Sampling, Histograms, LUT, And Exposure Indicators Flashcards
Digital imaging
-digital image receptors capture a wide range of exposures
-if the entire range were digitized, the values at the extreme high and low ends would also be digitized, resulting in an extremely low contrast image (too many shades of grey)
-with the CR system
🔹to prevent this, only the optimal density exposure range is processed by the data recognition system
Data recognition program
- works by finding the collimation edges and eliminating the scatter outside the collimation
- if the system fails to correctly locate the collimation edges, this can result in incorrect data collection and images may appear too bright or too dark
- anatomy needs to be centred to the imaging plate to help ensure the appropriate densities are located (failure to centre properly may result in images that are too bright/dark)
Histogram
- takes the image data from sampling, and a histogram is generated
- system finds the useful signal by locating the minimum and maximum signal within the anatomic regions of interest on the image
- all intensities of the signal are then plotted on the histogram
- shows the distribution of pixel values for any given exposure
Image sampling
The plate is scanned and determines: images location and the size of the signal
Histogram
- a graphical representation of the optimal densities within the collimated area
- horizontal axis (amount of exposure)
- vertical axis (number of pixels for each exposure)
- values at the left represent black
- values at the right represent white
- values in the middle are the medium tones
Histogram
-how will a dark image appear?
🔹most of the data points will be on the left
-how will a light image appear?
🔹majority of data points will be on the right
-the information collected within the collimated area is the signal used for image data
🔹this is the source of information for the exposure data indicators
Histogram analysis
- very complex
- shape of histogram is anatomy specific (should be fairly constant for each body part, ex histogram of knee image should look similar to another knee image of a different patient)
- important to select the correct ‘body part’ before processing the IP
- raw data (from IP) used to form the histogram is compared with a ‘normal’ histogram of the same body part by the computer, and the image correct will appear as needed
Nyquist theorem
- in digital imaging, at least twice the number of pixels needed to form the image must be sampled
- if there are too few sampled, there will be a loss of resolution
- if oversampling occurs, there will not be additional useful information gained (no need)
- PSP imaging- many conversions, with every conversion detail is lost
- imaging plate stores electrons for a while, but longer they are stored, they start to lose energy (should read as soon as possible to prevent signal loss)
- FPD systems have less signal loss to light spread vs PSP systems, but the nyquist theorem still applies to ensure that sufficient signal is sampled
Aliasing
-when the spatial frequency is greater than the nyquist frequency- sampling occurring less than twice per cycle or exactly at nyquist frequency
🔹information is lost and a fluctuating signal is produced
-aliasing-produces an image that looks like two superimposed images just slightly misaligned
🔹moire effect
🔹can get a similar effect with grid errors, tech needs to investigate
Automatic rescaling
-used for high or low exposures
-attempts to correct pixel display
-produces images with uniform density and contrast-regardless of exposure (leads to dose creep)
-exposure is too small (quantum mottle will occur)
-exposure is too large
🔹loss of contrast and edge sharpness
🔹due to increased scatter production
Look up tables
- used as a reference to evaluate the raw information and correct the pixel values during processing (raw data does not change)
- mapping function tells us what new value to ‘substitute’ in for each pixel value
- results in an image that will have the ‘appropriate’ look for the anatomic part
- appearance of image is modified
- values are manipulated to change what the ‘output’ values are going to be displayed as
- contrast can be increased/decreased (changing the slope of the graph)
- brightness (density) can be increased/decreased (moving the line along the y axis)
LUT
- Indicates what number is to be substituted for each pixel value after applied to the image (processing)
- straight line LUT does not change the image (shows the substituted number is the same as the original value)
- in order for the display characteristic to change, the LUT needs to substitute numbers that are different than the original values
- shape of the curve is similar to a characteristic curve
Advantages of LUT
can produce images with different contrast characteristics -slope of the curve represents contrast 🔹greater slope = high contrast 🔹less slope = low contrast 🔹inverted = black bone
Why do we need LUTS?
- original images recorded with digital systems are usually very low contrast
- every density is displayed
- use LUT to increase contrast for some section of the exposure range
Exposure indicators
- refers to the amount of exposure received by the IR
- can be used as a guide to monitor the dose to the patient while maintaining acceptable image quality
Exposure indicators
Following an exposure
-the IR is read and a histogram is formed
🔹”counts and plots” all the densities from the exposure into a graph
-exposure indicators represent the average grey shade value of your image
🔹value found by identifying the most exposed vs the least exposed area detected over the ROI (region of interest)
🔹based on collimated boundaries
-compared to a pre programmed histogram of the same body part
🔹histograms are matched/combined
Exposure indicators
Manufacturer/ vendor specific values
- EI numbers (exposure index), used in our labs with our CR equipment
- S numbers (sensitivity numbers)
- EI_s numbers (exposure index sensitivity), used in our labs with our DR rooms
- LGM (log numbers)
Exposure indicators
-EI_s numbers: linear response (as our exposure increases, our EI_s value increases)
-EI numbers: direct relationship (vendor specific)
-LGM numbers: direct relationship
🔹when exposure is doubled, our LGM value increases by a factor of 0.3
🔹when the exposure is hlaved, our LGM value decreases by a factor of 0.3
-S numbers: indirect relationship
🔹when the exposure is doubled, our S value decreases by a value of 200
🔹when the exposure is halved, our S value increases by a value of 200
exposure indicators
Ideal exposure indicator value
- typically the “halfway” value
- the point between the max and the min exposure value limits (unless specified by the manufacturer)
- ideal range= 200-800
- ideal value = 500
- provides the most diagnostic data for the least amount of patient exposure (the perfect balance)
Exposure indicators
With DR systems
- the ideal exposure indicator value for the patient is the lowest possible technical factor selection that will enable the closest value to under exposure, without being outside the recommended limits
- ex ideal range = 200-800
- ideal value = 500
- ideal value for patient = 200
SNR (signal to noise ratio)
Images consist of two components:
-SIGNAL: meaning pattern carrying information about the subject
-NOISE: chaotic pattern, carrying no information about the subject
🔹quantum noise: quantum mottle
🔹electronic noise: created during amplification of signal
-both of these components make up an image
SNR
- High signal to noise ratio: signal is higher than noise
- low signal to noise ratio: signal is lower than the noise, information is lost
Latitude in digital imaging
-refers to the amount of error that can be made and still maintain a quality image
-in digital the exposure can be up to:
🔹200% above the ideal exposure
🔹50% below the ideal exposure
