X-ray diagnostics and imaging - by pengtt


         Many different types image receptors are used in modern
diagnostic radiology. They all have in common that they form an image by
absorption of energy from the X-ray beam (after transmitting through the
body). The main characteristics will be discussed on the example of the
direct exposure film.

                             Direct-Exposure Film

          Direct exposure film has a relatively low absorption efficiency 
          for photons in the diagnostic range, however it is still used in
          many combinations of image receptor systems.

          Direct exposure film material has special design of two
          photographic emulsions with protective layers between to
          optimize the absorption efficiency  .

          Correct exposure is important to produce a reliable image on the
          film. Over- or underexposure will result in loss of contrast and
          therefore possibly in loss of diagnostic information.
                The proper film exposure can be obtained from the
       so-called characteristic curve of the film material.

          The blackening of the film after X-ray exposure is expressed in
terms of its optical density:

                              D = log10(I0/I)

   where I0 and I is the light intensities before and after passing through
   the exposed film material.
           The objective of this section is to correlate the optical
  density D (amount of blackening) with the received X-ray, exposure.
  This can be obtained with a simple model for the absorption process.

          A single emulsion of film material initially contains G silver-bromine
grains per unit area, the average cross section of the grain is b. After
irradiation with a flux of N X-ray photons per unit area a total number of g
grains per area are sensitized.

The number of sensitized grains per incoming X-ray photon is:

         with X as the received X-ray dose and k a conversion constant.
          The sensitized grains develop into a silver speck with an
average cross section a after the film development. If light hits one of
the silver specks in the developed film, it is completely absorbed (black



        To relate the number of sensitized grains to the optical density
D the absorption of the light in the film material of thickness t can be
described as:

        Using this relation the optical density can be calculated to:

                   This relation is known as Nuttings Law !
            Maximum optical density for an area on the film is obtained
   when all grains are sensitized:   g=G
         The correlation between the optical density D and the
maximum number of sensitized grains results in a relation between
the optical density D and the received dose X :

                                    This is displayed in the figure.

                                    The curve relating the optical density
                            to the film exposure dose is called the
                            characteristic curve of the film material.

                                   In the center part of the curve the relation
                          between optical density and the logarithm of the
                          dose is approximately linear:
        For a small contrast in dose DX= X1-X2 the associated
change in optical density is:

            The constant F is known as the film-gamma and ranges
   between 2-3. It corresponds directly to the slope of the linear
   section of the characteristic curve.

            For low exposure or high exposure the characteristic curve
  levels out, exposure differences do not translate into differences in
  optical density (blackening).
          A film with an optical density of D  0.5 appears
overall light, a film with optical density D  2 appears
overall black.

         To achieve best contrast the film must be exposed hi such
a way that the region of interest in the patient cause film doses
which are in the center part of the characteristic curve.
         Additionally the contrast can be affected by the energy
absorption efficiency of the image receptor material which in
general decreases with energy.

        Efficiency and / or sensitivity of film material

          The sensitivity of film material depends on size and
density of grains, emulsion thickness and X-ray absorption
         The figure shows a typical X-ray absorption efficiency 
for a double emulsion film as a function of energy.

                                       The efficiency drops rather rapidly
                              with increasing energy and is mainly
                              determined by the interaction probability of
                              the photons with the film material
                              (attenuation coefficient m ) and the thickness
                              of the material t.

            An important goal is to maximize the efficiency of the image
   receptor material.
                  The noise in the image may limit the contrast.

The noise in the receptor image arises from several sources:

          fluctuations in the number of absorbed X-ray photons per unit area

          fluctuations in the absorbed photon energy

          fluctuations in the number of silver halide per unit area of emulsion

                     The first and the last are the main sources for noise
             (quantum mottle and random darkening).
To calculate the effect of quantum mottle we replace:

               with A as area,  as interaction efficiency, and N as
               number of incident photons.

The resulting expression for the noise in the optical density of the image is:

                        The noise due to quantum mottle is proportional to the
               slope G of the characteristic curve of the receptor material!
         To determine the noise due to random darkening the
influence of granularity fluctuation (g  A)1/2 in the number of
developed grains (g  A) in area A for the fluctuation in optical
density DDG needs to be calculated:

                         after substituting for g.

