Chapter 5 Exposure

• digital image processing


  various computer algorithms applied to digital data for the purpose of optimizing the image for display
• what two stages describe digital image processing?

  preprocessing and post processing
• preprocessing


  preparing data for processing image
• postprocessing


  computer operations on taken image
• where do flaws occur?

  raw data
• how do you fix raw data errors?

  bad or dead pixel corrections and flat fielding corrections
• flat fielding corrections


  process of correcting the nonuniformity of pixel values throughout entire image
• Histogram


  graphic display of the distribution of pixel values
• what is another name for a graph?

  data set
• what does the graph include?

  all pixel values that represent the latent image after corrections applied
• x-axis on histogram

  represents range of pixel values assigned
• y-axis on histogram

  represents number of pixels per value
• location of graph on x-axis on histogram

  represents brightness levels
• shape of graph


  represents image contrast (grayscale)
• processing errors are less likely to occur in DR or CR IR's?

  DR IR's
• Histogram Analysis


  computer analyzes histogramusing processing algorithms and compares it with a pre-established (reference) histogram specific to anatomicpart being imaged
• Location and shape of exposure histogram impacted by what 4 things?


  Anatomic area imaged

  Exposure technique

  Positioning

  Collimation
• Values of Interest (VOIs)


  determines the range of the histogram data set included in the displayed image
• do computer softwares have stores histogram models?

  yes
• Exposure Indicator


  a numeric value that is displayed on the processed image to indicate the level of x-ray exposure received on the digital image receptor
• what does digital imaging processing allow?


  images to appear appropriately bright, regardless of IR exposure
• CR Vendor-specific EI's


  Fuji and Konica - Sensitivity (S) numbers

  Carestream (Kodak) - Exposure index (EI) numbers

  Agfa - Log median value (lgM) numbers
• Deviation Index (DI)


  a value that reflects the difference between the desired or target exposure to the image receptor and the actual exposure to the image receptor
• Variables that influence EIs:

  Collimation

  Kilovoltage peak

  Central ray centering
• what do the variables affect?

  level of noise and image quality
• who monitors the EI's?

  radiographer
• Lookup Tables (LUTs)


  provide a method of altering theimage to change the display of the digital image in avariety of ways.
• what are the LUTs based on ?

  the reference histograms
• what do the LUT's alter?

  brightness and grayscale of digital image using computer algorithms
• Automatic Rescaling


  occurs during histogram analysis and is used to maintain consistent image brightness despite over- or underexposure of the digital image receptor.
• Rescaling errors can occur from:


  Improper positioning

  Collimation errors

  Image artifacts
• what can rescaling errors result in?

  poor-quality digital images
• what do LUTs allow?

  grayscale and contrast adjustment
• If processed image not altered

  LUT shows a straightline
• If processed image altered,

  LUT shows a curve
• Alternative Algorithms


  a mechanism for the radiographer to reprocess the image data by selecting an alternative anatomical region or projection to improve its presentation at the display monitor
• Pro to using alternative algorithm:


  Could prevent need for repeat imaging.
• Cons to using alternative algorithm:


  Not assured to have improved image quality

  Could limit ability to further postprocess image

  Reprocessing info could cause storage and retrieval issues.
• Ways digital images can be manipulated:

  1. Dual-Energy Subtraction

  2. Contrast Enhancement

  3. Edge Enhancement

  4. Inversion

  5. Smoothing

  6. Equalization
• Dual-Energy Subtraction


  a technique that can remove superimposed structures so that the AOI becomes more visible; can subtract select brightness values to create an image without superimposed structures.