Post Processing And Artifacts Flashcards
Window and level
-most common image processing parameters (adjustments of brightness and contrast)
-window level: controls how bright/dark the image is
-window width: controls the contrast of the image
🔹wider= lower contrast
🔹narrower= higher contrast
Shuttering/masking
- when viewing an image (film/digital), the unexposed edges around the image can be distractingly bright (causing veil glare: interferes with the image reception of the eye)
- with digital, automatic shuttering is used to black out the white collimation borders, eliminating veil glare
- can be manually removed and adjusted
- this is NOT MEANT to replace collimation
Magnification
Two types
1. Magnifying glass function
🔹a box placed over a section of anatomy shows a magnified version of that section
🔹just like using a magnifying glass to enlarge JUST ONE AREA
2. zoom function
🔹enlarges ENTIRE IMAGE, so that it likely will not be able to fit on the screen- only a protion will be visible at a time
🔹need to use mouse to navigate the image
Spatial frequency filtering
-after signal is obtained for each pixel, to shorten the processing time and minimize storage requirements, the signals are averaged
🔹one pixel will be averaged with the signals of the neighbouring pixels
-if more pixels are involved in averaging, the image will appear smoother
-EDGE ENHANCEMENT is when fewer pixels are used for the averaging
🔹the fewer pixels are included in the average, the greater the enhancement is
Spatial frequency filtering
High pass filtering -when frequencies are amplified -sharpening effect -increases contrast and edge enhancement Low pass filtering -when frequencies are suppressed -smoothing -can lose small details, reduces noise and contrast
Image stitching
- special cases where the anatomy/area of interest is too large to fit on one cassette
- multiple images are ‘stitched’ together to create one image (usually 3-4 CR cassettes are used) software program processes image to display as one single image
Artifacts
-any undesirable density on the processed image (excluding scatter and fog)
-four main types
🔹imaging plate artifacts
🔹plate reader artifacts
🔹image processing artifcats
🔹printer artifcats
Imaging plate artifacts
- usually due to aging or wear
- plate can become prone to cracks due to the constant removal and replacement within the reader (cracks will appear as areas of radiolucency on the image)
- backscatter: worn down or cracked areas of the lead coating on the cassette allow scatter to reach the IP (backscatter can be seen as dark line artifacts)
- can occur for different reasons: incorrect part selection, incorrect sampling
- very important to set appropriate facts and choose the correct body part so that the algorithm can produce the desired image
- poor technique and positioning can cause misrepresentation of the image
Plate reader artifacts
-if there are extraneous line patterns that appear randomly (not regularly), this can be caused by issues with the electronics
-white lines parallel to the direction of the plate travel are caused by debris on the light guide
-requires cleaning or replacement by service personnel
-if multiple IPs are loaded into a single cassette
🔹very rare
🔹only one plate will be extracted, leaving the other one to be exposed multiple times
🔹similar effect as double exposing a film cassette
-insuffiecent erasure can result in residual information being on the IP before next exposure (can occur from over exposure or if the erasure lamp requires work
-moire artifact can occur with the orientation of a stationary grid (grid lines need to be perpendicular to the laser scan direction), not an issue with moving grids
Operator error
- collimation: inappropriate collimation for the body part can result in an improper calculation of the exposure indicator, affects histogram
- cassette use: ex if image is taken on the back of a cassette hardware will be present as artifact on the image
- technical factors: under exposure produces quantum mottle, over exposure reduces contrast
Flat panel artifacts (dead pixels)
- as detectors age, quantity of dead pixels increases, but may not need replacement if they are near edge
- software identifies dead pixels and uses an interpolation method to ‘fill in’ the dead pixels information from surrounding pixels
- can be an issue if there is too many dead pixels
Flat panel artifacts (incorrect gain calibration)
-used for the correction of flaws in the detector
-if there is an area with many dead pixels or poor connections between the conversion layer and the array-there will be a noticeable artifact on the screen
🔹gain calibration corrects this
🔹creates a mask of the defects, so when an image is taken, the software uses the mask to remove the unwanted densities so only the diagnostic information is displayed
Image lag
-some flat panel systems allow for the user to take images faster than the detector can accommodate
-if there is an image taken before the detector has finished releasing all the signal for the previous image, there may be a faint image of the previous image still visible
-like a type of double exposure
-caused by:
🔹taking images in rapid succession
🔹over exposure
🔹areas with little beam attenuation (marker)
Offset correction
-can correct image lag
-determines the amount of signal inherent in the detector
🔹creates a mask of the residual signal
-if this is done before residual signal has left the pixels, that information can be stored as inherent and cause a negative image of the signal- inverse image
