially placed acetabulum. Widened tear drop is commonly seen in acetabular dysplasia, but dysplastic hips can sometimes show a narrow or reversed tear drop making it less diagnostic. Tear drop shape. A normal tear drop is a few millimeters wide U shaped  xray finding. The medial limb of the U is usually the Ilio Ischial live and the lateral is the acetabular fossa. A wide tear drop is commonly associated with dysplasia and reversed tear drop where the acetabular fossa goes medial to the ilioischial line is a sign of coxa profunda. Neither of the findings are very specific and we have better ways to assess under and over coverage by looking at the sourcil and the two walls.

Ischial spine sign

The ischial spines are ideally not visible on the AP view. If one or both are visible on a good AP, that is called ischial spine sign.  It is an indirect sign of external rotation of the hemipelvis/ retroversion of the acetabulum. It is not used to quantify the deformity well. It is a normal finding if the X-ray is rotated to the same side or is an inlet view.

Radio graphic assessment will be most accurate after skeletal maturity


The acetabular morphology can be defined by three structures. Sourcil, anterior wall, and posterior wall. All three are seen adequately on the AP pelvis.

Sourcil is a radio graphic density of the subchondral bone. It represents the weight bearing dome of the acetabulum. It is not effected much by mild tilting or rotation of the pelvis x-ray. Its size, shape, and tilt are important to look at independently.

Size – The sourcil should cover about 80% of the head. Less than 70% is under coverage.  More than 90% is over coverage.  The size can also be measured with lateral center edge angle (LCE) which is ideally 25-35 degrees. Rarely, there may be good coverage laterally but the medial coverage may be deficient when the acetabular fossa is high. Medial center edge angle (MCE) can be measured separately and should be at least 20 degrees. Lateral over coverage with medial deficiency is a contraindication to do rim trimming because that makes the cartilage surface too small.

Shape – should be concave to match and contain the convex head. Lateral upslope of the sourcil appears as increased lateral joint space. It happens early when the triradiate is still open from remodeling of the lateral acetabulum to accommodate the asphericity of the femoral head. So, it is a sign of cam morphology that can cause cam impingement. It is often seen in slipped epiphysis. A small lucency surrounded by sclerotic border can be seen in the sourcil as  a normal variant called supra acetabular fossa

Sourcil angle / Tonnis angle/ Acetabular index – it is the angle between a transverse line connecting the bottoms of the tear drops and a straight line drawn from one end of the sourcil to the other end. Normally this should be 0-10 degrees up sloping and perpendicular to the primary compression trabeculae of the femur.

A good sourcil should extend 25 to 35 degrees lateral and at least 20 degrees medial to the center of the head, it’s shape should be concave to match the top of the femoral head, it should be flat to 10 degrees up being perpendicular to and centered over the primary compression trabeculae. The sourcil angle is more important than the size and shape. It imparts stability to the hip. A small sourcil placed flat on the head is better than a large sourcil that is tilted too far up or down. Stability is more important than the surface area.

The anterior wall is more horizontal, has an ‘S’ shape, and leads to the bottom of the superior pubic ramus. Ideally, it stays completely medial to the posterior wall and contacts it at the lateral edge of the sourcil covering 20% of the femoral head. Anterior wall crossing the posterior wall is always abnormal, but does not always cause symptoms or disease. Anterior wall coverage can also be measured with ACE on a faux profil view. I don’t need that view because I can see the anterior coverage on the AP view and ACE was found not to be accurate in later studies.

The posterior wall is more vertical, usually straight, and leads to the lateral side of the ischium. It normally covers half of the head and stays lateral to the anterior wall. So LCE measures posterior coverage unless there is a cross over sign. With cross over sign, I measure posterior wall LCE separately to quantify posterior coverage at the dome level.

Posterior wall passing medial to the center of the head is called posterior wall sign; a sign of posterior wall deficiency. Posterior wall sign is not quantitative because it does not tell us if the posterior wall is mildly deficient or severely deficient. You can also have the posterior wall crossing the center of the head, but deficient at the sourcil level – posterosuperior deficiency. Posterior wall LCE measures that well.

Ossified labrum can produce two lines representing the walls called Double wall sign.

Look at the three acetabular components individually. Anterior wall, posterior wall, and sourcil for shape and percent coverage. Ideally, the anterior wall should cover 20% of the head, posterior wall 50%, and sourcil 80%. More than these values is morphological over coverage and less is under coverage and the values quantify the coverage well both for classifying the morphology and for treatment purposes.

With each wall possibly being normal, deficient, or over covered you can get 9 types of acetabulae and only one is ideal and 5 of them can have a cross over sign.

