| Question | Answer |
| Describe SENSITIVITY - detector | the ability of the detector to detect and respond to very low levels of visible light or radiation.
Similar to DQE
Important because it affects patient dose |
| Describe DYNAMIC RANGE - detector | the ability of the detector to respond to a wide range of light intensity, from very dim to very bright/radiation exposure |
| What is the relationship of SENSITIVITY and DQE? | DIRECT
the more sensitive the detector, the higher the DQE |
| What is a CCD? | solid-state LIGHT sensitive device developed in the 70s |
| What is DQE? | Orth p307
detective quantum efficiency
a measurement of overall efficiency with with the detector can convert the x-ray input signal into a useful output image |
| Describe the advantages of digital over film screen. | Increased dynamic range/wide exposure latitude
Greater contrast resolution
immediate results
improved workflow |
| What are the limitations of digital to film screen. | lower spatial resolution
image quality depends on processing and display monitor characteristics |
| Describe SNR | Signal to Noise ratio
white noise interferes w/digital image;
measure of translation of latent image to electronic image;
measures contrast resolution P p48
std method to express noise vs signal strength |
| Describe CNR | contrast to noise ratio;
ratio between image contrast and image noise; can be used to describe image quality
how well the latent image subject contrast translates to image contrast |
| What impacts spatial resolution in detector? | size of the DELs/dexels
in CR, the type of phosphor |
| What is sampling ? | function of detecting and measuring radiation that occurs at the imaging plate or array of detectors
In DR, aperture is square - loss of info between pixels
In CR, aperture is round (due to laser shape) and no loss since it overlaps each sampling |
| What is sampling frequency? | how often analog signal reproduced in discrete digitized form. Increasing sampling frequency of analog signal increases pixel density & improves SR.
closer samples (increased sampling frequency), smaller sampling pitch, or distance BTWN sampling points |
| How does pixel pitch affect pixel size? | increased pixel pitch results in increased pixel size and reduced spatial resolution |
| What is fixed sampling versus fixed matrix? | Fixed matrix - number of pixels/DELs/dexels remains constant despite physical size (affects spatial resolution)
Fixed sampling - always sampling the same points |
| Describe phosphor layer of the detector. | a layer of photostimulable phosphor that interacts with the x-rays to create the latent image
in CR - commonly used BaFl and Europium |
| How does DEL/dexel size affect spatial resolution of the detector? | smaller DELs means better spatial resolution, until you get to the 80% fill factor. Then the spatial resolution requires too much mAs to justify it. |
| What is fill factor? | surface area on the dexel/DEL that absorbs/detects radiation for image production 80% (sweet spot)
20% remaining does not contribute to image - components like TFT and capacitor |
| What is the relationship of spatial resolution to DEL size? | smaller DELs, improved spatial resolution |
| Define LUMINESCENCE | the emission of light when stimulated by radiation |
| Define PHOSPHORESCENCE | emission of visible light during and after stimulation |
| Define FLUORESCENCE | emission of visible light only during stimulation |
| What is turbid phosphor and how does it affect spatial resolution? | Random size and spread of crystals - creates light spread - lower spatial resolution
Most CR phosphors are turbid |
| What is structured phosphor and how does it affect spatial resolution? | Linear, columnar filaments that funnel light down to next layer - increased spatial resolution |
| Are flat panel detectors direct or indirect capture? | BOTH direct and indirect
all have TFT array/AMA array composed of dexels/DELs |
| How do we increase spatial resolution with DELs/dexels? | decrease DEL size to increase SR nn increase in mAs to get enough SNR
