Graf Method in DDH: Standard Plane, Angles, and Errors
Clinical Task of the Graf Method
The Graf method is intended not for an approximate description of the acetabular shape, but for a reproducible morphological classification of the infant hip joint. In the review Graf, 2022, it is emphasized that the method works only as a sequential algorithm: first anatomical identification, then checking the suitability of the section, and only after that constructing lines and measuring angles α and β.
Practical takeaway for the ultrasound physician: no standard plane — no reliable angles. Attempting to 'stretch' the image with lines in an oblique section is the main cause of inter-expert variability and erroneous typing.
Child and Probe Position
The examination is performed with the child in a lateral position with an assessment of the hip joint in the coronal plane. The probe is placed laterally to obtain a longitudinal image of the iliac bone and the acetabular roof. The purpose of positioning is not just to 'show the joint,' but to strictly display the section for which the Graf angles are validated.
Excessive pelvic rotation, anterior or posterior tilt of the probe, pressure on soft tissues, and attempts to compensate for poor positioning by rotating the probe lead to an oblique section. In such an image, the iliac bone often does not appear as a straight line, the lower edge of the iliac bone disappears or shifts, and the labrum is visualized outside the required anatomical relationship.
Standard Plane: 3 Control Structures
According to the Graf method, measurements are permissible only in the standard plane. Control structures used for its confirmation must be visible simultaneously:
- lower edge of the iliac bone deep in the acetabulum;
- straight line of the iliac bone, suitable for drawing the baseline;
- acetabular labrum, necessary for constructing the cartilaginous roof line.
If at least one of these elements is not determined, the image should not be used for classification. Particularly dangerous is the situation when the labrum is visible, but the lower edge of the iliac bone is absent: the angle can be formally constructed, but it will not meet the method's standard.
Anatomical Identification Before Measurement
Before drawing lines, the main elements of the joint need to be recognized: the iliac bone, the bony roof of the acetabulum, the cartilaginous roof, the femoral head, the joint capsule, and the labrum. This is not a formality. Incorrectly taking a cartilaginous or capsular echo signal for the labrum changes β, and the wrong choice of the bony edge changes α.
In the logic of the Graf method, measurement geometry follows anatomy, not vice versa. Therefore, the protocol should record not only numerical angles but also the type of joint obtained on a suitable standard section.
How to Construct Angle α
Angle α reflects the bony roof of the acetabulum. It is constructed between the baseline and the bony roof line. The baseline is drawn along the straight lateral contour of the iliac bone. The bony roof line is drawn from the lower edge of the iliac bone to the bony edge of the acetabulum.
The key technical error is drawing the baseline along a tilted or curved iliac bone on an invalid section. In this case, even a slight change in probe orientation changes α and can shift the joint from a normal type to a dysplastic one or vice versa. The second common error is choosing the wrong point of the bony edge, especially with a rounded or smoothed transition of the bony roof.
How to Construct Angle β
Angle β characterizes the position of the cartilaginous roof and labrum. It is constructed between the baseline and the cartilaginous roof line, drawn through the bony edge of the acetabulum to the labrum. In the Graf classification, β is especially important for distinguishing between borderline centered and decentering types in the range of α 43–49°.
A typical error is drawing the line to a random bright echo signal in the capsule zone or to the edge of an artifact, rather than to the anatomically identified labrum. Another error is measuring β on an image where the labrum is visible, but the standard plane is not confirmed by the lower edge of the iliac bone.
Graf Classification: Working Thresholds
| Graf Type | Angle α | Angle β / Age | Morphological Meaning |
|---|---|---|---|
| I | ≥60° | β separates Ia and Ib | Mature centered joint |
| IIa | 50–59° | Age up to 3 months | Physiologically immature centered joint |
| IIb | 50–59° | Age over 3 months | Dysplastic centered joint |
| IIc | 43–49° | β ≤77° | Critical, still centered joint |
| D | 43–49° | β >77° | Decentering joint |
| III–IV | <43° | Position of the cartilaginous roof and labrum is assessed | Decentered/dislocated types |
These thresholds apply only after obtaining the standard plane. An α or β number measured on an incorrect section should not be used for type designation.
Probe Tilt Errors
Anterior or posterior tilt of the probe changes the geometry of the section. The image may remain 'beautiful,' but it is no longer standard. Signs of a problem: the iliac bone does not form a stable straight line, the lower edge of the iliac bone is not visualized, the roof appears excessively steep or, conversely, artificially flattened.
A practical control technique is not to start measuring immediately after the femoral head appears. One should consistently achieve a straight iliac line, find the lower edge of the iliac bone, and confirm the labrum. Only after this is the image fixed and lines are drawn.
Typical Measurement Errors
- Measuring angles on any coronal image of the joint without checking the standard plane.
- Absence of the lower edge of the iliac bone on the saved frame.
- Drawing the baseline along the skin, probe, or oblique bone fragment instead of the straight contour of the iliac bone.
- Incorrect choice of the bony edge of the acetabulum for the α line.
- Taking the capsule, fold, or artifact for the labrum when constructing β.
- Rounding the result without considering morphology: the Graf type is determined not only by the number but also by the correct anatomical picture.
Protocol Quality Control
In a systematic review of 131 publications discussed in the Graf 2022 review, it is shown that non-compliance with the method's criteria remains a frequent problem in research and practice. Therefore, quality control should be integrated into the workflow: save the standard section, document α and β, indicate the Graf type, and do not accept images for interpretation without control structures.
For training and audit, a simple checklist is useful: anatomy recognized; three structures of the standard plane visible; baseline drawn along the straight iliac bone; α line goes to the correct bony edge; β line goes to the labrum; type corresponds to thresholds and age. This approach reduces variability and makes the conclusion comparable among specialists.
Frequently asked questions
Can α and β be measured if the lower edge of the iliac bone is not visible?
No. According to the Graf method, this means that the standard plane is not confirmed. Such an image is unsuitable for typing.
Which is more important for typing: angle α or angle β?
α is the main indicator of the bony roof. β complements the assessment of the cartilaginous roof and is critically important for distinguishing IIc and D at α 43–49°.
Why can the same joint receive different angles from different doctors?
Most often due to differences in positioning, anterior-posterior tilt of the probe, choice of bony edge, and incorrect identification of the labrum. Checking the standard plane before measurement reduces this variability.