    Fluctuations therefore depend directly on fluctuations in grain size!
        The figure shows that the quantum mottle
corresponds directly to the film-gamma and the random
darkening due to the granularity distribution directly to the
characteristic curve.
                   Alternative Image Receptors

Intensifying Screens in front of the photo emulsion convert X-rays
into visible light. The film material is more sensitive to light photons
than to X-ray photons. This increases the energy absorp-tion
efficiency e by more than one order of magnitude, but the
resolution decreases due to additional noise components.
Image intensifiers convert X-ray photons to electrons by
photo-electric effect on a photocathode. The electrons are
focused onto by electrical fields on a fluorescent screen where
they form an in-tensified image which can be recorded on film
or viewed with a TV camera.
Xeroradiography is a dry non-silver photographic system, which
produces images on paper (Xerox copies). It is slower than
standard film receptors (which may result in higher doses) but has
better resolution and better energy absorption efficiency e
          lonography replaces the photographic film by a
position sensitive ion chamber. The X-rays induce by
ionization of the gas electron clouds which can be detected by
an electrode with good spatial resolution.
                   X-Ray Transmission Computed Tomography

Several problems exist with conventional radiography techniques:

            inability to distinguish soft body tissue because of limited contrast
            (see example blood-muscle); this can be fixed by the use of liquid
            contrast medium which has to be injected.

            inability to resolve spatially structures along the X-ray propagation
            axis resulting in loss of depth information (flat picture), because
            the three-dimensional body is projected on to a two-dimensional
      Computed tomography (CT) techniques allows sectional imaging .

         It is based on the principle that an image of an unknown object can
be obtained if one has an infinite number of projections through the object.

                 Two scans through the body gives an image in x- and y-direction
        (side view and front view). With increasing number of scans over a 360°
        angle range provides a set of images which allow to construct the three
        dimensional structure of the body.
        For X-ray tomography a planar slice of the body is defined
and X-rays are passed along this plane in all directions. To produce a
tomographic image typically 100 to 1000 scans are required.

    To store the multitude of images and process the data requires computer.

                                                        The two-dimensional
                                               image corresponds to a three
                                               dimensional section of the
                                               patient with the third dimension
                                               being the slice thickness which
                                               is typically a few millimeter

               The resulting spatial resolution is 1 mm, a density discrimination
      (contrast) of better than 1% can be obtained with this technique.
         Present CT machines have an rotatable X-ray source.
This allows to scan the patient who is located along the
rotational axis from all sides and angles. The image receptor
system is designed as a stationary ring of detectors around the
patient which receive the portions of the fanned X-ray beam.
        Each data point acquired by the detector array is a
transmission measurement through the patient along a given line
x between source and detector pixel.

         Each data point therefore follows the basic equation:

          The attenuation coefficient p represents the sum of all attenuation
coefficients along the line x:
From the image information the projection P(x) of the image is calculated:

                                                      Because the attenuation
                                            coefficient m corresponds directly to
                                            the density of the body tissue along
                                            the projection axis x, the projection
                                            corresponds to the density.

                     From a complete scan along different axes (x, f) the cross
            sectional density along the slice can be constructed using Fourier
            analysis methods.
From the image information the projection P(x) of the image is calculated :

                                        The projection is directly proportional to
                                        the summed attenuation coefficient and to
                                        the length x through the body.

                                         Because the attenuation coefficient m
                                         corresponds directly to the density of
                                         the body tissue along the projection axis
                                         x, the projection corresponds to the

                     From a complete scan along different axes (x, f) the cross
            sectional density along the slice can be constructed using Fourier
            analysis methods.
         The resulting CT image is a two-dimensional matrix of numbers
with each number corresponding to a spatial location in image and patient
along the plane of the slice. The matrix is constructed from 512x512 pixels.

           Each pixel has a value (up to 4096=12 bits) which corresponds to the
  level of gray (darkness, attenuation). This number is called CT number and
  contains the physical information ( attenuation  density ) about the
  corresponding body section.
          As the attenuation coefficient ranges between m=0 (air) and m=1 (metal
inlet) the attenuation is scaled to a maximum of 4096. The CT therefore
corresponds to a two-dimensional map of attenuation across the body slice.

            The CT number CT is normalized to the attenuation of water:

      If the attenuation coefficient for a given pixel is equal to water, the CT
      number=0, soft tissue material has CT number in the range of -100 to +100,
      dense tissues like bones have high CT numbers from 300 to 3000.
                CT numbers are rescaled attenuation coefficients!
Typical Radiation Doses in Radiographic and in CT Examinations

               The radiation dose received in CT is considerably
      higher than that of conventional screen radiography.

             To compare the two types of X-ray exposures the received
   dose D of X-ray radiation is converted to integral dose DI, which
   corresponds to the total amount of energy deposited in the body
   tissue of mass m:
In radiographic examinations the dose is not distributed evenly but
drops from the entrance dose at the skin D0 towards deeper layers.

The integral dose is described by:

         For the skull m  0.33 cm1, A  256 cm2, the typical entrance
dose is D0  4.8 mGy. This yields for the integral dose:

                                 DI = 3.7J

                    The integral dose in a seven slice head CT is about
           twenty times higher than the integral dose of a single X-ray

To top