Digital Image Characteristics, Receptors and Image Acquisition Notes

Digital Image Characteristics, Receptors, and Image Acquisition

Objectives

  • Compare and contrast the attributes of digital imaging systems.
  • Explain the digital characteristics of matrix and pixels.
  • Recognize the relationship among pixel size, field of view (FOV), and matrix size.
  • State the relationship between spatial frequency and spatial resolution and how it relates to sampling frequency.
  • Differentiate between computed radiography (CR) and direct radiography (DR) image receptors (IRs).
  • Define signal-to-noise and contrast-to-noise ratios and explain their importance to digital image quality.
  • Explain histogram analysis, automatic rescaling, and look-up tables, and their role during computer image preprocessing to create a quality digital image.
  • Differentiate among the vendor-specific types of exposure indicators.
  • Compare and contrast the types of display monitors used for diagnostic interpretation and image reviewing.

Digital Image Characteristics

  • The latent image is stored as digital data and must be processed by the computer for viewing on a display monitor.
  • Digital Radiography Types:
    • Computed Radiography (CR)
    • Indirect Digital Radiography
    • Direct Digital Radiography
  • Each system includes:
    • Capture element
    • Coupling element
    • Collection Element
  • Matrix: combination of rows and columns. Digital image receptors have a matrix that correlates to a display matrix.
  • Pixel: "picture elements" or smallest component of the matrix.
    • Recorded as a single numerical value representing a brightness level on a display monitor.
    • The location of a pixel in the matrix corresponds to an area within the patient or volume of tissue, called the spatial domain, so it can be placed at the correct "address" on the image.

Matrix Size

  • Digital image quality is improved with a larger matrix size (greater number of smaller pixels).
  • Computer processing time, network transmission time, and digital storage space will increase with increased matrix size.

Pixel Size, FOV, and Matrix Size

  • This section seems to be intentionally left blank in the transcript.

Pixel Bit Depth

  • The numerical value assigned to a pixel is determined by the relative attenuation of x-rays passing through the volume of tissue and the resultant electrical signal.
  • The value is applied by an analog-to-digital converter (ADC) built into the system.
  • Pixel bit depth or number of bits (2n)(2^n) in the ADC determines the number of shades of gray that can be displayed.
  • The greater the bit depth, the higher the contrast resolution.
  • 1 bit = bit, 8 bits = 1 byte, 2 bytes = 1 word

Pixel Density and Pitch

  • The greater the number of pixels in an image matrix, the smaller their size.
  • Pixel density: a greater number of pixels per unit area.
  • Pixel pitch: distance measured from the center of a pixel to an adjacent pixel.
  • Increasing pixel density and decreasing pixel pitch increases spatial resolution.

Spatial Frequency

  • Spatial frequency capability of a system is measured in line pairs per millimeter (lp/mm).
  • Subject Spatial Frequency (SF) vs. Image Spatial Frequency (SF):
    • Subject SF: Small objects have higher spatial frequency (they’re closely spaced), and large objects have lower spatial frequency (spaced farther apart).
    • Image SF: The ability of the image to demonstrate anatomy with high subject spatial frequency.
  • Increasing the number of lp/mm resolved in the imaging system improves spatial resolution.
  • Systems that resolve higher spatial frequency have higher spatial resolution.
  • Typically limited by the size of pixels.

Dynamic Range

  • The ability of the detector to accurately capture and represent the range of photon intensities that hit the IR with the appropriate shades of gray on the image.
  • Digital image receptors have a wide dynamic range, meaning they can accurately detect a wide range of radiation intensities.
  • The computer can process the raw pixel data to compensate for exposure errors in digital imaging.
  • Lower or higher than necessary exposure techniques will not guarantee a quality digital image.
  • Large Dynamic Range = Long gray scale and higher contrast resolution.
  • Cannot correct for quantum mottle.
  • Exposure latitude refers to the variety of technical factors that could be utilized to still obtain acceptable images.
  • Even digital imaging has its limits with exposure latitude.