Over coverage


Anterior over coverage




Posterior dysplasia


Posterior over coverage

Under coverage

Global dysplasia

Anterior dysplasia

Increased anteversion



Under coverage


Over coverage

Cross over sign means that anterior wall is more lateral than the posterior wall. It is not diagnostic of retroversion or any other morphology. It can be seen with 

1.True retroversion – anterior over coverage, posterior under coverage

2. Coxa profunda / global over coverage / protrusion – Anterior over coverage is more than posterior over coverage.

3. Global dysplasia. In 1/3 of dysplastics, posterior coverage is more deficient than anterior coverage 

4. Posterior dysplasia – Anterior coverage is normal and posterior is deficient

5. Anterior over coverage with normal posterior coverage.

Cross over sign, ischial spine sign, and tear drop shape are indirect signs and are not diagnostic. 

Cross over sign, prominent ischial spine, and posterior wall sign together suggest that the acetabulum is retroverted.

Treatment should be determined by the degree of posterior deficiency, risk of arthritis from it, and the age and preferences of the patient. Older than 25 may not be good for anteverting PAO because there may be anterior acetabular damage already.

With posterosuperior dysplasia, I would like to measure LCE angle of the posterior wall separately because the regular LCE measures the sourcil. Tend to consider an osteotomy if it is less than 20.

When we can quantify anterior and posterior wall coverage on a perfect AP view, retroversion is anterior coverage more than 20% and posterior coverage less than 50%.

Measuring wall coverage this way helps with making the correct diagnosis and recommending the appropriate treatment.

Radiology of the proximal femur

The normal femoral head is mostly spherical with thicker and wider cartilage superiorly. The epiphysis extends on to the neck anterolaterally often causing cam morphology, defined as an alpha angle of more than 55 degrees. Anterolateral cam can be seen on a 45 degree Dunn or a shoot through lateral with the hip in internal rotation. Lateral cam is seen on the AP view. Anterior and posterior cam are seen on frog lateral views. The best evaluation of head neck contour/cam is on radial MRI and 3D CT.

The physeal scar can be seen extending laterally as a sign of cam morphology called epiphyseal extension sign.

Sagging rope sign shows the lateral extent of the anterior head. A second sagging rope is the lateral extent of the posterior head.

Fovea is usually located inferomedially and does not contact the U shaped acetabular cartilage. On the AP view the fovea should be below the medial edge of the sourcil by at least a 10 degrees’ arc. Fovea higher than that is called Fovea Alta which is a risk factor for foveal impingement where the ligamentum teres rubs against the acetabular dome and causes pain.

The epiphysis can be tilted posteriorly on the neck. Southwick angle considers up to 12 degrees of posterior tilt to be normal. With idiopathic cam morphology the head has an anterior tilt. In SCFE, posterior tilt is seen in 99.9%, varus tilt in 98% and valgus tilt in 2%. Siebenrock calls it a varus slip if the neck axis passes above the fovea on the AP view.

Femoral neck

The primary functions of the neck are to increase abductor lever arm and allow more range of motion. The neck has a neck shaft angle of 130 degrees on the AP view, and 15-20 degrees of anteversion.

Neck shaft angle measurement is error prone, varies with hip rotation, and the angle itself is not important. This angle can be changed without breaking or bending the medial cortex between the head, neck, and shaft as we do with relative neck lengthening.

The height of the tip of the greater trochanter relative to the center of the head is a better estimate of the neck shaft angle. Above the center is coxa vara and below is valga. This method is less variable with hip rotation.

Coxa vara shortens the abductor lever arm and can cause weakness, limp, and fatigue pain. It also decreases abduction, both internal and external rotations in flexion and intra articular and extra articular impingement. Trendelenburg gait can decrease joint reaction forces, increase shear forces on the hip cartilage, and valgus stress on the knee joint.

Coxa valga increases joint reaction forces, makes the hip more unstable, brings the neck and lesser trochanter closer to the pelvis predisposing the hip to intra and extra articular impingement in extension.

Anteversion increases flexion and internal rotation, thereby decreasing the risk of anterior impingement. But it increases the risk of posterior impingement. Retroversion increases the risk of anterior impingement.

The neck length is typically one head diameter from the center of the head to the tip of the greater trochanter. Less than that is breva which has similar effects as coxa vara and us usually associated with vara. I am not aware of the name or effects if too long a neck. May predispose the hip to femoral neck fracture.

Concavity of the neck

A narrow neck with a spherical head gives greater range without impingement. A thick neck or a more medial head neck junction increase the risk of impingement. Head neck junction is the point where the sphericity of the head ends.