If we make the dexels smaller, the fill factor will go down. The 20% part cannot be made smaller, so we need to increase the mAs to decrease the SNR from the smaller fill factor |
| Describe DR plate elements. | Capture element
Coupling element
Collecting element |
| What is capture element? | Orth p307
the material that captures the x-ray photons
in CR, it is the photostimulable phosphor
in DR, it is the NaI, CsI, and GdOS for indirect
Direct DR - amorphous Selenium |
| What is conducting element? | Orth p273
Transfers the x-ray signal to the collection element;
CR - lens system or fiber optics;
Indirect DR - CCD or fiber optics or Contact Layer of amorphous silicon;
Direct - amorphous selenium |
| What is collecting element? | Orth p274
Collects the signal
CR - photodiode, photodetector, PMT, CCD (light sensing)
INDIRECT - CCD (scintillator phosphor) or TFT (charge sensing device)
DIRECT - TFT (collects electrons) |
| Describe elements of CR plates | Capture element - BaFl PSP (Europium) p266 Orth
Coupling element - lens(optical component)/fiber optics p270 Orth
Collecting element - photodetector (PMT, CCD, photodiode) |
| Describe elements of early Indirect DR plates. | Capture element - CsI
Coupling element - fiber optic bundles p274 Orth
Collecting element - CCD/CMOS |
| Describe elements of later Indirect DR plates | Capture element - CsI and GdOS p274 Orth
Coupling element - contact layer (amorphous silicon semiconductor) painted onto the glass substrate
Collecting element - TFT |
| Describe elements of Direct DR | Capture element - amorphous selenium p275 Orth
Coupling element - NONE
Collecting element - TFT |
| What is PMT? | photomultiplier tube - a light sensitive device used to convert the light into electronic signal |
| What is photostimulable luminescence? | secondary release of light;
phosphor releases trapped energy as light when re-excited by infrared or visible light
Helium Neon laser causes this in CR reader |
| Describe latent image. | invisible image captured on IP (CR or FPD (until activated by CR reader or processed by digital system to become visible image) |
| Describe PREPROCESSING. | a. segmentation/exposure field recognition/data recognition
b. Histogram analysis
c. Automatic rescaling/normalization (RISK of FAILURE)
d. LUT |
| What is a TFT? | Thin-film transistor - a charge collection electrode (senses ELECTRONS) - it is a switch that is closed during exposure, allowing the charge in each DEL to be collected and stored. |
| Where is the risk of failure? | AUTO-RESCALING - normalization - editing image during preprocessing (choosing incorrect histogram, tech uses incorrect technique and overexposes patient but still gets a good image) |
| Describe SEGMENTATION/EXPOSURE FIELD RECOGNITION. | Partitioned pattern recognition; determining # & orientation of views on IR. (CR)
Sys recognizes collimated borders of field smaller than IP, which allows computer to EXCLUDE this info.
Sys makes distinction btwn usable densities or scatter. |
| Describe HISTOGRAM ANALYSIS/construction. | A process in which a computer analyzes histogram using processing algorithms and compares with preestablished histogram specific to anatomic part imaged. |
| Describe HISTOGRAM ERROR. | analysis errors that cause issue to be rescaled, thereby affecting the curve of histogram; displayed image will have excessive gray scale/contrast |
| Describe SEGMENTATION for CR. | Partitioned pattern recognition.
sys determines nbr and orientation of each view on the IR.
Fields MUST be even numbered, and symmetrical.
RESTRICTIONS: scatter, artifacts, must MASK off non-used areas |
| describe CR - image artifacts | Errors that are specific to digital imaging - during extraction of latent image - scatter, image artifacts (double xpos, clothing, IV lines) |
| What is SEGMENTATION FAILURE? | errors creating image during segmentation because system doesn't recognize actual exposure field. Place the part as much as possible in the CENTER aligned lengthwise.
when views are non-symmetrical - |
| Describe histogram construction | histogram is a graphic representation of the optimal patient tissues brightness within a collimated area.
