Residency · Residency · Nuclear Medicine

The Gamma Camera: Principles and Quality Control

Gamma Camera Components

Overall Design (Anger Camera)

The gamma camera, invented by Hal Anger in 1958, remains the fundamental imaging device in nuclear medicine. Despite decades of refinement, its core design principles persist. The camera consists of a collimator, a NaI(Tl) scintillation crystal, an array of photomultiplier tubes, position logic circuitry, an energy analyzer, and a computer display system. Together, these components convert the spatial distribution of a gamma-ray-emitting radionuclide within the patient into a two-dimensional image.

Collimator

The collimator sits between the patient and the crystal and defines the directions from which photons can reach the detector. Made of high-Z material such as lead or tungsten, it contains holes or channels that accept only photons traveling along specific trajectories. The collimator is the primary determinant of both spatial resolution and sensitivity, and there is an inescapable trade-off between these two parameters: improving resolution by making the holes smaller or longer necessarily reduces sensitivity by rejecting more photons.

Collimator Types

Several collimator designs serve different clinical needs. The parallel-hole collimator is the most widely used; it projects an image the same size as the object, though resolution worsens as the source moves farther from the collimator face. Converging collimators have holes that angle toward a focal point beyond the patient, magnifying the image and increasing sensitivity for small organs. Diverging collimators have holes that splay outward, minifying the image to accommodate large organs on smaller camera heads, though they are rarely used today. The pinhole collimator uses a single small aperture to magnify objects close to the collimator, providing excellent resolution for small superficial structures such as the thyroid and pediatric joints. Fan-beam collimators converge in one direction and are parallel in the other, finding application in brain SPECT imaging.

Collimator Selection by Energy

Collimator selection must be matched to the radionuclide's photon energy. Low-energy general-purpose (LEGP) collimators are designed for Tc-99m and balance resolution with sensitivity. Low-energy high-resolution (LEHR) collimators offer superior resolution at the cost of lower sensitivity and are preferred for bone scans. Low-energy high-sensitivity (LEHS) collimators sacrifice resolution for count rate and are used in dynamic studies. Medium-energy (ME) collimators have thicker septa to handle the higher-energy photons of Ga-67 and In-111. High-energy (HE) collimators, with the thickest septa, are required for I-131 at 364 keV to prevent septal penetration.

NaI(Tl) Crystal

The NaI(Tl) crystal is the detection element of the gamma camera. The standard thickness of 9.5 mm (3/8 inch) is optimized for absorbing 140 keV photons from Tc-99m. Thicker crystals improve sensitivity for higher-energy photons but degrade intrinsic spatial resolution. Crystal diameters typically range from 40 to 50 cm for general-purpose cameras. Because sodium iodide is hygroscopic, the crystal must be hermetically sealed to prevent moisture damage. Crystal cracking or hydration from a compromised seal produces irreversible damage visible as artifacts on flood field images.

Position Logic and Energy Discrimination

Behind the crystal, an array of 30 to 100 photomultiplier tubes detects the scintillation light produced by each gamma ray interaction. The Anger logic circuit calculates the X and Y coordinates of the interaction based on the relative signal amplitudes from different PMTs. The sum of all PMT signals, known as the Z-pulse, represents the total energy deposited by the photon. A pulse height analyzer then accepts only those events whose energy falls within the photopeak window, typically set at 15 to 20% centered on the photopeak energy. Events outside this window -- including most scattered photons -- are rejected.

Gamma Camera Performance Parameters

Spatial Resolution

Spatial resolution has three components. Intrinsic resolution reflects the resolving power of the detector itself (crystal plus PMTs) and is typically 3 to 4 mm for modern systems. Collimator resolution is determined by the collimator geometry and worsens with increasing distance from the collimator face. System resolution combines both components, added in quadrature: system resolution equals the square root of (R_intrinsic^2 + R_collimator^2). In clinical practice, collimator resolution is the dominant factor, so the overall system resolution is largely determined by the collimator design and the source-to-collimator distance.

Sensitivity

Sensitivity, measured in counts per unit time per unit activity, reflects how efficiently the gamma camera detects emitted photons. The collimator is the main determinant: only about 0.01 to 0.05% of emitted photons successfully pass through a parallel-hole collimator. Higher sensitivity permits shorter imaging times or lower administered doses.

Uniformity

An ideal gamma camera should produce a perfectly uniform image when exposed to a uniform flood source. In reality, non-uniformities arise from PMT drift, crystal defects, damaged collimator septa, or energy window miscalibration. Even small non-uniformities are amplified during SPECT reconstruction, producing ring artifacts in tomographic images. Uniformity is quantified as integral and differential uniformity according to NEMA definitions. Correction is achieved by acquiring high-count flood images and generating uniformity correction maps that are applied during image processing.

Energy Resolution

Energy resolution quantifies the detector's ability to distinguish photons of different energies. It is expressed as the full width at half maximum (FWHM) of the photopeak divided by the photopeak energy, multiplied by 100%. For NaI(Tl), energy resolution is approximately 10% at 140 keV. Better energy resolution translates directly into better scatter rejection and improved image contrast.