Alpha angle measures how medial the head neck junction is and Head neck offset measures the depth of the neck relative to the head.

None of the risk factors like over coverage, cam morphology, or coxa vara are diagnostic of hip impingement.

Hip dysplasia is always a radio graphic diagnosis while fai is mostly a clinical diagnosis and treated by correcting radiographic abnormalities.

Radiographic signs of dysplasia

In addition to measurable decreased coverage of the head by the sourcil, anterior wall, and posterior wall, hip dysplasia is significant if the hip also shows signs of instability. Subluxation is the unequivocal sign of instability and is seen as widened medial joint space, break in Shenton’s line, acetabular rim fracture, and cyst formation from rim loading.

Radio graphic signs of impingement.

Sclerosis, ridge or groove formation at head neck junction, cyst formation under the cam bump are early signs, while narrowing of lateral joint space, medial osteoarthritis, subtle lateral subluxation of the head into small lateral acetabular defect, double wall sign, labral calicification and head neck osteophytes are late signs of damage.

Femoral head & Acetabulum influence each other’s development

You probably heard of experiment of placing a square object in the acetabulum and the acetabulum becoming square.  It is quite common to see lateral joint space widening in idiopathic cam, scfe, and Perthes from acetabular remodeling induced by the aspherical head in a growing child. On the other head, it is very common to see cam morphology with acetabular dysplasia probably from inadequate remodeling/rounding of the head because the socket was not deep enough to influence the head shape. Protrusio hips have a very spherical head with a narrow neck (low alpha angle).  Protrusio and cam combination is less common and is very likely to become symptomatic early.

Interaction of morphological features

Though radiographic variations are very common, symptoms and damage are less common because most variation/ abnormaility can be compensated by another morphological feature. Examples are 1. Lateral dysplasia by coxa vara or flattening of the sourcil angle2. Coxa profunda by coxa valga and increased head-neck offset 3. Cam morphology and mild scfe by increased anteversion or shallow acetabulum 4. Anterior dysplasia by femoral retroversion 5. Posterior dysplasia by femoral anteversion.

In summary, each morphological variation has a specific effect and the effects of each variation can be additive to cause disease early or they can compensate for others and allow good function. As mechanical hip disease is activity related, more active people are more likely to become symptomatic early. To add to the morphology , biology dictates who becomes more symptomatic from the same activity level. So, the prognosis of each variation is unknown in that particular person. The only way to correct these variations is by surgery, As surgery has cost, risk of complication, no guaranteed  outcome, we offer surgical treatment only  to symptomatic and active patients who understand all this. The exception is dysplasia in a growing child that gets worse without causing pain. We push surgery harder when poor prognosis without surgery is well known.

Pelvis imaging for trauma

AP pelvis is less useful in trauma because the pelvis is tilted anteriorly.

Ideal inlet view is an axial view of S1. You want the anterior border of S1 & 2 to overlap perfectly. You see what axial cuts of CT show on this view. Superior displacement of the hemipelvis is seen as posterior displacement on this view and CT scans.

Inlet also shows rotational deformity very well and sacral alar buckling as in LC1 is seen.

Ideal outlet view is AP of S1. Typically the top of pubic symphysis is placed at the bottom of S1. This shows the entire sacrum and pubic rami in profile, and vertical displacement and flexion injuries are seen well.

The standard angles of the xray machine to get good inlet and outer views may not be good for all patients because the pelvic tilt is variable.

A useful tip is to measure the angles of these views preoperatively on supine pelvic CT to get AP and axial of S1 and start the flouro at those angles and fine tune as needed.

Almost all pelvic reduction and fixation can be done with these two views, but a perfect lateral of the sacrum, though difficult to get, helps with the starting point adjustment and placing trans sacral screws in S1 and S2.

Another view of occasional use is the obturator outlet to see the tear drop above the acetabulum that represents the supra acetabular column that goes from AIIS to PSIS. Great corridor for large front to back or back to front screws or ex fix pins. Ideal view is to place the teardrop a few millimeters above the acetabulum and there should be minimal Iliac overlap laterally. Iliac oblique can confirm screw path above the greater sciatic notch and for length of the screw.

MRI rules out pathologies like AVN and tumor, shows labral tears sometimes and damage when it is advanced. I use it when the clinical diagnosis is not clear and to differentiate between borderline dysplasia versus cam damage looking at the labral size. Circumferential alpha angle measurements are best done on radial sequences.

AP pelvis is less useful in trauma because the pelvis is tilted anteriorly.