6.2 PET/CT Instrumentation, Coincidence Detection & Crystals

PET/CT Instrumentation and Physics

Positron Emission Tomography (PET) represents a fundamentally different imaging paradigm compared to SPECT. While SPECT relies on single-photon emitters (like Tc-99m) and physical collimation, PET utilizes positron-emitting radiopharmaceuticals (like F-18 FDG) and electronic collimation through coincidence detection. Modern systems combine PET with Computed Tomography (CT) into a single gantry, providing perfectly co-registered metabolic (PET) and anatomic (CT) information.

Positron Emission and Annihilation

The physics of PET begins with the radioactive decay of a positron-emitting isotope.

1. Emission: A proton-rich nucleus decays by converting a proton into a neutron, emitting a positron ($e^+$) and a neutrino. A positron is the antimatter counterpart of an electron—it has the same mass but a positive charge.
2. Travel: The emitted positron travels a short distance through the surrounding tissue, losing kinetic energy. This distance is called the positron range and is typically 1-3 mm for F-18. Positron range represents a fundamental physical limit to the spatial resolution of a PET scanner.
3. Annihilation: Once the positron loses nearly all its kinetic energy, it encounters an electron ($e^-$) in the tissue. The matter and antimatter collide and annihilate each other.
4. Twin Photons: According to Einstein's mass-energy equivalence equation ($E=mc^2$), the combined mass of the positron and electron is converted entirely into energy. This results in the creation of two gamma photons, each with exactly 511 keV of energy. To conserve momentum, these twin photons travel in approximately opposite directions, approximately 180 degrees apart.

Coincidence Detection and the LOR

Unlike a gamma camera that uses a lead collimator to determine the direction of incoming photons, a PET scanner relies on a ring (or multiple rings) of detectors surrounding the patient and a concept called electronic collimation.

The Line of Response (LOR)

When the two 511 keV annihilation photons hit opposing detectors almost simultaneously, the system registers a valid event. The scanner assumes that the annihilation event occurred somewhere along the straight line connecting those two specific detectors. This imaginary line is called the Line of Response (LOR).

Types of Coincidence Events

For an event to be recorded, the two photons must arrive within a very narrow timeframe, called the coincidence timing window (typically 4-12 nanoseconds, depending on the crystal type). However, not all detected pairs represent useful information.

1. True Coincidence: Two photons from the same annihilation event strike opposing detectors within the timing window without undergoing significant scatter. This creates a valid LOR that accurately passes through the site of radiotracer uptake. This is the desired signal.
2. Scatter Coincidence: Two photons from the same annihilation event are detected, but one or both have undergone Compton scattering within the patient's body prior to hitting the detectors. Scattering changes the photon's path. Therefore, the resulting LOR is incorrect and does not pass through the actual annihilation site, adding background noise and reducing image contrast.
3. Random Coincidence: Two photons from different, unrelated annihilation events happen to strike opposing detectors within the coincidence timing window. The system incorrectly pairs them, drawing an invalid LOR between them. Randoms increase significantly at higher doses of radioactivity and degrade image contrast.

Scintillator Crystals: LSO/LYSO vs. BGO

The choice of scintillator crystal is the most critical hardware component in a PET scanner. The crystal must efficiently stop 511 keV photons and quickly convert them into light.

Key Crystal Properties

• Stopping Power (Density & Effective Z): Higher density and higher effective atomic number ($Z_{eff}$) mean the crystal is better at stopping the high-energy 511 keV photons, increasing scanner sensitivity.
• Light Yield: The amount of light photons produced per keV of deposited energy. Higher light yield improves energy resolution, allowing better rejection of scattered photons.
• Decay Time: How quickly the crystal flashes and resets. A shorter (faster) decay time allows the scanner to handle higher count rates (reducing dead time) and enables smaller coincidence timing windows (reducing random events).

Evolution of Crystals

• Bismuth Germanate (BGO): The standard crystal for many years. It has excellent stopping power (high density and high $Z_{eff}$) and is relatively inexpensive. However, its major drawback is a very slow decay time (~300 ns) and low light yield. This limits its ability to handle high count rates and precludes the use of Time-of-Flight technology.
• Lutetium Oxyorthosilicate (LSO) and Lutetium Yttrium Oxyorthosilicate (LYSO): These are the modern standards. They possess stopping power nearly equal to BGO but offer drastically superior performance in other areas. They have a very high light yield (improving energy resolution) and an incredibly fast decay time (~40 ns). This rapid speed allows for extremely narrow coincidence timing windows and is the enabling technology for Time-of-Flight PET.

Time-of-Flight (TOF) PET

In standard PET, a true coincidence event establishes an LOR, but the system does not know where along that line the annihilation occurred. During reconstruction, the probability is distributed evenly along the entire length of the LOR.

Time-of-Flight (TOF) takes advantage of the ultra-fast decay times of LSO/LYSO crystals. Light travels at approximately 30 cm per nanosecond. If an annihilation event occurs closer to Detector A than Detector B, the photon will strike Detector A a fraction of a nanosecond before the twin photon strikes Detector B.

A TOF-capable scanner measures this minute time difference (timing resolution, often ~300-500 picoseconds). Using the speed of light and the time difference ($\\Delta t$), the system can localize the annihilation event to a specific, narrow segment along the LOR, rather than smearing it across the whole line.

Benefits of TOF:

• Massive improvement in Signal-to-Noise Ratio (SNR).
• Superior image quality, especially in larger patients where attenuation and scatter are severe.
• Better lesion detectability and contrast recovery.
• Potential for reduced scan times or lower injected doses.

The CT Component

In a PET/CT scanner, the CT gantry is mounted coaxially with the PET gantry. The CT utilizes a standard x-ray tube and solid-state detectors to generate high-resolution anatomical images. In the context of PET/CT, the CT scan serves two vital purposes:

1. Anatomical Localization: Providing a high-resolution map to precisely locate the areas of increased radiotracer uptake seen on the PET scan.
2. Attenuation Correction (CTAC): The CT data provides a highly accurate density map of the patient, which is mathematically essential for correcting the severe photon attenuation that occurs in PET imaging (detailed further in Section 6.3).