Biology Reference
In-Depth Information
Traditional X-rays
X-rays, generated by a cathode ray tube, have a specific wavelength that allows them to
pass through the body. The X-ray beam is generated when excited by a high-voltage power
supply (
Bronzino, 2005
). All X-ray systems (traditional X-ray, CT, DEXA) have an X-ray
source called a collimator. The collimator sends the rays that pass through the body. The
more dense bone slows down (or attenuates) the beam moreso than the soft tissues. On
the other side of the subject is an X-ray detector, which can be X-ray film, an image intensifier,
or a set of detectors (
Bronzino, 2005
). In traditional radiography, the X-ray then reacts with
a phosphor coating on photographic film (i.e., X-ray film). The difference in the amounts
of absorption and attenuation of the X-rays by the different tissues of the body causes the
shadowing on the photographic paper (
Bronzino, 2005
).
Biplanar Radiography
Traditional radiographs (X-rays) provide an image of the cortical thickness of a bone in
a single, two-dimensional plane. This can be used to estimate cortical area, but it assumes
a circular cross-section of the bone. Combining the breadth measurements of the cortical
bone from two separate planes reduces the error and changes the assumption of a circular
cross-section to one that is elliptical (
Ruff, 1981
). Imagine breaking a doughnut in half
and seeing the cross-section revealed of the inside of the doughnut. If the doughnut is
a perfect sphere, the cross-section you see will well represent the parts that you do not
see. Unlike the doughnut, bone is not a perfect sphere nor is it a perfect ellipse. But
an ellipse can better approximate the area of a bone's cross-section than if a spherical
shape is assumed. By looking at the bone from two separate planes, you are able to
view four separate quadrants of the bone area, not just two. In this way, biplanar radio-
graphs are an improvement over traditional X-ray technology and can be used to better
approximate the three-dimensional cross-sectional shape of a bone. Biplanar radiographs
of a single femur from two different and perpendicular planes can be used to approxi-
mate the cortical area at the midshaft and to establish a ratio of I
max
/I
min
in order to esti-
mate activity levels.
When using biplanar radiography, the overall shape of the bone must be inferred from two
different planes (this is also called ellipse model method (EMM), as explained below). This
inevitably introduces a certain amount of error. Validation studies have been undertaken
to compare the accuracy of the shape predictions when using biplanar radiography to actual
cut cross-sections or CT images. One study found that use of an asymmetrical model from
biplanar radiography improves estimates of cross-sectional geometry of the mandible
(
Biknevicius and Ruff, 1992
). Van Gerven (1969), however, found errors of 25% or more
when comparing cut cross-sections to estimations of midshaft cortical area from AP radio-
graphs. Another study encountered errors of up to 40% from the cross-sectional areas of
the radiographic breadths from several long bones (
Ruff, 1981
). “Radiographic shadowing”
is another problem encountered when using radiographs to infer morphological shape,
which results from the overlap of structures (e.g., the tibial crest can obscure the lateral tibial
curvature) (
Ruff, 1981
).