Dose Monitoring

  • Dose: amount of kinetic energy deposited in a volume of air; described in Air Kerma.
  • Dose Area Product (DAP): a digital system process that measures the air exposure, then computes an estimate of absorbed dose in the patient.
    • Depends on technique, FOV, and total tissue volume being irradiated.
  • Kerma Area Product (KAP): similar to DAP but uses the Air Kerma at the entrance to the patient.

Modulation Transfer Function (MTF)

  • Measurement of the imaging system’s ability to display the contrast and spatial resolution of anatomic objects of all sizes and brightnesses.
  • MTF is responsible for converting subject object contrast into contrast intensity levels in the image (image contrast).
  • MTF of 1 = 100%
  • High-frequency subject objects are harder to visualize, so most MTFs are much lower than 1.
  • Performed by an analog-to-digital converter.
  • Dependent on sampling frequency and bit depth.

Detective Quantum Efficiency (DQE)

  • DQE is a measurement of the efficiency of an image receptor in absorbing photons and converting exposure to a quality radiographic image.
  • The higher the DQE of a system, the lower the radiation exposure needed to produce a quality image; decreases patient exposure.
  • DQE of 1 = 100% or no loss.
  • Affected by the type of material used in the IR and the energy of transmitted photons.
  • Capture layer thickness and Atomic composition (k shell).
  • DR has better DQE than CR; direct DR has higher DQE than indirect at higher kVp levels.

Signal-to-Noise Ratio (SNR)

  • Describes the strength of the exposure (signal) in comparison to the amount of noise apparent in the digital image.
  • Increasing the SNR means that the strength of the signal is high compared to the amount of noise.
  • Increasing the SNR will improve the quality of the digital image.
  • Increased noise will decrease the visibility of anatomic details.
  • Noise caused by mottle and scatter.

Quantum Noise

  • Limited photons captured creating a low signal-to-noise ratio.
  • Noise reduces visibility.
  • Other noise sources: Electronic, Poor Capture, Poor display.

Contrast-to-Noise Ratio (CNR)

  • CNR describes the contrast resolution compared with the amount of noise apparent in a digital image.
  • Impacts contrast on the image.
  • Increasing the CNR increases the visibility of anatomic details.
  • Increasing scatter decreases CNR.

Digital Image Receptors

  • Computed Radiography (CR):
    • Two-step process for image acquisition: image capture and image readout.
    • Imaging plate (IP): Photostimulable phosphor.
  • Indirect Digital Radiography (DR):
    • Utilizes capture, coupling, and collection elements.
    • Relies on scintillation to be produced by the capture element and transferred to the collection elements via the coupling element.
  • Direct Digital Radiography (DR):
    • Combines image capture and image readout in one process.
    • Does not require scintillation to produce an image.
    • Coupling and collecting components are combined.
    • Utilize TFT flat panel detectors with amorphous selenium.

Computed Radiography (CR)

  • Imaging plate phosphor (Barium fluorobromide/fluorohalide and Europium) absorbs transmitted x-ray intensities and ionizes through photoelectric (PE) interaction.
  • Energy absorbed by PSP excites an electron, which becomes trapped in the conduction band just beyond the Valence shell (metastable state).
    • The number of electrons trapped in this state is proportional to the number of photons transmitted through the patient.
    • If these electrons escape the conduction band, light will be released.
    • Electrons remaining in the conduction band become the latent image to be processed.
  • The imaging plate is scanned with a red laser.
    • Electrons released to normal energy state during laser beam scanning in the reader unit produce light.
  • Light is captured and transmitted to the collecting component via the fiberoptic coupling component.

