Gamma Imaging Devices: Principles & Applications

• Gamma imaging devices are instruments that spatially localize gamma-ray sources using position-sensitive detectors and collimation methods like pinhole, coded-aperture, and Compton scattering.
• They employ multiple physical principles to achieve high spatial and energy resolution, enhancing applications in nuclear safeguards, medical imaging, and astrophysics.
• Recent advances focus on miniaturization, improved detection efficiency, and multi-modal imaging, driving innovations in nuclear security and environmental monitoring.

Gamma imaging devices are instruments capable of detecting and spatially localizing gamma-ray sources via the measurement of gamma-ray interactions within a position-sensitive detection system. By leveraging various physical principles—most notably Compton scattering, pinhole or coded-aperture collimation, and pair production—gamma imagers have been deployed across a broad range of fields, including nuclear security, nuclear safeguards, medical diagnostics, non-destructive assay, astrophysics, and environmental monitoring. Recent advances emphasize miniaturization, enhanced position and energy resolutions, increased detection efficiency, and multi-modality (e.g., gamma-neutron) operation.

1. Foundational Physical Principles of Gamma Imaging

Gamma imaging fundamentally relies on the spatial and energetic detection of gamma-ray interactions within a segmented or pixelated detector medium, coupled with a geometric or electronic collimation scheme to deduce source direction. The principal methodologies include:

• Pinhole and coded-aperture imaging: A mechanical collimator (single hole or pattern) restricts gamma trajectories to reconstruct a source image via the geometric correlation between photon arrival position and angle. The spatial resolution is constrained by the collimator's geometry (hole size $d$ and mask-to-detector separation $D$), with $\Delta x \approx d$ for a pinhole of diameter $d$ located at distance $D$ (Caballero et al., 2018).
• Compton imaging: Electronic collimation via Compton scattering exploits the kinematics of the Compton effect. A photon deposits energy $E_1$ at position $\mathbf{x}_1$ (scatter detector), then is fully absorbed with energy $E_2$ at $\mathbf{x}_2$ (absorber). The incident angle $\theta_C$ is reconstructed from energy and positional measurements using the Compton formula:

$$\cos\theta_C = 1 + \frac{m_ec^2}{E_2} - \frac{m_ec^2}{E_1 + E_2},$$

constraining the possible photon origin to the surface of a cone about axis $\mathbf{x}_2 - \mathbf{x}_1$ (Sinclair et al., 2016).

• Pair-production/polarimetry (GeV regime): High-energy gamma photons produce $e^+e^-$ pairs in a converter stack (e.g., nuclear emulsion films). The reconstructed pair plane and energies enable incident photon reconstruction and, under favorable conditions, polarization analysis (Rokujo et al., 2017).

Further advancements incorporate simultaneous neutron and gamma imaging by exploiting isotope-specific capture or scattering, and hybrid architectures (e.g., "active masking" with scene data fusion) for 3D radiological mapping (Vavrek et al., 12 Mar 2025).

2. Detector Architectures and Materials

Gamma imaging devices encompass diverse detector-module configurations and readout technologies:

Device Type	Detector Media	Readout Scheme
Compton camera	Scintillator (CsI, NaI, CeBr₃, LaCl₃), semiconductor (CZT, CdTe, Si/Ga), emulsion	PMT, SiPM, semiconductor pixellation, SOI CMOS, MAPMT, Gaseous Photomultiplier
Pinhole/coded	CeBr₃, CdTe, CsI, HPGe	PSPMT, ASICs, SDD, FPGA, ARM CPU
Pair telescope	Emulsion film (converter/calorimeter)	Automated optical scan
Hybrid/advanced	CLLBC (gamma/neutron), LXe	SiPM, THGEM, LiDAR/SLAM fusion

• Pixelated semiconductors (e.g., CZT arrays): 2D or 3D pixellated anode/cathode structures at $\sim1-2$ mm pitch directly provide depth-of-interaction resolution and enable Compton reconstruction at high spatial and energy resolution (Brücken et al., 29 Oct 2025, Saariokari et al., 2024).
• Scintillator arrays with SiPM readout: Cubic or bar-shaped crystals (e.g., CsI(Tl), NaI(Tl), LaCl₃, GAGG) read out by SiPM arrays or PMTs provide the spatial segmentation for event localization in both planar and volumetric architectures (Sinclair et al., 2016, MacLeod et al., 2015, Marco et al., 2022).
• Liquid noble-gas scintillation: Capillary-filled LXe detectors with gaseous photomultiplier readout achieve high spatial resolution via photon transport control and readout segmentation; fast-neutron and MeV gamma discrimination is possible (Israelashvili et al., 2017).
• Silicon-on-insulator (SOI) pixel scatter detectors: 30 μm pitch monolithic SOI pixel arrays enable precise electron track imaging for advanced Compton tracking concepts (Shimazoe et al., 2015).
• Coded-aperture systems: CdTe-based hybrids (e.g., Caliste-O, 16×16 pixels, 0.89 mm pitch, 2-mm thick) coupled with FPGA/ARM SoC for real-time imaging and advanced search/deconvolution techniques are now feasible in sub-kilogram portable systems (Maier et al., 2020).
• Neutron/gamma dual-mode: CLLBC-based active-masked imagers integrate fast neutron and gamma imaging by combining pixelated crystal arrays, LiDAR-driven scene data fusion, and advanced likelihood-based algorithms (Vavrek et al., 12 Mar 2025), while GN-Vision types exploit layered scintillators and pinhole moderation (Marco et al., 2022).

