Positron emission tomography

Safety

PET scanning is non-invasive, but it does involve exposure to ionizing radiation. For a typical dose (245 MBq) of FDG, one of the most common radiotracers used for PET neuroimaging and cancer patient management, the effective radiation dose is 4.7 mSv.

For combined PET–CT scanning, the radiation exposure contributed by the CT scan may be substantial - ranging from around 3–26 mSv (for a 70 kg person, and depending on the coverage and intended use of the scan).

The amount of radiation in a typical FDG PET-CT scan is similar to the effective dose of spending one year in the American city of Denver, Colorado (12.4 mSv/year). For comparison, radiation dosage for other medical procedures range from 0.02 mSv for a chest X-ray and 6.5–8 mSv for a CT scan of the chest. Average civil aircrews are exposed to 3 mSv/year, and the IAEA recommend that whole body effective dose for nuclear energy workers is 20 mSv/year (when averaged over 5 years) and does not exceed 50 mSv in a given year.

Operation

Radionuclides and radiotracers

The radioactive substance used clinical practices in PET scanning is a simple sugar called fluorodeoxyglucose (FDG). Many different tracers may be used for research or clinical investigations various imaging purposes, depending on the target process within the body, such as:

• Fluorodeoxyglucose ([¹⁸F]FDG or FDG) is commonly used to detect cancer, as cancer cells need sugar to grow.
• [¹⁸F]Sodium fluoride (Na¹⁸F) is widely used for detecting bone formation.
• Oxygen-15 (¹⁵O)-water is used to quantify myocardial blood flow.
• Carbon-11 (¹¹C)-methionine is used to image brain tumors.

Radionuclides are incorporated either into compounds normally used by the body such as glucose (or glucose analogues), water, or ammonia, or into molecules that bind to receptors or other sites of drug action. Such labelled compounds are known as radiotracers. PET technology can be used to trace the biologic pathway of any compound in living humans (and many other species as well), provided it can be radiolabeled with a PET isotope. Radiotracers for new target molecules and processes are continuing to be synthesized. As of this writing there are already dozens in clinical use and hundreds applied in research. In 2020 by far the most commonly used radiotracer in clinical PET scanning is the carbohydrate derivative FDG. This radiotracer is used in essentially all scans for oncology and most scans in neurology, thus makes up the large majority of radiotracer (>95%) used in PET and PET–CT scanning.

Due to the short half-lives of most positron-emitting radioisotopes, the radiotracers have traditionally been produced using a cyclotron in close proximity to the PET imaging facility. The half-life of fluorine-18 is long enough that radiotracers labeled with fluorine-18 can be manufactured commercially at offsite locations and shipped to imaging centers. Recently rubidium-82 generators have become commercially available. These contain strontium-82, which decays by electron capture to produce positron-emitting rubidium-82.

The use of positron-emitting isotopes of metals in PET scans has been reviewed, including elements not listed above, such as lanthanides.

Immuno-PET

The isotope ⁸⁹Zr has been applied to the tracking and quantification of molecular antibodies with PET cameras (a method called "immuno-PET").

The biological half-life of antibodies is typically on the order of days, see daclizumab and erenumab by way of example. To visualize and quantify the distribution of such antibodies in the body, the PET isotope ⁸⁹Zr is well suited because its physical half-life matches the typical biological half-life of antibodies, see table above.

Immuno-PET can use both intact monoclonal antibodies and smaller antibody-derived constructs, including antibody fragments and nanobodies. The radionuclide is generally selected according to the pharmacokinetics of the antibody or antibody-derived construct. Intact antibodies circulate slowly and may therefore be labelled with radionuclides that permit imaging at later time points, including iodine-124, copper-64 and zirconium-89. By contrast, the faster clearance of smaller antibody-derived constructs allows the use of shorter-lived radionuclides, including gallium-68 and fluorine-18.

Emission

To conduct the scan, a short-lived radioactive tracer isotope is injected into the living subject (usually into blood circulation). Each tracer atom has been chemically incorporated into a biologically active molecule. There is a waiting period while the active molecule becomes concentrated in tissues of interest. Then the subject is placed in the imaging scanner. The molecule most commonly used for this purpose is FDG, a sugar, for which the waiting period is typically an hour. During the scan, a record of tissue concentration is made as the tracer decays.