CR Reader

  • The IP is removed from the cassette and scanned with a helium-neon (red) laser beam or solid-state laser diode.
  • The laser releases trapped electrons, resulting in the release of light (Photostimulable luminescence).
  • Light is funneled to fiberoptic couplers that transfer it to the Collection system. On the way it is filtered.
  • Collection system:
    • Photomultiplier Tube (PMT) or Photodetector (PD) or Charge-Coupled Device (CCD).
    • Collects, amplifies, and converts the light to an analog electric signal proportional to the range of energies stored in the IP.
    • Sends the signal to the ADC.
  • ADC:
    • Samples the analog signal at different intervals (frequencies) for amplitude.
      • Higher sampling frequencies are used for increased pixel density and improved spatial resolution.
      • The closer the samples are together, the smaller the sampling pitch between pixels.
    • Emits a binary digital signal representative of the amplitude of the light signal (quantization).
      • Options for quantity based on bit depth.
    • The signal is sent to the computer to begin the histogram process.

CR Sampling (Cont.)

  • Sampling frequency must be at least 2X the highest spatial frequency of the incoming signals to be accurate.
  • The Nyquist Frequency is the sampling frequency necessary to accurately convert analog data to digital data.
  • Increasing sampling frequency increases pixel density and spatial resolution.

Fixed Sampling Frequency

  • Fixed matrix size.
  • Imaging Plate Size.

Indirect Conversion Detector

  • Scintillator-type detector converts exit radiation into visible light.
    • Capture element: Cesium iodide (CsI) or Gadolinium Oxysulfide.
    • Coupling element: Fiber optics or Amorphous Silicon.
  • Photodetector in the TFT, CCD, or complementary metal-oxide-semiconductor (CMOS) converts visible light to proportional electrical charges.
  • Electrical signals are sent to the ADC for digitization.
    • Some ADCs are now in the detection elements themselves!

CCD’s vs. CMOS

  • CCDs may show seaming (tiling) that must be accommodated for in preprocessing.
    • Tiling occurs when multiple CCD devices are joined, acting as one.
    • CCD is used with the following:
      • CsI capture
      • Fiberoptics/lenses coupling
      • CCD collection
  • CMOS
    • Scintillators -capture
    • Fiberoptics/lenses/crystal light tubes- couple
    • CMOS DEL- collects
    • Each DEL has its own amplifier, photodiode, storage capacitor and is surrounded by transistors
  • Use less power and are inexpensive compared to CCDs.

CR and Indirect Conversion Detector

  • May be structured or unstructured.
  • Structured reduces the spread of light
    • Needle-point crystals of CsI
    • Improved Spatial Resolution
  • Unstructured (turbid)
    • Not manufactured in columns
    • Blobs
  • The thicker the layer, the higher the DQE but the lower the spatial resolution.

Direct Radiography

  • Each detector element (DEL) is associated with transmitted radiation and converts its intensity to proportional electrical signals for digitization.

Direct Digital Radiography

  • Flat-panel detectors are solid-state image receptors with a large active matrix array of electronic components.
  • The detector contains layers to receive radiation and convert photons to electrical charges.
    • Signal storage, signal readout, and digitizing electronics are integrated into the flat-panel device.
  • IR is made of several layers:
    • X-ray converter (capture element) - Amorphous Selenium
    • Coupler - Amorphous Selenium
    • Collector - Thin-film transistor (TFT) array
  • TFT:
    • Divided into square detector elements.
    • Each detector element (DEL) has a capacitor to store electrical charges and a switching transistor for readout.
    • Electrical charges are read out separately from each DEL and then sent to the ADC for digitization.
    • Each DEL corresponds to a smaller pixel in the matrix.
    • Each DEL only captures a percentage of the x-rays, known as the fill factor.
    • Decreased DEL size improves Spatial resolution but also decreases fill factor.
    • Glass substrate

Direct Conversion Detector

  • Amorphous selenium-coated detector converts exit radiation directly into electrical charges.
    • Captures and Couples
    • Selenium has a fairly low atomic # (Z=34Z=34), so it is thick (1mm) to maximize absorption.
    • The selenium layer has an electric field to limit the lateral spread of electrons.
  • Collection TFT creates an electrical signal, then sends it through the ADC
    • Stored in the capacitor.
    • Next, it goes on to be amplified and processed in the computer.

