Gamma Imagers for Nuclear Security and Nuclear Forensics: Recommendations based on results from a side-by-side intercomparison

This paper presents the results of a side-by-side intercomparison between semiconductor and scintillator-based gamma imagers to provide guidance on their optimal utilization in a tiered nuclear security and forensic response strategy.

The Gist

Imagine you are a detective trying to find a hidden, glowing treasure (radioactive material) in a large, dark forest. You have two different types of "magic glasses" to help you see where the glow is coming from. This paper is a report on a test where the authors put these two very different glasses side-by-side to see which one works better for different parts of the job.

Here is the breakdown of their findings in simple terms:

The Two "Magic Glasses"

The researchers tested two specific devices designed to not only detect radiation but to take a picture of exactly where it is coming from, overlaying that picture onto a normal photo of the surroundings.

1. The "Lightweight Binoculars" (H3D H420):

   • What it is: A small, portable device using a special semiconductor crystal (Cadmium Zinc Telluride).
   • How it works: It's like a high-quality camera that sees radiation from all directions at once (360 degrees). It is very precise at identifying what the radiation is, but it's a bit "faint" when looking at the big picture.
   • The Analogy: Think of this as a detective with a very sharp magnifying glass. It can read the fine print on a single clue perfectly, but it takes a long time to scan a whole room, and if you try to scan while running, it gets too blurry to see anything.

2. The "Wide-Angle Flashlight" (SCoTSS 3×3):

   • What it is: A larger device made of hundreds of tiny crystals (Scintillators) arranged in a grid.
   • How it works: It acts like a powerful, fast-moving spotlight. It can see a lot of radiation very quickly and create a clear, smooth image of where the glow is.
   • The Analogy: Think of this as a detective with a super-bright flashlight. It can scan a whole field in seconds and tell you exactly where the glow is. However, if the glow is behind you or to the side, the flashlight beam gets a bit dimmer and fuzzier, making it harder to see details in those specific spots.

The Test Drive

The authors set up a controlled experiment in Ottawa. They placed radioactive sources (like glowing marbles) at specific distances and angles. They then took turns using the two devices to "photograph" the sources.

• The Single Source Test: When there was just one glowing marble, both devices found it.

  • The Wide-Angle Flashlight (SCoTSS) created a very smooth, clear picture almost instantly. It was so good that it could find the source in just 2 seconds.
  • The Lightweight Binoculars (H3D) took 2 minutes to build a picture that was a bit grainy (noisy), but it still found the source.

• The "Behind the Back" Test: They moved the source to the side and rear of the devices.

  • The Binoculars kept working just as well. It didn't matter if the source was in front, behind, or to the side; the picture quality stayed the same.
  • The Flashlight struggled a bit. The picture got fuzzier and the source looked weaker when it wasn't directly in front of the device.

• The "Two Sources" Test: They placed two glowing marbles close together.

  • The Flashlight was amazing at separating them. It showed two distinct dots clearly, even when they were close.
  • The Binoculars saw them as one big, blurry blob. It couldn't tell there were two separate sources, only that the glow was spread out.
  • However, when the two sources were very far apart (one in front, one behind), the Flashlight had a problem: it was so focused on the front source that it "hid" the back source in the final picture because of how the software filtered the image. The Binoculars, being fair to all directions, showed the spread of both.

The Verdict: Which Tool for Which Job?

The paper concludes that you shouldn't just pick one device for everything. Instead, you need a "tiered" approach, like using different tools for different stages of a search:

1. Stage 1: The Wide Search (Mobile Survey):

   • Best Tool: The Wide-Angle Flashlight (SCoTSS).
   • Why: When you are driving a car or flying a drone looking for a radioactive source over a huge area, you need speed. This device can find the "hot spots" in seconds and draw a map for you. It replaces the old, "direction-blind" detectors that just say "radiation is here" without saying "it's over there."

2. Stage 2: The Close-Up Investigation (In-Situ Characterization):

   • Best Tool: The Lightweight Binoculars (H3D).
   • Why: Once the wide search finds a suspicious spot, a team walks up to it and stands still for 15 minutes. Here, you don't need speed; you need precision. This device gives a very clear, fair view of the radiation no matter where it is coming from, helping experts figure out exactly what the material is without missing anything hidden behind an object.

The Bottom Line

The paper doesn't claim these devices are perfect yet, but it proves that different technologies excel at different stages of a nuclear security mission.

• If you are running to find a problem, use the fast, heavy crystal imager.
• If you are standing still to analyze a problem, use the precise, all-seeing semiconductor imager.

The future of nuclear security, according to the authors, involves using both types of imagers in a team, replacing old "blind" detectors with these new "visionary" ones to make the whole process safer and more accurate.