3. Imaging Performance Metrics and Limiting Factors

Performance is appraised in terms of spatial/angular resolution, energy resolution, detection efficiency, field of view, and sensitivity. Representative metrics include:

• Compton Angular Resolution Measure (ARM):

$$\mathrm{ARM} = \theta_{\text{recon}} - \theta_{\text{geom}}$$

where $\theta_{\text{recon}}$ is from the Compton formula and $\theta_{\text{geom}}$ is the true source-scatter angle. ARM widths (standard deviation or FWHM) are empirically $2.8^{\circ}$–$4.7^{\circ}$ (662 keV–344 keV) for SCoTSS (Sinclair et al., 2016), $<4^{\circ}$–$20^{\circ}$ (662 keV) for SCoTSS 3×3 vs. $20^{\circ}$–$30^{\circ}$ for CZT-based H420 (Sinclair et al., 31 Jan 2026).

• Energy Resolution (FWHM): 7.5%–7.9% (CsI+SiPM, NaI+PMT at 662 keV, SCoTSS) (Sinclair et al., 2016); $\leq$1.1% for H3D H420 (CZT) and 0.3% for GeGI (HPGe) (Brücken et al., 29 Oct 2025).
• Spatial Resolution: Pixel-limited ($\sim$pitch/$\sqrt{12}$): e.g., 2 mm pitch yields $0.58$ mm RMS (Saariokari et al., 2024). For pinhole devices, spatial resolution $\Delta x \approx d$, e.g., 3 mm for GUALI (Caballero et al., 2018). LXe/GPM achieves 12 mm FWHM (prototype, 1.33 MeV) and $<$4 mm projected (optimized, simulated) (Israelashvili et al., 2017).
• Detection Efficiency: Compton photopeak efficiency for 662 keV: SCoTSS, $0.60\%$; GUALI, $0.0046\%$ (unfiltered, 662 keV, 19 cm) (Sinclair et al., 2016, Caballero et al., 2018); 11–13× increases in full absorption efficiency with large CZT crystals over legacy designs (Saariokari et al., 2024, Brücken et al., 29 Oct 2025).
• Field of View (FOV): Compton systems routinely achieve $>$80° (±40°), with many architectures supporting full 4$\pi$ coverage (CZT, CLLBC array) (Sinclair et al., 31 Jan 2026, Vavrek et al., 12 Mar 2025).

Limiting factors include electronic noise (imposing thresholds), physical pixel size (position resolution), Doppler broadening, energy resolution limitations, dead material (spurious scattering), and detection geometry constraints.

4. Reconstruction Methods and Signal Processing

Imaging reconstruction is based on signal processing and event selection tailored to the specific device physics:

• Back-projection: For Compton cameras, cones of possible emission direction are superimposed to form an image; pixel intensity correlates to the intersection density of cones (Sinclair et al., 2016, MacLeod et al., 2015).
• Likelihood-based algorithms: List-mode MLEM (Maximum Likelihood Expectation Maximization) iteratively maximizes the Poisson likelihood of observing recorded events given a model of the spatial distribution. The EM step recursively updates estimate $\lambda_j^{(k)}$ according to

$$\lambda^{(k+1)}_j = \frac{\lambda^{(k)}_j}{s_j} \sum_{i} \frac{t_{ij}}{\sum_{k'} t_{ik'} \lambda^{(k)}_{k'}},$$

where $t_{ij}$ is the event–voxel probability kernel (Yabu et al., 2021).

• Centroiding and moment analysis: For monolithic scintillators (GUALI), Anger logic is used for fast position estimation (Caballero et al., 2018).
• Coded mask deconvolution: Detector shadowgrams are cross-correlated with known mask patterns to obtain source maps, achieving arcminute to degree-scale localization (Campana et al., 2018, Maier et al., 2020).
• Hybrid multi-modal fusion: Scene Data Fusion with SLAM-generated 3D maps fuses radiometric data with position and environmental context for quantitative, spatially resolved mapping in arbitrary environments, employing Poisson-based global ML-EM or GPSL/ASPL localization (Vavrek et al., 12 Mar 2025).

Calibration methods involve energy and spatial mapping with known sources, polynomial distortion corrections, LUT-based uniformity compensation, and time-of-flight/event-coincidence windows for background rejection.