As the radioisotope undergoes positron emission decay (also known as positive beta decay), it emits a positron, an antiparticle of the electron with opposite charge. The emitted positron travels in tissue for a short distance (typically less than 1 mm, but dependent on the isotope), during which time it loses kinetic energy, until it decelerates to a point where it can interact with an electron. The encounter annihilates both electron and positron, producing a pair of annihilation (gamma) photons moving in approximately opposite directions. These are detected when they reach a scintillator in the scanning device, creating a burst of light which is detected by photomultiplier tubes or silicon avalanche photodiodes (Si APD). The technique depends on simultaneous or coincident detection of the pair of photons moving in approximately opposite directions (they would be exactly opposite in their center of mass frame, but the scanner has no way to know this, and so has a built-in slight direction-error tolerance). Photons that do not arrive in temporal "pairs" (i.e. within a timing-window of a few nanoseconds) are ignored.

Localization of the positron annihilation event

The most significant fraction of electron–positron annihilations results in two 511 keV gamma photons being emitted at almost exactly 180 degrees to each other. Hence, it is possible to localize their source along a straight line of coincidence (also called the line of response, or LOR). In practice, the LOR has a non-zero width as the emitted photons are not exactly 180 degrees apart. If the resolving time of the detectors is less than 500 picoseconds rather than about 10 nanoseconds, it is possible to localize the event to a segment of a chord, whose length is determined by the detector timing resolution. As the timing resolution improves, the signal-to-noise ratio (SNR) of the image will improve, requiring fewer events to achieve the same image quality. This technology is not yet common, but it is available on some new systems.

Image reconstruction

The raw data collected by a PET scanner are a list of 'coincidence events' representing near-simultaneous detection (typically, within a window of 6 to 12 nanoseconds of each other) of annihilation photons by a pair of detectors. Each coincidence event represents a line in space connecting the two detectors along which the positron emission occurred (i.e., the line of response (LOR)).

Analytical techniques, much like the reconstruction of computed tomography (CT) and single-photon emission computed tomography (SPECT) data, are commonly used, although the data set collected in PET is much poorer than CT, so reconstruction techniques are more difficult. Coincidence events can be grouped into projection images, called sinograms. The sinograms are sorted by the angle of each view and tilt (for 3D images). The sinogram images are analogous to the projections captured by CT scanners, and can be reconstructed in a similar way. The statistics of data thereby obtained are much worse than those obtained through transmission tomography. A normal PET data set has millions of counts for the whole acquisition, while the CT can reach a few billion counts. This contributes to PET images appearing "noisier" than CT. Two major sources of noise in PET are scatter (a detected pair of photons, at least one of which was deflected from its original path by interaction with matter in the field of view, leading to the pair being assigned to an incorrect LOR) and random events (photons originating from two different annihilation events but incorrectly recorded as a coincidence pair because their arrival at their respective detectors occurred within a coincidence timing window).

In practice, considerable pre-processing of the data is required – correction for random coincidences, estimation and subtraction of scattered photons, detector dead-time correction (after the detection of a photon, the detector must "cool down" again) and detector-sensitivity correction (for both inherent detector sensitivity and changes in sensitivity due to angle of incidence).

Filtered back projection (FBP) has been frequently used to reconstruct images from the projections. This algorithm has the advantage of being simple while having a low requirement for computing resources. Disadvantages are that shot noise in the raw data is prominent in the reconstructed images, and areas of high tracer uptake tend to form streaks across the image. Also, FBP treats the data deterministically – it does not account for the inherent randomness associated with PET data, thus requiring all the pre-reconstruction corrections described above.

Statistical, likelihood-based approaches: Statistical, likelihood-based iterative expectation-maximization algorithms such as the Shepp–Vardi algorithm are now the preferred method of reconstruction. These algorithms compute an estimate of the likely distribution of annihilation events that led to the measured data, based on statistical principles. The advantage is a better noise profile and resistance to the streak artifacts common with FBP, but the disadvantage is greater computer resource requirements. A further advantage of statistical image reconstruction techniques is that the physical effects that would need to be pre-corrected for when using an analytical reconstruction algorithm, such as scattered photons, random coincidences, attenuation and detector dead-time, can be incorporated into the likelihood model being used in the reconstruction, allowing for additional noise reduction. Iterative reconstruction has also been shown to result in improvements in the resolution of the reconstructed images, since more sophisticated models of the scanner physics can be incorporated into the likelihood model than those used by analytical reconstruction methods, allowing for improved quantification of the radioactivity distribution.

Research has shown that Bayesian methods that involve a Poisson likelihood function and an appropriate prior probability (e.g., a smoothing prior leading to total variation regularization or a Laplacian distribution leading to ${\displaystyle {\ell }_1}$-based regularization in a wavelet or other domain), such as via Ulf Grenander's Sieve estimator or via Bayes penalty methods or via I.J. Good's roughness method may yield superior performance to expectation-maximization-based methods which involve a Poisson likelihood function but do not involve such a prior.