Technical Summary

Technical Summary: Intercomparison of Semiconductor and Scintillator Gamma Imagers for Nuclear Security

Problem Statement
Nuclear security and forensic operations rely on a tiered response strategy, beginning with wide-area searches using high-sensitivity mobile spectrometers and followed by high-precision, long-dwell characterization using solid-state detectors. While current methods effectively detect the presence and quantity of radioactivity, they possess limited ability to determine the precise location of a source. Recent advances in gamma imaging technology, derived from medical physics and astrophysics, offer the capability to map radioactive distributions and overlay them on optical photographs. However, the proliferation of different technological approaches (e.g., semiconductor vs. scintillator) and the lack of standardized performance reporting make it difficult for operators to select the optimal imager for specific mission spaces. Previous intercomparisons were limited by testing devices in different locations and background conditions.

Methodology
This study presents a preliminary side-by-side intercomparison of two gamma imagers based on fundamentally different technologies, deployed at the same location under identical experimental configurations in Ottawa, Canada, in October 2019.

• Instruments:
  • H3D H420: A commercial imager utilizing a >19 mm³ Cadmium Zinc Telluride (CZT) semiconductor detector. It features high energy resolution (≤1.1% FWHM at 662 keV) and a 4π field of view, but a smaller sensitive mass.
  • SCoTSS 3×3: A custom imager based on a scintillator design using 288 CsI(Tl) crystals arranged in a "scatter-layer" and "absorber-layer" configuration read out by silicon photomultipliers. It offers higher stopping power (comparable to a 4 L NaI(Tl) crystal) and an unlimited 4π field of view, though with lower energy resolution (7% FWHM at 662 keV) compared to the CZT device.
• Experimental Setup: Both detectors were alternately mounted on a rotating turntable at a fixed position. Two 160 MBq Cs-137 point sources were deployed at 10 meters distance. The study utilized both single-source runs (at -20° and -120° azimuths) and two-source runs (with separations of 10°, 30°, and 100°).
• Data Processing: To ensure a fair comparison, raw "list-mode" data from both instruments were processed through a unified analysis code. This involved strict event selection criteria: requiring exactly two energy deposits, summing them to the Cs-137 photopeak (662 keV), and applying a "lever arm" cut for the SCoTSS to minimize position smearing. Imaging was performed using Compton back-projection algorithms, with a 50% threshold applied to contour display to minimize false positives from statistical fluctuations.

Key Results
The intercomparison revealed distinct performance characteristics suited to different operational tiers:

1. Single Source Localization:

   • The SCoTSS 3×3 demonstrated superior image smoothness and angular resolution, particularly for sources in the forward direction. It successfully localized a source in a 2-second acquisition window, indicating suitability for mobile survey and imaging in motion.
   • The H3D H420 produced images with significant statistical scatter due to lower efficiency but maintained a uniform response across the 4π field of view. It required longer acquisition times (e.g., 2 minutes) to reconstruct a usable image and was deemed unsuitable for rapid mobile survey.

2. Multi-Source and Extended Source Resolution:

   • SCoTSS 3×3: While capable of resolving two sources separated by 30° (-20° and -50°) with high precision, its performance degraded significantly for sources at wide angles (e.g., -120°). The angular resolution and efficiency varied strongly with source location, causing the off-axis source to be suppressed below the 50% display threshold when a second, on-axis source was present.
   • H3D H420: Due to its uniform 4π response, the H3D H420 successfully indicated the extent of radioactivity for widely separated sources (e.g., -20° to -120°) and extended sources. However, its broad angular resolution prevented the discrimination of two distinct point sources separated by small angles (e.g., 10° or 30°), often appearing as a single elongated blob.

Significance and Conclusions
The paper concludes that no single gamma imager technology is optimal for all stages of a nuclear security response. Instead, the results support a tiered approach:

• Wide-Area Search: The scintillator-based SCoTSS 3×3 is well-suited for mobile survey operations where rapid data accumulation and high efficiency are required to map hot zones, effectively acting as a directional replacement for traditional NaI(Tl) survey spectrometers.
• Follow-on Characterization: The semiconductor-based H3D H420 is best suited for fixed, in-situ characterization. Its high energy resolution and uniform 4π response make it ideal for detailed isotopic analysis and localizing sources in complex or distributed scenarios, provided longer dwell times are available.

The authors emphasize that this preliminary work highlights the necessity of side-by-side intercomparisons to fairly evaluate technological strengths and weaknesses. They note that future directions include optimizing event sequencing, correcting for non-uniform response functions in scintillator imagers, and the eventual replacement of non-directional instruments with imaging counterparts as electronics become smaller and more integrated.