5. Representative Applications and Operational Scenarios

Gamma imaging is deployed in diverse settings:

• Nuclear safeguards and waste characterization: Passive Gamma Emission Tomography (PGET) verifies spent nuclear fuel; pixelated CZT or HPGe imagers localize, count, and distinguish isotopes in storage drums or fuel assemblies (Saariokari et al., 2024, Brücken et al., 29 Oct 2025, Caballero et al., 2018).
• Emergency response and security: Portable scintillator-based Compton imagers (SCoTSS) deliver 1$^\circ$ localization of 10 mCi point sources at 40 m within a minute over $\pm 45^\circ$ fields (Sinclair et al., 2016).
• Medical imaging: 3D-PSCC and Si/CdTe Compton cameras enable tomographic and real-time monitoring of dose delivery in proton therapy, achieving mm-scale spatial and 10% energy resolutions for prompt gamma emission (Koide et al., 2018, Yabu et al., 2021). Advanced SPECT architectures exploit multi-pinhole/curved detector arrays for enhanced sensitivity and resolution (Bhusal, 2020).
• Astrophysics: Balloon-borne emulsion telescopes (GRAINE) with sub-micron resolution in 0.01–100 GeV range enable high-statistics observations of cosmic gamma sources with intrinsic polarization sensitivity (Rokujo et al., 2017).
• Neutron/gamma dual imaging: Devices such as GN-Vision or CLLBC-based arrays offer simultaneous, real-time 3D mapping of neutron and gamma sources for nuclear safety, inspection, and decommissioning (Marco et al., 2022, Vavrek et al., 12 Mar 2025).
• Survey and mapping: Free-moving systems integrate LiDAR and omnidirectional gamma detection to enable rapid, quantitative mapping over complex terrains (urban, industrial, environmental) (Vavrek et al., 12 Mar 2025).

6. Technological Comparison and Selection Considerations

Selection of gamma imaging devices for field or laboratory work is dictated by a balance of sensitivity, spatial/angular and energy resolution, operation speed, spectral range, and system complexity.

Use Case	Preferred Solution	Rationale
Wide-area mobile survey	Scintillator-based Compton imager (SCoTSS)	High efficiency, rapid source localization, robustness
High-precision isotope ID	CZT/HPGe semiconductor imager (H420)	Superior energy resolution, 4$\pi$ FOV, in situ spec.
3D radiological mapping	Active-masked array + SDF (CLLBC, LiDAR)	Quantitative mapping, omnidirectional, scene-aware
Medical SPECT/SPECT-CT	Curved multi-pinhole SPECT (hemi-ellipsoid)	High sensitivity, sub-5 mm FWHM, dose/time tradeoff
Proton therapy verification	3D position-sensitive Compton camera	mm-scale Bragg peak imaging, high energy operation
Nuclear astrophysics	Emulsion pair telescope (GRAINE), XGIS	Sub-degree to degree angular, polarization, wide band

No single detector concept is optimal across all applications: scintillator-based Compton systems offer higher efficiency for forward sources; semiconductor imagers deliver best energy resolution and uniform 4$\pi$ response; hybrid designs (e.g., CLLBC, coded-aperture) enable flexibility and new capabilities (Sinclair et al., 31 Jan 2026). Ongoing research focuses on integrating detector miniaturization, algorithmic advances (MLEM, advanced deblurring, Bayesian inference), depth-of-interaction correction, and autonomous operation with site context (LiDAR/SLAM fusion).

7. Outlook and Future Directions

Future gamma imager development targets include:

• Materials and architectures: Implementation of advanced room-temperature semiconductors (TlBr, perovskites), denser/faster scintillators (LaBr₃, CeBr₃), and hybrid multi-layer or curved geometries for 4$\pi$ uniformity and depth-of-interaction sensitivity (Sinclair et al., 31 Jan 2026).
• Electronics and readout: Upgrades to digital pulse processing enabling sub-50 keV thresholds, real-time event-level Compton analysis, and high-throughput acquisition are under consideration (Brücken et al., 29 Oct 2025).
• Algorithmic innovation: Full-system Monte Carlo MLEM, system-matrix response inclusion, and GPU-accelerated real-time 3D imaging algorithms are key for next-generation deployment (Vavrek et al., 12 Mar 2025).
• Integration and miniaturization: Low-power FPGA-based electronics, embedded GNSS/IMU for continuous mobile tracking, and fusion with vision/sensor packages for augmented-reality overlays in field surveys (Vavrek et al., 12 Mar 2025).
• Multi-modality: Expansion to neutron/gamma co-imaging, spectral-polarimetric measurement, and environmental robustness (temperature, vibration, immersion) (Marco et al., 2022, Vavrek et al., 12 Mar 2025).
• Application-driven R&D: Adaptive instruments for specific emerging use cases—range verification in radiotherapy, miniaturized probes for intraoperative assessment, lightweight platforms for airborne surveillance, and large-area platforms for astrophysics (Rokujo et al., 2017, Koide et al., 2018).

Significant improvements remain possible especially in the trade-off between system complexity, cost, and field performance, as well as in the integration of multi-modal imaging with contextual environmental modeling. The field is moving toward highly versatile, portable, and quantitative gamma imaging enabling prompt, actionable decision making across safety, security, scientific, and medical domains.