Direct IR

  • May be permanently mounted in a Bucky system or mobile.
  • Highly dose efficient.
  • Image is immediately available.
  • Spatial resolution is improved over CR.
  • Spatial resolution is limited by DEL size.
  • Very expensive.
  • Manufactured to operate in indirect or direct conversion.

The Digital Image

  • Regardless of the type of digital image receptor, the varying electrical signals are sent to the ADC for conversion to digital data.
  • The digitized pixel intensities are patterned in the computer to form the image matrix.
  • The image matrix is a digital composite of the varying x-ray intensities exiting the patient.
  • Each pixel has a brightness level representing the attenuation characteristics of the volume of tissue imaged.

Digital Image Processing

  • After ADC, digital data must be processed before display.
  • Digital image processing describes various computer manipulations that optimize the appearance of an image.
  • An essential process is Histogram Analysis:
    • Image processing technique to identify the edges of the image and compare to a stored pre-established histogram specific to the anatomic part imaged.
    • All digital images have the ability to evaluate original image data through histogram analysis.
    • A histogram is a plot of the frequency of appearance of a given object’s characteristics.
    • It is a discrete plot of values developed off the study and averages of many exposures of the same anatomy.

Histograms

  • Each image has a distribution of values of pixels.
    • Averages of these distributions over many images are obtained to produce a histogram.
    • The histogram for each image type (knee, hand, chest, spine, etc.) is unique to that part.
    • Alternative histograms may be applied to images.
    • Radiographers must select the most appropriate histogram.
  • The computer processes the histogram using processing algorithms and compares it to a pre-established histogram specific to that anatomic part.
  • Values of Interest (VOI): the range of histogram data set that should be included in the displayed image.
    • If edges are not properly recognized, a histogram error may occur.
  • Each histogram will have a unique shape.
  • Attempts to maintain common brightness in under/overexposed images with automatic rescaling.

Histogram Analysis Error

  • In CR imaging, the entire imaging plate is scanned to extract the image from the photostimulable phosphor.
  • If at least three edges are not identified, a histogram analysis error could occur.
  • Artifacts, improper collimation, poor alignment, and poor centering.

Histogram Analysis

  • Maintains consistent image brightness despite overexposure or underexposure to the image receptor.
  • Automatic rescaling: a process of mapping the gray scale to the VOI to present a specific display of brightness.

Histogram Analysis (Cont.)

  • Exposure indicator: provides a numerical value indicating the level of radiation exposure to the digital image receptor.
  • Exposure indicator values should be within the optimum range for that digital imaging system.
  • Optimal ranges are vendor-specific and may vary between different parts imaged.

Deviation Index (DI)

  • Recommended universal standard exposure indicator.
  • DI is a value that reflects the difference between the desired or target exposure to the IR and the actual exposure to the IR.
  • A DI of 0 indicates no difference between the desired EI and the actual EI.
  • A DI above 0 indicates increased exposure.
  • A DI below 0 indicates decreased exposure.
  • Department standards should be followed for how over- and underexposures are handled before repeating the image.

EI/DI Limitations

  • Collimation
  • kVp
  • Centering
  • Inaccurate anatomic part selection
  • Inaccurate data extraction or computer rescaling
  • Evaluate noise along with EI/DI numbers

Lookup Tables

  • A method of altering the image to change the display (brightness or grayscale) of the digital image.
  • Use computer algorithms.
  • Original pixel values can be altered to change the display of the digital image.
  • May allow image inversion.

Lookup Tables (Cont.)

  • If the original image is altered, the original pixel values would be different in the processed image, and the graph would no longer be a straight line.
  • The pixel intensities would be calculated to display an image with a different contrast level.

Digital Imaging Artifacts

  • Numerous causes.
  • Usually a bright or dark appearance.
  • Often due to the complexity of electronics.
  • Can be classified as:
    • Prior to ADC
    • During ADC
    • After ADC
    • Image display
  • Examples:
    • CR: particulates, scratches, cracks, and fogging
    • DR: Calibration of DEL, rows or columns, electronic readout malfunction or electronic components after ADC

Image Display

  • Soft copy viewing of the digital image occurs at a computer workstation with specialized postprocessing software.
  • The performance of the display monitor affects the quality of the image.
    • Luminescence, resolution, ambient lighting, etc.
  • Primary display monitors used for diagnostic interpretation must be of high quality.

Display Monitors

  • Cathode Ray Tube (CRT):
    • Creates an image by accelerating and focusing electrons through an electron gun to strike the faceplate, composed of a fluorescent screen.
    • The image is scanned on the screen in lines.
      • The number of lines affects image quality. Must be at least 525 lines per 1/30 second.
    • Typically have a curved faceplate.
    • Being replaced by LCD or plasma
  • Liquid Crystal Display (LCD):
    • Passes light through liquid crystals to display the image on the glass faceplate.
    • Electrical signals can vary the light waveforms as they pass through the crystals for viewing on the faceplate.
    • Flat screen/thinner dimension
  • Plasma monitors: similar to LCD but with pixels instead of crystals.

Display Monitors (Cont.)

  • Placement of display monitors and ambient lighting can affect soft copy viewing of digital images.
  • The monitor matrix size should be at least as large as the image matrix size.
  • Monitors that have a higher luminance ratio are capable of displaying a greater gray-scale range.
  • Pixel depth affects contrast.
  • Monitors with higher luminance can display greater grayscale.
  • Visualization can be affected by spatial resolution, luminance, and contrast resolution.

Performance Criteria

  • QC devices:
    • Require American Association of Physicists in Medicine (AAPM) or Society of Motion Picture and Television Engineers (SMPTE) Test patterns.
    • These patterns test for:
      • Geometric distortion
      • Luminance
      • Resolution
      • Contrast resolution
      • Noise
      • Veiling glare

Postprocessing

  • Software options are available to allow manual manipulation of the displayed image.
  • Electronic masking: altering the regions viewed on the displayed image (AKA Shuttering).
  • Window level: sets the midpoint of the range of brightness visible in the image.
    • Allows brightness to be increased or decreased throughout the entire image.
  • Window width: varies the range of shades of gray visible in the image.
    • Wide window width lowers contrast (all shades of gray visualized).
    • Narrow window width increases contrast (fewer shades of gray visualized).
  • If these changes are saved to the image, it may affect the radiologist’s ability to manipulate the image.

Other Postprocessing Techniques

  • Subtraction: removal of superimposed structures so the anatomic area becomes more visible
    • The technologist would select a certain brightness value to remove.
  • Contrast enhancement: alters pixel values to increase enhancement.
  • Edge Enhancement (high pass filtering): improves the visibility of small high contrast structures but may cause noise.
  • Inversion: reverses gray scale.
  • Smoothing: suppresses noise but degrades spatial resolution.
  • Equalization: underexposed areas are made darker, and overexposed areas are made brighter, lowering contrast.
  • ROI
  • Stitching

Digital Communication Networks

  • Picture Archival and Communication System (PACS) is a computer system designed for digital imaging that can capture, store, distribute, and display digital images.
  • Digital Imaging and Communication in Medicine (DICOM) is a communication standard between PACS and imaging modalities.
  • Hospital Information System (HIS) and Radiology Information System (RIS) are computer systems for medical information.
  • Health Level Seven Standard (HL7) is a communication standard for medical information.
  • Systems are linked by networks to allow access to all information in a convenient location in little time.

Summary

  • Digital systems rely on matrices to obtain and display images.
  • Matrices with smaller pixels have greater spatial resolution.
  • Spatial resolution is higher with high spatial frequency.
  • Bit depth determines the number of possible shades of gray and is a component of the ADC, which controls contrast resolution.
  • CR and DR IRs differ in construction, processes, and function.
  • The sampling frequency in the ADC is responsible for the accurate demonstration of anatomic parts and is higher with higher pixel density.