constructed by counting # of dexels, then counting the number of grays |
| Describe histogram graph. | Horizontal axis X - rep value of each gray level; Left represents whiter/lighter grays - Right rep darker/blacker grays (fat/gas/air);
Middle rep soft tissue;
Vertical axis Y - rep # of pixels;
typically is bell curve with spike on right |
| Are collimated field borders important to image processing? | YES
Segmentation needs the borders to exclude unnecessary info |
| What is algorithm? | PP p5
formula or set of rules for solving problem |
| What is histogram? | PP p5
graphical representation of different body tissues
SPATIAL RESOLUTION |
| What is LUT? | PP p5
Look Up Table
histogram of luminance values (image display) for pixel values CONTRAST
p165 Papp
maps the image grayscale values into some visible output intensity on the monitor |
| Describe Segmentation. lecture 2hr 20min | 1st step - Segmentation - feature that allows system to pull info off IR -
Partitioned Pattern Recognition; scan the plate, look for data, ID # of exposures, extract information and apply to histogram analysis |
| How does tech affect PREPROCESSING? | Selecting correct exam (choose unique histogram) tells system how to process image. If choose incorrect exam, system must post-process data resulting in loss of info, unless repeat;
choose correct technique;
Steps - select exam, histogram applied, LUT |
| Describe Ideal, appropriate and inappropriate collimated fields. | IDEAL - all four borders are visible and CENTERED on IR
APPROPRIATE - 2/3 borders visible and SYMMETRICAL
INAPPROPRIATE - 1/2 borders visible and OFF CENTER |
| Describe HISTOGRAM ALGORITHM function. | displays anatomic part/tissues |
| Describe LUT ALGORITHM function. | changes pixel values to display appropriate contrast |
| What is HISTOGRAM ANALYSIS? | system compares image data histogram with stored image histogram
reads data from left to right (low to high)(white to black) |
| On a histogram graph, what does the white side represent? | White grays are on the left side and represent collimation, scatter then bone
Possible prosthesis/Barium which can cause histogram analysis error - use correct HISTOGRAM |
| On a histogram graph, what does the black side represent? | Blacker grays are on the right side and represent no anatomy, O/S skin margins, AIR FAT GAS |
| Describe histogram elements. | Histogram peaks
Histogram troughs
Width of histogram |
| What is exposure latitude in a histogram? | Increase kVp decrease mAs = converts absorption to transmission;
Flatter and wider histogram;
range of techniques that can be used to produce image w/o violating ALARA |
| Describe histogram contrast. | Affected by kVp driven by LUT
more shades shown with increased depth |
| What is attenuation? | attenuation is loss of energy of photons |
| What is VOI? | Occurs during preprocessing with LUT
volume of interest
VALUE OF INTEREST (Mrs. Lofton prefers)
anatomic part histo compared to stored histogram; ID collimated field to exclude extraneous values (not part of image) when rescaling |
| What is ROI? | region of interest - occurs on computer |
| what does kVp influence? | Increase kVp decreases subject contrast - flattens and widens histogram (fewer peaks/valleys - flat)
Decrease kVp increases subject contrast - more peaks/valleys, narrower histogram |
| Histogram analysis shapes | Shape stays fairly constant for each part exposed (anatomy specific
IE - Shape of histogram for a chest radiograph on a large adult is different than a ped knee |
| Define PREPROCESSING | image acquisition processing takes place in computer using processing algorithms (segmentation, histo analysis, auto-rescale, LUT) that analyze and ID components and characteristics of image |
| Define POST PROCESSING. | all manipulation and adjustments done by the tech through various user functions after pre-processing has occurred (once visible image is displayed)
Tech must be careful not to POST PROCESS necessary data from image. |
| What is the difference in CR/DR during pre-processing? | Only the detector plate (CR uses IRD/CR reader)
All other steps are the same
SEGMENTATION
HISTOGRAM ANALYSIS
AUTO RESCALE
LUT |
| What is the purpose of Automatic Rescaling/Normalization | p298/299 Orth/p164 Papp
to display image with appropriate brightness level regardless of the exposure
tech FAILURE possible here - select correct TECHNIQUE (ALARA) |
| What is gradient processing, rescaling image brightness, rescaling image contrast? | auto rescaling process |
| Why does the image initially appear as a washed out gray image? | due to DYNAMIC RANGE - the computer processes millions of shades of gray |
| How much dynamic range can the computer work within? | exposure of 500% over ideal and 60% below ideal |
| What does auto-rescaling do? | It stretches the pixel values toward optimal image values (stored histogram)
Can also be an error in processing |
| When does auto rescaling fail? | It can fail when exposure field is not correctly detected by pattern recognition software - will not correctly recognize the data (magnifies the error) |
| What is a LUT? | Processing algorithm applied to the data set of the image based on specific assumptions about the histogram which determines the final displayed contrast (image contrast).
reference histogram as a cross reference to transform raw data |
| What are the types of LUT? | A Priori
1. Extremity model - type 1 (common)
2. General model - type 2 (common)
3. Metal model - type 3 (fixed algorithm)
Neural - AI |
| What is stage 1 error ? | any error where system does not recognize data or collimation |
| What is LUT used for? | to correct values of pixels;
to give desired final image contrast |
| What is mapping function of LUT?