Attenuation correction: Attenuation occurs when photons emitted by the radiotracer inside the body are absorbed by intervening tissue between the detector and the emission of the photon. As different LORs must traverse different thicknesses of tissue, the photons are attenuated differentially. The result is that structures deep in the body are reconstructed as having falsely low tracer uptake. This must be corrected for when using quantitative PET imaging.

One option for estimating attenuation coefficients is via a transmission scan, using a ⁶⁸Ge rod source that is rotated around the patient Transmission scans directly measure attenuation values at 511 keV. With PET-CT scanners the 511 keV photon attenuation coefficients can be estimated using the Hounsfield units from the CT scan. Attenuation correction in PET-MRI scanners is more challenging, as the MRI images do not directly contain information on about the tissue attenuation coefficients, and the scanner bores are too narrow for a transmission scan. Several methods have been proposed for PET-MRI attenuation coefficient, including machine learning, algorithms to directly estimate attenuation coefficients from the MRI or emission data, and atlas-based approaches (where the MRI scan is co-registered with a database of CT images).

While attenuation-corrected images are generally more faithful representations, the correction process is itself susceptible to significant artifacts. As a result, both corrected and uncorrected images are often reconstructed and read together.

2D/3D reconstruction: Early PET scanners had only a single ring of detectors, hence the acquisition of data and subsequent reconstruction was restricted to a single transverse plane. More modern scanners now include multiple rings, essentially forming a cylinder of detectors.

There are two approaches to reconstructing data from such a scanner:

1. Treat each ring as a separate entity, so that only coincidences within a ring are detected, the image from each ring can then be reconstructed individually (2D reconstruction), or
2. Allow coincidences to be detected between rings as well as within rings, then reconstruct the entire volume together (3D).

3D techniques have better sensitivity (because more coincidences are detected and used) hence less noise, but are more sensitive to the effects of scatter and random coincidences, as well as requiring greater computer resources. The advent of sub-nanosecond timing resolution detectors affords better random coincidence rejection, thus favoring 3D image reconstruction.

Time-of-flight (TOF) PET: For modern systems with a higher time resolution (roughly 3 nanoseconds) a technique called "time-of-flight" is used to improve the overall performance. Time-of-flight PET makes use of very fast gamma-ray detectors and data processing system which can more precisely decide the difference in time between the detection of the two photons. It is impossible to localize the point of origin of the annihilation event exactly (currently within 10 cm). Therefore, image reconstruction is still needed. TOF technique gives a remarkable improvement in image quality, especially signal-to-noise ratio.

Combination of PET with CT or MRI

PET scans are increasingly read alongside CT or MRI scans, with the combination (co-registration) giving both anatomic and metabolic information (i.e., what the structure is, and what it is doing biochemically). Because PET imaging is most useful in combination with anatomical imaging, such as CT, modern PET scanners are now available with integrated high-end multi-detector-row CT scanners (PET–CT). Because the two scans can be performed in immediate sequence during the same session, with the patient not changing position between the two types of scans, the two sets of images are more precisely registered, so that areas of abnormality on the PET imaging can be more perfectly correlated with anatomy on the CT images. This is very useful in showing detailed views of moving organs or structures with higher anatomical variation, which is more common outside the brain.

At the Jülich Institute of Neurosciences and Biophysics, the world's largest PET–MRI device began operation in April 2009. A 9.4-tesla magnetic resonance tomograph (MRT) combined with a PET. Presently, only the head and brain can be imaged at these high magnetic field strengths.

For brain imaging, registration of CT, MRI and PET scans may be accomplished without the need for an integrated PET–CT or PET–MRI scanner by using a device known as the N-localizer.

Limitations

The minimization of radiation dose to the subject is an attractive feature of the use of short-lived radionuclides. Besides its established role as a diagnostic technique, PET has an expanding role as a method to assess the response to therapy, in particular, cancer therapy, where the risk to the patient from lack of knowledge about disease progress is much greater than the risk from the test radiation. Since the tracers are radioactive they are generally not used with those who are pregnant.

