🔬 physics

Positron Emission Tomography with quantum-entangled Compton events: first imaging results at clinically relevant activities

This paper presents the first imaging results of a Positron Emission Tomography demonstrator that exploits the quantum entanglement of annihilation quanta via polarization-correlated Compton events, demonstrating spatial resolutions between 3.6 and 4.9 mm and a potential 10% sensitivity increase over conventional methods at clinically relevant activities.

Original authors: Mihael Makek, Ana Marija Kožuljević, Tomislav Bokulić, Darko Grošev, Zdenka Kuncic, Siddharth Parashari, Luka Pavelić

Published 2026-06-25

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Mihael Makek, Ana Marija Kožuljević, Tomislav Bokulić, Darko Grošev, Zdenka Kuncic, Siddharth Parashari, Luka Pavelić

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to take a perfect photograph of a tiny, glowing firefly inside a dark room. In the medical world, this "firefly" is a radioactive tracer inside a patient's body, and the "camera" is a Positron Emission Tomography (PET) scanner.

For decades, these cameras have worked by catching two "photons" (particles of light) that fly out in opposite directions when the tracer decays. The camera catches them and draws a line between the two points where they hit, guessing the firefly was somewhere on that line. This is like trying to find a lost coin by seeing two people point in opposite directions; you know the coin is somewhere on the line between them, but you don't know exactly where.

The New Idea: The Quantum "Handshake"
This paper introduces a clever new trick. When those two photons are born, they are "quantum entangled." Think of this like a pair of magical dice. If you roll them, they don't just land on random numbers; they are secretly connected. In this case, their "polarization" (the direction they are vibrating) is locked in a specific relationship: they are always perpendicular to each other, like the hands of a clock at 12 and 3.

Standard PET cameras ignore this secret handshake. They just catch the photons. This team built a special camera that looks for that handshake. They do this by letting the photons bounce off the camera's sensors (Compton scattering) before being caught. By measuring the angle of that bounce, they can tell if the two photons are holding hands (entangled) or if they are just random noise.

The Experiment: Building a Prototype
The researchers built a prototype camera using a single layer of special crystals (like a grid of tiny, sensitive eyes) connected to silicon sensors. They didn't build a full ring around a patient yet; instead, they used a rotating arm to simulate a full ring, testing it with two different-sized circles (one the size of a human head, one smaller).

They tested it with two types of "fireflies":

Tiny Rods: Two very thin, glowing lines (like glowing straws) to test how sharp the picture is.
A Phantom: A plastic model filled with glowing liquid that mimics a human body, used to test how well the camera sees details in a complex shape.

What They Found
Here is the breakdown of their results, using simple terms:

The Sharpness Trade-off:
- The Old Way (Single-Pixel): When the camera just catches the photon directly, the picture is very sharp. They got a resolution of about 2.5 mm. This is like taking a photo with a high-end lens.
- The New Way (Polarization-Correlated): When the camera tries to use the "handshake" (the entangled bounce), the picture gets a bit fuzzier, ranging from 3.6 mm to 4.9 mm. This is because it's harder to guess exactly where the photon hit first after it bounces. It's like trying to trace a path through a maze; you know the general direction, but the exact spot is slightly uncertain.
The Noise Filter:
- Standard cameras often get confused by "random" photons that happen to arrive at the same time by pure chance (like two strangers walking into a room at the same time).
- The new method is much better at filtering these out. The "entangled" photons they selected had 20% less background noise than the standard method. It's like wearing noise-canceling headphones that only let through the specific song you want to hear.
The Gain:
- By adding these "entangled" events to the standard ones, they didn't just get a cleaner picture; they got more data. They found they could increase the camera's sensitivity (its ability to see the signal) by up to 10%.
- Crucially, even with the slightly fuzzier "entangled" events, when they mixed them with the sharp "standard" events, the final picture remained high quality.

The Bottom Line
The researchers successfully proved that you can use the quantum "handshake" of particles to improve PET imaging. While the "entangled" events alone aren't as sharp as the standard method, they are much cleaner (less noisy).

By combining the two, they estimate they can get a 10% boost in sensitivity without ruining the image quality. This means doctors could potentially use less radioactive dye or scan patients faster in the future. However, the paper notes that to get the full benefit, they need to figure out a better way to pinpoint exactly where the "bounced" photons hit, so the picture doesn't get too blurry.

In short: They built a camera that listens for a secret quantum code between particles. It's a bit fuzzier than the old camera, but it hears the signal much more clearly, allowing for a potentially better and more efficient scan.

