Positron emission tomography (PET) is a painless, non-invasive medical imaging technique, widely used for detection cancer, heart disease and brain disorders such as dementia. This process involves detection of metabolism within bodily tissue, enabling construction of 3-dimensional images which show how organs and tissue in the body function, rather than producing an image that solely shows appearance.
How does it work?
A radioisotope is an atom with an unstable nucleus. Nuclei lie at the centre of atoms, composed of protons and neutrons. Unstable nuclei have excess energy, which causes them to change their fundamental composition and transform into a less energetic, more stable atom – this process is known as radioactive decay.
A radiotracer is a molecule that has at least one of its atoms replaced with a radioisotope. Radiotracers are used in a number of medical imaging techniques because, once injected into the body, they emit radiation which can be detected externally and then used to construct detailed pictures of the body. The tracer used in a PET scan depends on its purpose; F-18-fluoro-deoxyglucose (FDG) is commonly used for cancer and brain imaging; the compound acts much like glucose and thus the brain treats it in a similar way to how it uses glucose for its metabolism.
Upon injection into the body, FDG molecules decay relatively quickly into a stable isotope of oxygen (Oxygen-18), emitting positrons in the process; this is known as beta decay. The molecules have a short half life, minimising the radiation dose administered to the patient – an instrumental property for the mitigation of long-term risks posed by excessive exposure to radiation.
Positrons are a type of antimatter, which describes matter with opposite properties to normal matter. More specifically, they are the corresponding antiparticle of electrons, similar in every way except with an opposite (positive) charge.
Soon after production, each positron will inevitably collide with a nearby electron in the body, which leads to annihilation. Annihilation is where the mass of each particle is transformed into an amount of energy, which can be calculated using Einstein’s famous energy-mass relation (E=mc2). This energy is released in the form of 2 gamma rays that travel in opposite directions to conserve momentum, as shown in the diagram below.
A circular detector, composed of inorganic scintillator crystals, surrounds the patient, registering the simultaneous arrival of gamma rays and transforming them into an electrical signal for analysis by a computer. Electronic collimation involves the extrapolation of these signals to pinpoint the origin of the signal to develop a 3-dimensional picture of how the body interacts with the radiotracer.
Interpreting the Image
Cancerous tissue uses more glucose than normal tissue due to accelerated growth; this causes excess accumulation of FDG in comparison to healthy tissue, so cancerous areas show up darker on the PET scan. It is important to note that there are a number of factors that can cause uncertainty in these images; we will briefly look at 2 of the most prominent challenges for optimising image clarity, namely attenuation correction and temporal mismatches.
Gamma rays travelling in opposite directions travel through different parts of the body, consequently interacting with different matter and experiencing differing amounts of attenuation. As a result, there will generally be a very slight disparity in the time taken for each of them to reach the detector. In compound to this effect, detectors may have slightly different signal processing times.
To account for these effects, if 2 gamma rays arrive within 6-10ns of each other, the detector still recognises and attributes the gamma rays to the same annihilation event. This short time frame is known as the coincident time window. When 2 photons are recorded within this time frame, without intermediate interactions with bodily matter, we call this a true coincidence. However, due to the leniency of the coincident time window, the detector will inevitably also attribute unrelated photons as coincidence events; these are known as accidental or random events.
Scatter coincidences occur when, due to scattering of gamma rays as they interact with bodily matter, 2 photons arrive at the PET detector within the coincidence time-window, appearing to originate from a different location from reality. This latter type of coincidence contributes around 40% of total recorded coincidences, but various mechanisms that revolve around energy analysis of scattered photons are employed to effectively eliminate as many scatter coincidences as possible.
As gamma rays travel through the body, their intensity drops off exponentially. To put this in perspective, by the time that the radiation has travelled from one side of the abdomen to the other, the intensity has decreased to just 1/50th of the initial amount. This causes significant degradation in image contrast and quality, especially in larger patients.
The main cause of this attenuation is a process known as Compton scattering. When gamma rays interact with electrons, they transfer energy to the electrons. This energy loss causes them to be scattered away from the detector, rendering them unavailable for image formation, decreasing the resolution of the final image that is formed.
SPECT vs PET
Single photon emission computerised tomography (SPECT) is an alternative, cheaper tomographic technique that is commonly employed. As the name suggests, SPECT involves the release of just 1 gamma photon rather than 2, meaning that the detector must locate the source of the photon via a less accurate mechanism compared to that which is used for PET scans. The more advanced diagnostic ability of PET scans are owed to its superior image contrast and spatial resolution, albeit at a higher financial cost.