Quiz 3 Study Guide – Digital Radiography
⸻
Preprocessing vs Postprocessing
• Preprocessing = happens before the image is displayed
• Corrects raw data from the detector
• Examples: flat-field correction, dead pixel correction
• Postprocessing = happens after image acquisition
• Adjusts image appearance
• Examples: brightness, contrast, edge enhancement
⸻
Postprocessing Domains
Spatial Domain
• Works with pixel location
• Affects detail/resolution
• Example: smoothing, edge enhancement
Intensity Domain
• Works with pixel brightness values
• Affects contrast
• Example: windowing, LUT
Frequency Domain
• Works with patterns (frequencies) in the image
• Separates noise vs detail
• Example: filtering (remove noise or enhance edges)
⸻
Histogram Analysis
What the Histogram Represents
• Graph of pixel intensity distribution
• X-axis = brightness (black → white)
• Y-axis = number of pixels
⸻
Types of Histogram Analysis
• Type 1
• Simple exams (extremities)
• Single peak
• Type 2
• Two main tissue types (chest)
• Two peaks
• Type 3
• Complex anatomy (abdomen)
• Multiple peaks
• Neural (AI-based)
• Uses pattern recognition instead of fixed shapes
⸻
Construction of Histogram
1. Image acquired
2. Exposure field recognized
3. Pixels analyzed
4. Histogram created from pixel values
⸻
Histogram Analysis Errors
• Wrong body part selected
• Collimation too wide/narrow
• Artifacts (prosthetics, shielding)
• Multiple exposures on one plate
👉 Leads to incorrect brightness/contrast
⸻
Look-Up Table (LUT)
• Converts pixel values → visible grayscale
• Controls contrast and brightness appearance
• Different LUTs = different exam types
⸻
Dynamic Range Compression (DRC)
• Reduces wide exposure range into visible range
• Helps see both dark & light areas
⚠️ Effects:
• Can reduce contrast
• Can hide pathology if overused
⸻
Smoothing
• Reduces noise
• Makes image look softer
• ↓ spatial resolution
⸻
Edge Enhancement
• Increases sharpness
• Highlights borders
• Can increase noise/artifacts
⸻
SNR vs CNR
• SNR (Signal-to-Noise Ratio)
• Signal vs background noise
• Higher = cleaner image
• CNR (Contrast-to-Noise Ratio)
• Difference between structures vs noise
• Higher = better visibility of anatomy
⸻
Dose Creep & Exposure Index (EI)
• Dose creep
• Gradual increase in exposure over time
• Happens because images still look good even at higher dose
• Exposure Index (EI)
• Indicates how much radiation reached detector
• Used to monitor proper exposure
⸻
Segmentation
• Identifies area of interest
• Removes background from analysis
• Important for accurate histogram
⸻
Grid Line Suppression
• Removes visible grid lines digitally
• Prevents moiré patterns
⸻
Kernels
• Mathematical filters applied to image
• Types:
• Smooth (reduce noise)
• Sharp (increase detail)
⸻
Postprocessing Controls (Tech Can Adjust)
• Window level (brightness)
• Window width (contrast)
• Magnification
• Edge enhancement
• Smoothing
⸻
Sequence of Preprocessing Events
1. Exposure detection
2. Analog → digital conversion
3. Flat-field correction
4. Dead pixel correction
5. Exposure field recognition
6. Histogram creation
7
44