Technical Summary: Positron Emission Tomography with Quantum-Entangled Compton Events

Problem Statement
Positron Emission Tomography (PET) relies on the coincident detection of two 511 keV gamma photons from positron annihilation. While conventional PET systems have improved sensitivity and image quality through advances in detector components and reconstruction algorithms, a potential enhancement remains unutilized: the quantum entanglement of the annihilation quanta. The polarization of these photons is correlated (orthogonal) due to the conservation of angular momentum and parity in the $e^+e^-$ system. This correlation offers a method to discriminate true coincidences from random background events independent of energy criteria. However, implementing this in clinical settings is challenging because traditional Compton polarimeters require multi-layer detector systems to track recoil electrons and scattered photons, which would significantly increase the complexity and cost of a clinical PET camera.

Methodology
To address the cost and complexity barrier, the authors developed a PET demonstrator based on a single-layer scintillator polarimeter concept. In this design, a segmented scintillator matrix acts as both the Compton scatterer and the absorber, eliminating the need for multiple detector layers.

Experimental Setup: The system comprised two detector modules, each containing 256 GAGG:Ce crystals (3.0 mm $\times$ 3.0 mm cross-section, 20 mm length) coupled to SiPM arrays. The modules were mounted on a rotating gantry allowing for imaging at diameters of 430 mm and 620 mm.
Data Acquisition: Using a TOFPET2 system, the setup triggered on coincidence events. The study focused on Polarization-Correlated Events (PCE), defined as events where both gamma photons undergo Compton scattering within their respective modules.
Event Selection & Reconstruction:
- Events were selected based on kinematic criteria: exactly two pixels firing per module with energy sums near 511 keV, and scattering angles consistent with Compton kinematics.
- Since the order of interaction (recoil electron vs. scattered photon) cannot be distinguished with certainty in a single layer, the pixel with lower energy deposition was assumed to be the recoil electron.
- Azimuthal scattering angles ( $\phi$ ) were reconstructed, and the distribution of the difference ( $\phi_1 - \phi_2$ ) was acceptance-corrected to extract the modulation factor ( $\mu$ ).
- Images were reconstructed using the Ordered Subset Expectation Maximization (OSEM) algorithm. The study compared three datasets: Single-Pixel Events (SPE, the conventional benchmark), PCE-only, and a merged SPE+PCE list-mode dataset.
Phantoms: Experiments utilized two $^{68}$ Ge rod sources (45 MBq each) and a NEMA NU-4 phantom filled with $^{68}$ Ga solution (378 MBq initial activity).

Key Results
The study presents the first imaging results of clinically relevant activities using polarization-correlated Compton events.

Spatial Resolution:
- SPE (Benchmark): Achieved a spatial resolution of 2.5 ± 0.1 mm (at 430 mm diameter) and 2.7 ± 0.1 mm (at 620 mm diameter).
- PCE-only: Resolutions ranged from 3.6 ± 0.3 mm to 4.9 ± 0.3 mm, depending on the selection criteria (specifically the inter-pixel distance $d_{1,2}$ and scattering angles). The degradation in resolution is attributed to the uncertainty in assigning the Line of Response (LOR) to the correct pixel position due to the unknown interaction order.
- Merged (SPE+PCE): When PCE were merged with SPE, the spatial resolution remained close to the SPE benchmark (e.g., 2.52 mm to 2.61 mm), demonstrating that adding PCE does not significantly degrade the image resolution if selection criteria are optimized.
Signal-to-Background Ratio (SBR): PCE exhibited up to 20% higher average SBR compared to SPE. This indicates that polarization-correlated events contain a lower fraction of random coincidences, effectively filtering out noise.
Image Quality Metrics:
- Signal-to-Noise Ratio (SNR): Improved with the inclusion of PCE.
- Contrast and CNR: Showed dependence on the spatial resolution of the PCE contribution. Improvements were observed when the PCE-only resolution was below 4 mm.
Sensitivity Gain: By merging PCE with SPE, the system gained up to 17% additional events for reconstruction. The authors estimate a net sensitivity increase of up to 10% while maintaining high image quality.

Significance and Claims
The paper claims to be the first to demonstrate PET imaging using quantum-entangled Compton events with clinically relevant activities (up to 378 MBq). The primary significance lies in proving the feasibility of a single-layer detector architecture for polarization measurement, which is compatible with current clinical scanner designs and avoids the prohibitive costs of multi-layer systems.

The authors conclude that while the current LOR assignment method limits the spatial resolution of PCE-only images, the trade-off is favorable. By exploiting polarization correlations, it is possible to:

Reduce random background contamination.
Increase sensitivity by approximately 10%.
Maintain or slightly improve overall image quality (SNR, CNR) when merging PCE with conventional SPE data.

The study suggests that future improvements in LOR assignment (e.g., via machine learning or depth-of-interaction detectors) could further unlock the benefits of quantum-entanglement in PET, but even with current methods, a measurable sensitivity gain is achievable.

Technical Summary: Positron Emission Tomography with Quantum-Entangled Compton Events

More like this