KNOW THIS | all pixels are changed to new gray value
image will have appropriate brightness and contrast
LUT is a table that maps the image grayscale values into some visible output intensity |
| How is LUT graphed? | original values - horizontal X axis;
new values - vertical Y axis |
| What is window width? | changing slope of LUT graph to adjust contrast |
| What is window level? | moving the slope of graph up and down the Y axis - adjusts brightness (Pixel Values) |
| What is exposure latitude? | The range of techniques that will produce a quality image that has an acceptable appearance (low noise) and does not violate ALARA |
| What is the biggest difference between digital and film screen? | ability to manipulate pixels |
| Describe WINDOW WIDTH. | Varies the range of shades of gray visible in the image
Wide window width - lowers contrast (more grays avail)
Narrow window width - increases contrast (fewer grays avail) |
| What is electronic collimation? | Altering the regions viewed on displayed image (masking) |
| What are the effects of excessive processing? | Degrades visibility of specific structures.
Increases chances of misdiagnosis |
| What controls contrast resolution? | PIXEL BIT DEPTH |
| describe image enhancement | LOCAL spatial frequency adjustment
generating image that is pleasing to observer
Includes - contrast, edge, spatial and frequency filtering, image combining and noise reduction |
| Describe Windowing | Local spatial frequency processing
to make changes to contrast and brightness of an image
Window width - CONTRAST
Window level - BRIGHTNESS |
| What is Edge Enhancement (high pass filtering) | Local spatial frequency processing
sharpens a blurred input image in spatial domain |
| What is smoothing (low pass filtering)? | local spatial frequency processing
a low pass filter operates on the input image with the goal of smoothing , resulting in blurred output image
Used to decrease noise in image; compromises image detail |
| What is unsharp masking? | Local spatial frequency processing that uses the blurred image from smoothing, subtracts it from the original image to produce a sharp image. |
| What is DOSE CREEP? | slow rise of patient dose in digital x-ray studies which occurs over time due to digital's ability to compensate for overexposure |
| How do you avoid DOSE CREEP? | p288 Orth
1. kVp 15% rule
2. kVp is seen as exposure and provides adequate data to computer
3. Increasing kVp maintains contrast resolution while decreasing patient dose
4. Very low kVp = quantum mottle = repeat |
| What does autorescaling do to an overexposed image? | p298 Orth/PP p8/graph 22.6 p299
processing stretches values toward optimal exposure (values that are more appropriate for specified anatomy) brings blacks to grays
SUBTRACTS values from pixels in image to align with stored histo |
| What does autorescaling do to an underexposed image? | processing stretches image pixel values toward optimal exposure - but image will appear mottled or with lots of whites
ADDS values to image pixel values |
| What is rescaling image brightness? | p298 Orth/PP p8/graph 22.6 p299
algorithms are applied to actual data from projection taken to align its position with that of the idea reference histogram. |
| Describe image artifacts | Any unwanted image on pic
- may obscure anatomic info
- PLUS density (B)
MINUS density (W)
GHOST/2x exp's
Scratches/tears
fogging
noise
heat blur
histo error
non-parallel collimation
linear scanning artifacts
Moire
Aliasing
Light bulb |
| What is PLUS density? | blacks on image |
| What is MINUS density? | whites on image |
| What is GHOST artifact? | electronic memory artifact |
| What is MOIRE pattern? | in CR, can occur when sampling frequency is not uniform or when GRID and Scan Frequency are similar and oriented together resulting in a zebra pattern |
| Tell me about Quantum Noise. | NOT ENOUGH PHOTONS REACH RECEPTOR
brightness fluctuations; photon dependent ; completely useless; more visible in digital imaging |
| What are exposure indicator scales? | ranges set by manufacturer and radiologists to help tech determine if image is acceptable
S# - Fuji, Konica, Philips INVERSE
EI# - Kodak - DIRECT
LgM - Agfa - DIRECT .3 |
| Describe inversely proportional scales (S#). | Center of scale represents speed class.
S value is low - OVERexposure
S value is high - UNDER exposure |
| Describe Logarithmic scales/exposure indicator. | Based on 300 point system
Increase EI by 300 = 2x exp
Decrease EI by 300 + 1/2 exp
to double receptor exposure, double the mAs
to halve receptor exposure, 1/2 mAs |
