10.1 Gamma Camera Quality Control
Key Takeaways
- A scintillation gamma camera combines a large NaI(Tl) crystal, an array of PMTs, a collimator, PHA electronics, and a computer that maps event position and energy into images
- Intrinsic uniformity uses a point source without collimator; extrinsic (system) uniformity uses a flood or sheet source with the collimator—both are core daily QC
- Spatial resolution and linearity are checked periodically with bar or hole phantoms; LEHR favors resolution, LEAP/LEGP favors sensitivity
- Recognize classic artifacts: dead PMT (cold circle), cracked crystal (linear cold/hot defect), edge packing (bright rim), and wrong energy window
- For Tc-99m, set the photopeak window on 140 keV (commonly ±10% or ±20% per protocol); off-peak windows degrade contrast and can mimic pathology
10.1 Gamma Camera Quality Control
Quick Answer: Daily floods (intrinsic and/or extrinsic per SOP), correct Tc-99m 140 keV window, and collimator matched to the study. Spot PMT failure, cracked crystal, and edge packing before you image patients.
Domain IV continues with imaging systems. The Anger (scintillation) gamma camera remains the workhorse for planar and SPECT nuclear medicine. Technologists must know components, QC tests, collimators, energy windows, and artifact patterns at CNMT depth.
Camera Components
| Component | Role |
|---|---|
| NaI(Tl) crystal | Converts γ energy to scintillation light; large thin slab for imaging |
| Light guide / optical coupling | Couples crystal light to PMT faces |
| PMT array | Multiplies light → electrical pulses; weighted signals encode X–Y position |
| Collimator | Projects only selected photon directions onto the crystal (spatial encoding) |
| PHA / electronics | Accepts events in energy windows; rejects scatter outside the window |
| Computer / console | Digitizes events, stores matrices, applies corrections, displays images |
Without a collimator, photons strike the crystal from all angles and spatial information collapses. With a collimator, only photons traveling nearly parallel to the holes (for parallel-hole designs) form a useful projection of the activity distribution.
Collimator Types
Parallel-hole collimators dominate clinical work. Hole diameter, length, and septal thickness trade sensitivity against spatial resolution and determine usable photon energy.
| Collimator | Typical use | Teaching trade-off |
|---|---|---|
| LEAP / LEGP (low-energy all-purpose / general-purpose) | Many Tc-99m planar studies when counts matter | Higher sensitivity, coarser resolution |
| LEHR (low-energy high-resolution) | Bone, many SPECT Tc-99m protocols | Finer resolution, lower sensitivity |
| Medium energy | Ga-67, In-111 (medium-energy photons) | Thicker septa reduce penetration |
| High energy | I-131 imaging | Thickest septa for high-energy γ |
| Pinhole | Thyroid, small parts, magnification | Excellent resolution for small FOV; low sensitivity |
Trap: Imaging In-111 or I-131 with a low-energy collimator produces septal penetration (star artifacts, blurred edges). Always match collimator energy rating to the radionuclide’s primary photons.
Pulse-Height Analysis and Energy Windows
Each accepted event contributes to the image only if its pulse height falls inside the photopeak window. For Tc-99m, the primary photopeak is 140 keV. Common teaching windows are approximately ±10% (about 126–154 keV) or ±20% (about 112–168 keV), depending on manufacturer defaults and departmental protocol.
| Window issue | Effect |
|---|---|
| Window too wide | More scatter accepted → reduced contrast |
| Window too narrow / off-peak | Count loss, noisy images, apparent cold areas |
| Dual/multiple peaks (e.g., Ga-67, In-111) | Must set correct multi-peak windows per protocol |
Peaking: verify the photopeak is centered in the window (or auto-peak per SOP) before floods and patient studies. Drift from high voltage or temperature can shift the peak.
Uniformity (Floods)
Uniformity is the most frequent imaging QC test. A uniform activity field should produce a uniform image after corrections.
| Type | Setup | Purpose |
|---|---|---|
| Intrinsic | Collimator removed; point source far from crystal (≈5× FOV distance common teaching) on the crystal axis | Crystal + PMT + electronics uniformity without collimator effects |
| Extrinsic (system) | Collimator on; sheet/flood source (Co-57 sheet or Tc-99m fillable flood) on collimator face | Whole imaging chain including collimator |
Daily extrinsic and/or intrinsic floods (per manufacturer and department SOP) detect sudden PMT failure, cracked crystal, loose collimator, or correction-map problems. Quantitative integral and differential uniformity percentages are compared with action limits. Extrinsic floods also reveal collimator damage (dents, bent septa) that intrinsic tests miss.
Correction maps: modern cameras store energy and linearity correction tables. Floods failing after service may need remapping by field engineers—do not “normalize away” a broken detector for clinical use.
Spatial Resolution and Linearity
Spatial resolution is the ability to distinguish closely spaced activity (often characterized as FWHM of a line-spread function). Linearity is the straightness of imaged lines—nonlinear mapping warps anatomy.
| Phantom | What it tests |
|---|---|
| Four-quadrant bar phantom | Visual resolution at four bar frequencies; common weekly/periodic check |
| Orthogonal-hole / parallel-line equal-space (PLES) | Resolution and linearity patterns |
| Line sources | Quantitative FWHM measurements |
Resolution worsens with distance from the collimator face (especially parallel-hole). Keep the organ of interest as close as practical. LEHR shows finer bars than LEAP at the cost of counts.
System Sensitivity
Sensitivity is counts per unit activity under defined geometry, collimator, and window (e.g., cpm/µCi). Periodic sensitivity checks with a known source flag unexpected count loss (wrong collimator, wrong window, electronic failure). Sensitivity alone does not prove spatial resolution is adequate.
Common Artifacts
| Artifact | Appearance | Likely cause |
|---|---|---|
| PMT failure | Round cold (or hot, if gain wrong) region over one tube | Dead/failing PMT, HV, or preamp |
| Cracked crystal | Sharp linear or irregular defect; may change with angle | Mechanical shock, thermal stress; NaI is brittle and hygroscopic if sealed poorly |
| Edge packing | Bright ring at FOV edge | Light reflection / position algorithm edge effects |
| Collimator damage | Localized hot/cold streaks or hexagonal patterns | Dropped collimator, bent septa |
| Wrong window / off-peak | Low counts, mottling, false cold lesions | Energy calibration drift |
| Contamination on collimator/crystal cover | Focal hot spots on floods and patients | Spill residue |
Daily Workflow Checklist (Planar)
- Peak / energy window for the radionuclide in use (Tc-99m at 140 keV with the departmental % window).
- Intrinsic and/or extrinsic flood per SOP; review image and quantitative uniformity % against limits.
- Confirm correct collimator mounted and locked; inspect face for dents or contamination.
- Verify corrections (energy, linearity, uniformity maps) are current if the console flags them.
- For resolution days: acquire bar or hole phantom images and archive with QC log.
If any step fails, stop clinical imaging on that head/camera until corrected or an alternate system is validated. Document results, initials, and corrective action—accreditation and license conditions expect a paper or electronic trail.
Count-rate awareness: very high activities near the camera (hot spills, unshielded syringes) can produce pile-up and distorted images; shield sources and keep non-patient activity away from the detector face.
Memory aid: Daily flood + correct window + correct collimator prevent most planar disasters; pattern-match artifacts before blaming the patient.
Intrinsic flood uniformity on a gamma camera is performed with which setup?
A technologist images In-111 leukocytes with a low-energy high-resolution (LEHR) collimator. The most important expected problem is:
A daily extrinsic flood shows a well-defined circular cold region that was not present yesterday. The most likely cause is: