Radionuclide Imaging MII Positron Emission Tomography (PET)

Size: px

Start display at page:

Download "Radionuclide Imaging MII Positron Emission Tomography (PET)"

Andrea Spencer
6 years ago
Views:

1 Radionuclide Imaging MII 3073 Positron Emission Tomography (PET)

2 Positron (β + ) emission Positron is an electron with positive charge. Positron-emitting radionuclides are most commonly produced in cyclotron by bombarding a stable element with protons, deuterons or helium nuclei. The produced radionuclides have an excess of protons and decay by the emission of positrons. A proton inside the nucleus is converted into a neutron. The excess energy is emitted as a pair of particles, a positron (β+) and a neutrino.

3 Radioisotope Atomic number (protons) Number of nucleons (protons + neutrons) Approximate half-life (min) Carbon Nitrogen Oxygen Fluorine Common positron emitters for PET and/or PET/CT imaging

4 When a positron is emitted, it travels for a short distance from its site of origin, gradually losing energy to the tissue through which it moves. When most of its kinetic energy has been lost, the positron reacts with a resident electron in an annihilation reaction. This reaction generated two 511 kev gamma photons, which are emitted in opposite direction at about 180 o from each other. In a PET scanner, these photons interact with the detector ring at opposite sites, which defines a line along which the annihilation reaction occurred and permits localization of the reaction. By using many such events, an image can be reconstructed.

5 511 kev β + e kev Positron- electron annihilation reaction

6 PET scanner/camera Dedicated PET camera contains with multiple rings detectors that consist of scintillation crystals coupled with PMTs. The ring design takes advantage of the fact that two photons detected by two opposed detectors in the ring are likely to be from a single annihilation event. Such a simultaneous detection is called a coincidence. The simultaneous detection of two photons provides location information in that the annihilation event can be assumed to occur somewhere on a line between the two detectors (the line of response; LOR).

7 β + e - Coincidence detection

8 The many coincidence events recorded by the PET scanner constitute a raw data set representing projections of the distribution of the positron radiopharmaceutical in the body. These data are then reconstructed by using a filtered back projection algorithm or an iterative algorithm to produce cross-sectional images. Because photons travel at the speed of light, PET cameras require very fast electronics to determine if two detected photons were likely produced by a single annihilation event.

9 In a PET scanner, each annihilation photon reaching a detector generates a single electronic pulse. For this photon to be accepted and used in the PET image, it must be in a specific energy range (ideally approaching 511 kev) and be paired with another photon reaching another detector simultaneously. Coincidence circuitry connecting the many detectors in the rings determines whether two such single pulses fall within a short coincidence time window, typically 6 to 12 nanoseconds. If so, they are deemed to constitute a coincidence event and are recorded in the resultant image.

10 Because of detector ring geometry and photon attenuation through scatter and absorption, many annihilation events result in one of the two 511 kev photons interacting with the PET camera detectors (single event). PET scanners use only photon pairs meeting the coincidence criterion in constructing PET images. Therefore, single events can be identified and discarded. In practice, about 99% of detected photons are rejected by the coincidence circuitry of the PET system. However, this principle of coincidence detection provides a virtual electronic collimation of the events and makes PET scanners inherently more efficient than traditional gamma cameras.

11 Events detected by PET scanners include true, scattered, and random events. All of these may be recorded as coincidences, providing both annihilation photons are actually detected and fall within the coincidence window. True coincidences are those that result when both 511 kev photons from an annihilation reaction are detected within the coincidence time window, neither photon having undergone any form of interaction before reaching the detector. These true coincidence events provide the desired information for constructing accurate images of the distribution of a PET radiopharmaceutical in clinical imaging.

12 Scattered coincidences occur when one or both annihilation photon undergo Compton interaction in body tissues and are deflected away from their expected path but still reach the detectors within the time window and are recorded as a coincidence event. Because the direction of the scattered photon has changed during the Compton interaction, the resulting coincidence event is likely to be assigned as an inaccurate LOR that no longer passes though the point of annihilation, leading to erroneous localization information and decreasing image contrast.