Count Rate Performance and Dead Time

At high count rates, the detector and electronics cannot process every event, and some counts are lost due to dead time. Two models describe this behavior. In a paralyzable (extending) system, each event restarts the dead time clock, so at very high count rates, the measured count rate can actually decrease and theoretically fall to zero. In a non-paralyzable (non-extending) system, dead time is fixed per event, and the count rate simply plateaus at a maximum value. Most gamma cameras exhibit mixed behavior. Count rate limitations become clinically relevant during first-pass cardiac studies and imaging immediately after therapeutic doses.

Quality Control Protocols

Daily QC

Daily quality control begins with a photopeak check, using a point source to verify that the energy window is properly centered on the photopeak. This is followed by a uniformity flood, acquired either intrinsically (without collimator, using a point source at distance) or extrinsically (with collimator, using a sheet source). A minimum of 2 to 5 million counts should be collected for visual assessment. The technologist examines the flood for non-uniformity patterns, PMT dropout (appearing as cold spots), or crystal damage. Integral uniformity should be less than 5% for the useful field of view (UFOV) and less than 3% for the central field of view (CFOV).

Weekly QC

Weekly quality control includes spatial resolution and linearity assessment using a four-quadrant bar phantom or orthogonal-hole phantom. The technologist verifies that straight lines in the phantom appear straight on the image (no warping or distortion) and assesses resolution by identifying the smallest resolvable bar groups.

Annual/Semi-Annual QC (Medical Physicist)

Comprehensive performance testing by a medical physicist includes full NEMA characterization: high-count intrinsic uniformity (30 to 100 million counts), intrinsic and system spatial resolution, energy resolution, count rate performance, center of rotation calibration for SPECT, sensitivity measurements, and multi-window spatial registration for multi-energy isotopes.

SPECT-Specific QC

SPECT imaging demands additional quality assurance measures. Center of rotation (COR) calibration ensures that the axis of camera rotation aligns with the reconstruction axis; misalignment causes characteristic blurring artifacts in reconstructed images. A high-count flood (greater than 30 million counts) is acquired to generate the uniformity correction map used during SPECT reconstruction. SPECT phantom imaging with a Jaszczak phantom (containing cold spheres and resolution rod sections) verifies overall system performance including uniformity, resolution, and contrast.

<image>An exploded-view diagram of an Anger gamma camera showing all major components in layers from top to bottom: patient, collimator (showing parallel-hole design with lead septa), NaI(Tl) scintillation crystal, optical light guide, array of photomultiplier tubes in hexagonal arrangement, position logic electronics, and computer display. Arrows should trace the path of a single gamma ray from the patient through the collimator, interaction in the crystal, light production, PMT signal generation, and final image formation. Include labels for each component.</image>

<image>A four-panel comparison of collimator types used in nuclear medicine: parallel-hole (showing equal object and image size), converging (showing image magnification), diverging (showing image minification), and pinhole (showing inverted magnified image). Each panel should show a cross-section of the collimator geometry with ray tracings from a source through the holes to the crystal. Include annotations about typical clinical applications for each type: parallel-hole for general imaging, pinhole for thyroid, converging for brain SPECT.</image>

<image>A quality control reference sheet showing examples of normal and abnormal gamma camera flood field images. Include: (1) a normal uniform flood, (2) a flood showing a photomultiplier tube defect (cold spot), (3) a flood showing cracked crystal artifact (linear cold streak), (4) a flood showing collimator damage (focal cold defect), and (5) a bar phantom image showing normal resolution pattern and one showing resolution degradation. Label each abnormality clearly with the likely cause.</image>

Clinical Pearls

The collimator is the weakest link in the gamma camera imaging chain. It alone determines both spatial resolution and sensitivity, and every design is a compromise between these two competing demands.

Matching the collimator to the radionuclide energy is essential. Using a LEHR collimator (designed for Tc-99m) to image I-131 causes severe septal penetration artifacts and unacceptable image degradation because the 364 keV photons pass through the thin septa meant for 140 keV photons.

Non-uniformity of even 1 to 2% in the flood field can produce significant ring artifacts when data are reconstructed into SPECT images. This extreme sensitivity to uniformity is why daily flood QC is mandatory before any SPECT acquisition.

Pinhole collimators are the collimator of choice for thyroid imaging because they magnify small superficial organs, dramatically improving effective spatial resolution compared to parallel-hole collimators at the same source distance.

Center of rotation misalignment is a frequent and treatable cause of degraded SPECT image quality. It produces a characteristic blurring of point sources in reconstructed images and should be checked regularly.

For board examinations, remember that spatial resolution worsens with increasing source-to-collimator distance for parallel-hole collimators. The practical implication is that the patient should always be positioned as close to the collimator face as possible.

Crystal hydration from a breach in the hermetic seal produces irreversible damage visible as non-uniform regions on flood images. The only remedy is crystal replacement.

References

• Cherry SR, Sorenson JA, Phelps ME. Physics in Nuclear Medicine. 4th ed. Elsevier; 2012. Chapters 13-14.
• NEMA Standards Publication NU 1-2018: Performance Measurements of Gamma Cameras.
• ACR-AAPM Technical Standard for Nuclear Medicine Physics. Revised 2023.
• Hines H, et al. National Council on Radiation Protection (NCRP) Report No. 99: Quality Assurance for Diagnostic Imaging Equipment.

Read this lecture as Markdown