Ch. 20 Carlton

Historical Development

• 1970s-1990s: Digital computerization of CT and Ultrasound (U/S).

• Creation of “electronic data sets” applied to all imaging modalities.

• Visualization in multiple planes and computer manipulation via post-processing.

Types of Digital Radiography Systems

• Computed Radiography (CR): Uses photostimulable imaging plates (PSP, IP).

• Digital Radiography (DR):

  • Direct conversion (without scintillator).

  • Indirect conversion (with scintillator).

  • DR receptors are flat panel detectors.

• Healthcare reimbursements encourage DR migration.

Digital Image Formation

• Electrical signals from receptors are in analog format.

• Must be converted to digital language

  • Analog-to-Digital Conversion (ADC) is required.

• Binary machine language:

  • Two-symbol alphabet: 0 and 1.

  • Bit versus Byte: 8 bits = 1 byte.

ADC

• Involves two important steps sampling and quantification.

• Digital data consists of bit values (binary digit).

• Bit strings connected as bytes.

• Computer memory expressed in total bytes (Megabytes, Gigabytes, Terabytes).

Digital Image Characteristics

• Matrix: Made of Pixels and Voxels.

  • Field of View (FOV).

  • Spatial resolution depends on matrix size.

• Pixel: Picture element.

• Voxel: Volume element.

Pixel Pitch and Size

• Inversely related to spatial resolution.

• Sampling frequency expressed as pixels/mm.

• Dependent on matrix size and image receptor size.

Bit-Depth

• Each pixel contains bit-depth

• Determines number of assigned gray shades to pixel value.

• Greater bit depth yields greater range of gray shades.

• 2^12 bit depth yields 4,096 gray shades.

• Greater bit depth increases data set size and volume data for image processing.

Image File Size

• Affected by pixel size, matrix, and bit depth.

Gray Scale Bit Depth

• Ranges from 8 to 32 bits.

• Digital Imaging and Communications in Medicine (DICOM) is the digital standard for imaging.

Histogram Characteristics

• Graphical computation of signal values of data set.

• Arranges signal values from minimum to maximum.

• Shape varies depending upon anatomy and exposure. (Short scale vs. long scale)

• Used to identify Pixel Values of Interest (VOI) and Exposure Index (EI).

Histogram Rescaling

• Acquired histogram compared to reference histogram.

• Rescaling compensates for under/over exposure by shifting histogram.

• Produces consistent image appearance regardless of exposure.

• Excessive overexposure can’t be compensated (Dose Creep).

Look-Up-Table (LUT)

• Adjustments to image contrast.

• LUT values assigned to data points (pixels) in histogram.

• Produces contrast-look according to reference contrast scale for exam view.

• LUT’s vary between exams and manufacturers.

• Sub-optimum exposure values cannot be compensated with LUT’s.

LUT Adjustments

• Image display adjusted from LUT (graph of processed pixel values).

• Permits changes in optical density or contrast.

• Can enhance pathologies; used when pathologies difficult to visualiz.

• djustment similar to changing DLOGE curve of film emulsion.

Digital Image Quality

• Spatial Resolution: Matrix size increases, pixel size decreases, spatial resolution increases.

• Density Resolution: Gray scale bit depth increases, density resolution increases.

• Contrast Resolution: Low Contrast Resolution (LCR) is the ability to represent small energy values; CR/DR receptors have excellent LCR and greater dynamic range.

Noise

• Undesirable signal values impair diagnostic value.

• Systems can suppress noise to a point, including electronic (system) noise and quantum noise.

• Measured as Signal-to-Noise Ratio (SNR).

Detective Quantum Efficiency (DQE)

• Measure of how sensitive and accurate incoming data is converted to output viewing.

• DQE = (SNRo)^2 / (SNRi)^2

• DQE of 1 = 100% or no loss of information.

• DR systems have a DQE between 30% and 70%.

• Higher DQE means lower dose.

Exposure Index (EI)

• Provides information about exposure to image receptor.

• Acceptable ranges for best image quality expressed as Target Exposure Indicators (EIT).

• Calculated using histogram values and Pixel Values of Interest (PVOI) mid-points.

• Varies between vendors.

• Responsible for understanding and applying EI values for image quality assessment and ALARA compliance.

Exposure Index Systems

• No universal system; different manufacturers use different systems.

• Comparing Sensitivity number (S number), Exposure Index (EI), or Log Median Exposure (LGM) values.

Deviation Index (DI)

• Created by AAPM and IEC to standardize exposure values between manufacturers using EI#’s.

• Uses department EIT values.

• Comparison between acquired exposure (KIND) and target exposure (KTGT).

• Expressed as a scale of DI values.