13 Random coincidences arise when two photons, each origination from a different annihilation reaction, reach any detector within the time window and thus appear to represent a true coincidence. Using detectors that allow very precise timing permits the recognition and exclusion of random events with a resultant improvement in image quality. If left uncorrected, both scattered and random coincidences add background to the true coincidence distribution, thereby increasing statistical noise, decreasing contrast, and causing the radioisotope concentrations to be overestimated.

14 a b c

15 There are a number of methods available to reduce the image degrading impact of scattered coincidences. Most scattered photons are not detected because they are absorbed in tissues of the body, are scattered away from the detector rings, or have lost significant energy during Compton scattering. These lower energy scattered events can be rejected by using an energy window designed to exclude photons of certain energies. The success of such rejection depends on the energy resolution characteristics of the detectors being used.

16 Because crystal detectors have only a finite energy resolution, if one were to measure only photons approaching 511 kev and exclude scattered photons of slightly different energies, a large number of true events would also be excluded, thereby either reducing image statistics or increasing image acquisition times unacceptably. Therefore, a rather broad energy window is used that allows some scattered events to be recorded as true events. Another method to reduce scatter from outside the plane of a detector ring is to use thin lead or tungsten septa positioned between the detector elements.

17 Imaging with lead septa is called 2D imaging because most of the photons counted originate in the plane of a single detector ring. 2D imaging improves image quality by reducing image noise. It also minimizes count losses due to system dead time by incidentally reducing the very large numbers of photons reaching the detectors that may occur at high count rates. However, although this reduces the number of scattered events originating outside the field of view (FOV), it also significantly reduces the true counts and increases imaging times.

18 Faster detector crystals and faster electronics in PET instruments have made imaging without septa, so called 3D imaging possible. This allows imaging from the volume defined by the entire FOV of the multiple detector rings of the camera and permits detection of true coincidence events that occur in different detectors on different rings. Compared with 2D imaging, 3D acquisitions increase sensitivity of the system by five folds or more. However, because both true coincidence and scatter rates are increases, better temporal and energy resolutions are needed to accurately eliminate scatter and random events.

19 2D imaging 3D imaging

20 PET scintillation detectors All positron systems use the principle of scintillation whereby the photon interacting with a crystal produces a flash of light, which is then detected and localized by PMTs. The ideal PET crystal detector would have: 1. high stopping power for 511 kev photons providing high efficiency and optimum spatial resolution 2. fast, intense light output with rapid decay of the light for decreased system dead time 3. good energy resolution for accurate scatter rejection 4. Low Compton scatter inside detector crystal 5. Matching of wavelength of fluorescence to response of light detector

21 Stopping power is best for crystalline materials with high density and high effective atomic number (Z value). There are several types of crystalline detector materials used for PET imaging include NaI, BGO, LSO and GSO. The light signal produced by scintillation detectors is not discrete in time but occurs over a short time interval (scintillation decay time, nanoseconds), which includes the period over which the light fades to background. Along with the speed of processing electronics, this decay time is an important determinant of system dead time.

22 Dead time is the brief period during which a crystal- PMT detector is busy producing and recording a scintillation event and having the scintillation light decay so that the next distinct scintillation event can be recognized and recorded. During this time additional arriving events cannot be processed and are lost. High count rate capability of PET instruments is particularly important in 3D acquisitions and in settings requiring high activities of very short-lived radionuclides (e.g., Oxygen-15; physical half-life 124s). Current count rate capabilities are about counts/second.

23 PET detector geometry State-of-the-art PET scanners is multiple full rings detector system that axially surround the patient (360 o ). These cameras have multiple adjacent detector rings that significantly increase the axial FOV of the patient. A larger FOV allows more counts to be detected for a standardized administered radiopharmaceutical dose and a fixed scan time by allowing more time at each table position. The most common detector arrangement consists of rings of individual detector modules of small crystal arrays or cut block scintillation crystal (usually BGO or LSO) coupled with PMTs.