Limitations to the widespread use of PET arise from the high costs of cyclotrons needed to produce the short-lived radionuclides for PET scanning and the need for specially adapted on-site chemical synthesis apparatus to produce the radiopharmaceuticals after radioisotope preparation. Organic radiotracer molecules that will contain a positron-emitting radioisotope cannot be synthesized first and then the radioisotope prepared within them, because bombardment with a cyclotron to prepare the radioisotope destroys any organic carrier for it. Instead, the isotope must be prepared first, then the chemistry to prepare any organic radiotracer (such as FDG) accomplished very quickly, in the short time before the isotope decays. Few hospitals and universities are capable of maintaining such systems, and most clinical PET is supported by third-party suppliers of radiotracers that can supply many sites simultaneously. This limitation restricts clinical PET primarily to the use of tracers labelled with fluorine-18, which has a half-life of 110 minutes and can be transported a reasonable distance before use, or to rubidium-82 (used as rubidium-82 chloride) with a half-life of 1.27 minutes, which is created in a portable generator and is used for myocardial perfusion studies. In recent years a few on-site cyclotrons with integrated shielding and "hot labs" (automated chemistry labs that are able to work with radioisotopes) have begun to accompany PET units to remote hospitals. The presence of the small on-site cyclotron promises to expand in the future as the cyclotrons shrink in response to the high cost of isotope transportation to remote PET machines. In recent years the shortage of PET scans has been alleviated in the US, as rollout of radiopharmacies to supply radioisotopes has grown 30 percent per year.

Because the half-life of fluorine-18 is about two hours, the prepared dose of a radiopharmaceutical bearing this radionuclide will undergo multiple half-lives of decay during the working day. This necessitates frequent recalibration of the remaining dose (determination of activity per unit volume) and careful planning with respect to patient scheduling.

History

The concept of emission and transmission tomography was introduced by David E. Kuhl, Luke Chapman and Roy Edwards in the late 1950s. Their work would lead to the design and construction of several tomographic instruments at Washington University School of Medicine and later at the University of Pennsylvania. In the 1960s and 70s tomographic imaging instruments and techniques were further developed by Michel Ter-Pogossian, Michael E. Phelps, Edward J. Hoffman and others at Washington University School of Medicine.

Work by Gordon Brownell, Charles Burnham and their associates at the Massachusetts General Hospital beginning in the 1950s contributed significantly to the development of PET technology and included the first demonstration of annihilation radiation for medical imaging. Their innovations, including the use of light pipes and volumetric analysis, have been important in the deployment of PET imaging. In 1961, James Robertson and his associates at Brookhaven National Laboratory built the first single-plane PET scan, nicknamed the "head-shrinker".

One of the factors most responsible for the acceptance of positron imaging was the development of radiopharmaceuticals. In particular, the development of labeled 2-fluorodeoxy-D-glucose (FDG—firstly synthethized and described by two Czech scientists from Charles University in Prague in 1968) by the Brookhaven group under the direction of Al Wolf and Joanna Fowler was a major factor in expanding the scope of PET imaging. The compound was first administered to two normal human volunteers by Abass Alavi in August 1976 at the University of Pennsylvania. Brain images obtained with an ordinary (non-PET) nuclear scanner demonstrated the concentration of FDG in that organ. Later, the substance was used in dedicated positron tomographic scanners, to yield the modern procedure.

The logical extension of positron instrumentation was a design using two two-dimensional arrays. PC-I was the first instrument using this concept and was designed in 1968, completed in 1969 and reported in 1972. The first applications of PC-I in tomographic mode as distinguished from the computed tomographic mode were reported in 1970. It soon became clear to many of those involved in PET development that a circular or cylindrical array of detectors was the logical next step in PET instrumentation. Although many investigators took this approach, James Robertson and Zang-Hee Cho were the first to propose a ring system that has become the prototype of the current shape of PET. The first multislice cylindrical array PET scanner was completed in 1974 at the Mallinckrodt Institute of Radiology by the group led by Ter-Pogossian.

The PET–CT scanner, attributed to David Townsend and Ronald Nutt, was named by Time as the medical invention of the year in 2000.

Cost

In Canada, PET scan funding varies by province. In Ontario, as of October 2025, there is no charge to the patient for any PET scans funded as part of the PET Scans Ontario Program. The program funds routine PET services for indications where sufficient evidence exists that the scan benefits the patient and offers advantages over other imaging tests.

In the United States, PET scan costs vary significantly by location, facility type, and insurance status. According to Medicare's 2026 payment data, a whole-body PET scan has a national average total cost of $872, with the patient paying approximately $174 after Medicare coverage.

In the United Kingdom, the National Health Service provides PET-CT scans free at point of use for eligible patients.

In Australia, Medicare subsidies are available for eligible PET services.

Quality control

The overall performance of PET systems can be evaluated by quality control tools such as the Jaszczak phantom.