Digital Image Processing

1
Digital Image Processing

• Thanks to the work of Dr. Perry Sprawls of Emory
  University and the Sprawls Educational
  Foundation, this material is available on-line.

2
Digital Image Processing

• A major advantage of having images in digital
  format is the ability to perform a variety of
  image processing procedures with a computer.
• The procedures are selected and adjusted to
  change the characteristics of the images.

3
Digital Image Processing

• Processing purposes
• Improve quality
• Optimize characteristics for maximum visibility

4
Changing Image Characteristics

• Processing can change most image properties.
• Three possibilities include
• Reduce Image Noise
• Increase Detail
• Adjust and Optimize image contrast

5
Contrast Adjustment with Processing

• While there are many processes that can be used
  to adjust contrast, the most common ones are
• Look Up Table (LUT) processing
• Windowing
• Both are used in DR and other imaging modalities.

6
Using Digital Processing to Change Image Contrast

• Here is a very simple image to show how digital
  processing can be used to change contrast.
• The image consists of a background area with a
  small square in the middle.

7
Using Digital Processing to Change Image Contrast

• The background has a value of 40 and the small
  square a value of 30.
• The numerical contrast would be 10 (40-3010).
• A numerical value of 10 would be low contrast.

8
Using Digital Processing to Change Image Contrast

• Look up tables are data stored in the computer
  that is used to substitute new values for each
  pixel.
• In this simple example we have values of only 30
  and 40.

9
Using Digital Processing to Change Image Contrast

• As we see here, the processing uses a LUT that
  substitutes 90 for 40 and 10 for 30.
• The effect is to increase image contrast.
• 90-1080 or improved contrast.

10
Using Digital Processing to Change Image Contrast

• As we will discover, it is usually possible for
  the user to select from a variety of LUTs, each
  one designed to produce specific contrast
  characteristics.

11
LUT described by Graphs

• Lets recall that a LUT indicates what number is
  substituted for each pixel value during image
  processing.
• It is very helpful to show this relationship
  between the original pixel value and the new
  pixel values.

12
LUT described by Graphs

• This is a simple straight line or linear graph
  that show the same values for the original and
  processed value.
• Processing with this LUT does not change the
  image, it just introduces us to the concept of
  LUTs.

13
LUT Curve

• If LUT processing is to change the contrast
  characteristics, it must substitute numbers that
  are different from the original pixel values.

14
LUT Curve

• You should recognize this curve as being similar
  to that of a radiographic film H D curve.
• That is intentional.
• By processing a digital image with this
  characteristic it takes on some of the familiar
  contrast look of an image recorded on film.

15
Selection of LUTs

• One of the advantages of digital processing is
  that the processing parameters (factors) can be
  selected to produce images with different
  contrast characteristics.
• Here we have three different LUTs.

16
Selection of LUTs

• Like the characteristic curve of film, the slope
  of the curve at every point represents how the
  contrast will be changed by processing.
• If the slope is greater than 45º the contrast
  will be increased.

17
Selection of LUTs

• If the slope is less than 45º, the contrast will
  be reduced by processing.
• The curve with greater slope will produce higher
  contrast.
• The one with less slope will produce more
  latitude or more like wide latitude type film
  used in chest radiography. We will look at
  inverted later.

18
LUT Selection

• In the typical digital radiography system, a
  variety of LUTs are installed.
• The appropriate LUT is then automatically
  selected to give the desired contrast
  characteristics for that procedure (chest,
  extremity, spine, etc) that is designated by the
  operator.

19
Processing to Increase Image Contrast

• If the original image was recorded without image
  processing, it would be relatively low contrast
  as we see on the left. This is because of the
  wide dynamic range and linear response of digital
  receptors.
• This is desirable as it is what gives us the wide
  range of exposure response.

20
Processing to Increase Image Contrast

• The digital receptor is not limited by the narrow
  response of film.
• The usual procedure is to use processing as
  illustrated here to increase the contrast for
  some portion of the exposure range.
• The processing simulates a high contrast film
  image.

21
Processing to Increase Image Contrast

• We do not generally use high contrast for chest
  radiography because it produces an image with too
  much area contrast (like the dark lungs shown
  here).

22
Processing a Chest Image

• Here is a LUT processing that is more appropriate
  for chest radiography.
• In general it simulates an image recorded on a
  latitude type film used for chest radiography.

23
Processing the Chest Image

• Compare the high contrast processing to the
  latitude processing. The later has good object
  and anatomical structure contrast and visibility
  throughout the image.

24
Brightness Scale Inversion

• Many radiologists find value in having an
  inverted (black bone) image in addition to the
  conventional (white bone) image at the time of
  viewing.
• Some objects and structures may be more visible
  and easier to detect on the inverted image.
• The inversion can be done with this type LUT.

25
Digital Image Windowing

• The ability to window is a valuable feature of
  all digital images.
• Windowing is a process of selecting some segment
  of the total pixel value (dynamic range of the
  receptor) and then displaying the pixel values
  within that segment over the full brightness
  (shades of gray) range from white to black.

26
Digital Image Windowing

• Contrast will be visible only for pixels within
  the selected window.
• All pixel values that are either above or below
  the window will be either all black or all white.

27
Digital Image Windowing

• The person controlling the window can adjust both
  the center and width of the window.
• The combination of these two parameters will
  determine the range of pixel values that will be
  displayed with contrast in the image.

28
Effect of Changing the Window Level

• One of the advantages of windowing is that it
  makes it possible to display and enhance the
  contrast in selected segments of the total pixel
  value range. This is compared to not being able
  to adjust contrast on the processed film image.

29
Effect of Changing the Window Level

• With windowing we can create many displayed
  images each focusing on a specific range of pixel
  values.

30
Effect of Changing the Window Level

• When the window is set for the lower pixel range
  we can enhance contrast in the medistimum.
• When set for the upper range of pixel values, the
  lung is enhanced.

31
Enhancing Visibility of detail

• The Blurred Mask Subtraction is often used in
  digital radiography, to enhance the visibility of
  detail in certain clinical procedures.
• This does not un-blur an image and recover detail
  that was completely lost because of blurring from
  the focal spot, motion of the receptor. What it
  does do is to increase the visibility (contrast)
  of some objects where the visibility is limited
  by the large area contrast such as in the chest.

32
Enhancing Visibility of detail

• The process consists of two distinct steps.
• First a blurred copy of the original image is
  produced.
• A common method is to average the pixel values in
  the pixels area.
• This results in a blurred image that removes all
  detail from the image.

33
Enhancing Visibility of detail

• Now we have two images.
• The original image contains the general large
  area contrast background plus some detail.
• The blurred image contains only the large area
  contrast background.
• The final step is the computer reduces the large
  area contrast background in relation to the
  contrast of detail.

34
Enhancing Visibility of detail

• The final result is that the contrast and
  visibility of the detail (small objects and
  structures) is enhanced.