24 In crystal arrays, multiple separate very small scintillation crystals are grouped together in blocks, often arranged in 6 X 6 or 8 X 8 blocks (more economic and cost-effective). These blocks are then assembled to form a crystal ring and coupled to PMTs (about four per block). For multiple rings PET camera, the intrinsic spatial resolution is a function of the crystal size. Thus, the small sizes of the crystal faces allowed by block design permits optimization of intrinsic resolution. Furthermore, a large number of small independent detectors will significantly reduce dead time count losses and allow camera operation at higher count rates.

25 Multiple rings detector Scintillator crystals PMT PMT Block detector unit

26 Sensitivity and resolution The sensitivity is defined as the recorded true coincidence rate (without scatter and random events) divided by the activity concentration (the true emitted events from the source). Sensitivity of a PET camera is determined by multiple factors like scanner geometry, crystal efficiency and photon attenuation in tissue. Spatial resolution in PET scanner is, in large part, a function of detector size, with smaller detectors increasing the resolving capability of the system. The ultimate limit of spatial resolution when using FDG is about 1 mm. However, the practical spatial resolution for clinical imaging is about 4 to 6 mm.

27 Many photons emitted from the patient (about 99%) are not detected because they are emitted in all directions from the patient and the detector rings cover only a fraction of the patient s body surface. When attenuation by absorption or scatter is considered, current systems record substantially less than 0.1% of the true events. However, because state-of-the-art PET scanners typically image in 3D mode, their efficiency for detecting emitted radiation is still considerably greater than that for SPECT imaging.

28 Time-of-flight (TOF) To improve resolution, some systems also measure time-of-flight (TOF) under the assumption that the location of the annihilation can be determined along the LOR of the coincident photons by measuring the time of arrival of each of the photons at the opposing crystals. Except the event that occurs in the exact centre of the detection ring, one of the photons will arrive before the other. The time difference will be proportional to the difference in distances traveled by the two photons and can be used to calculate the position of the event along the LOR.

29 Time-of-flight PET systems

30 Attenuation correction The attenuation in PET imaging, that is, loss of counts due to absorption of photons before they arrive at the detector, is compensated for arithmetically by using data from transmission scans. Depending on the camera design the transmission source can be a positron source, a high energy single photon source or a CT x-ray source. By using this transmission scan data, a patient specific attenuation correction map can be generated. An attenuation-corrected PET image has better image quality as compared with non-corrected one.

31 Non-attenuation corrected PET image Attenuation corrected PET image

32 PET transmission scan PET emission scan + = PET transmission image PET emission image (non-attenuation corrected) Attenuation corrected PET image

33 Standardized uptake value (SUV) Although visual assessment of PET images is often sufficient for image analysis and interpretation, the quantitative image analysis using the standardized uptake value (SUV) is also commonly used to augment the finding by measuring the degree of the tracer uptake on the area of the suspected lesion. SUV is a semiquantitative measurement that normalises the measured radioactivity concentration in a tissue to the body weight and total injected radioactivity, and a unitless ratio.

35 Limitations of PET imaging Several limitations of PET imaging are: 1. The attenuation map from the transmission is usually noisy and sensitive to the emission activity in the injected patient. 2. A long acquisition time may limit the throughput capability of the PET scanner. 3. The spatial resolution of PET images is generally poor compared with radiographs, CT or MRI image. 4. The cost of all PET instruments and facilities is relatively high as compared to other conventional imaging modalities. 5. The cost of each PET imaging is also high, which mainly depends on the cost of radiopharmaceuticals used. 6. Patient may suffer to an internal whole body exposure with a non-uniform distribution. 7. The radiation dose largely depends on the amount of injected radioactivity.

Chapter 2 PET Imaging Basics

Chapter 2 PET Imaging Basics Timothy G. Turkington PET Radiotracers Positron emission tomography (PET) imaging is the injection (or inhalation) of a substance containing a positron emitter, the subsequent