Functional neuroimaging with PET predates the development of fMRI. PET imaging takes advantage of the fact that unstable elements (such as 150, nC, or 18F, which possess too few neutrons relative to protons) go through a rapid process of decay involving the release of a positron (positively charged electron) from the nucleus. Once released, the positron collides with an electron, which causes the annihilation of both the electron and the positron and the production of two high-energy (511 keV) photons that travel at 180° from each other (see Figure 13.2). PET cameras consist of rings of crystals that produce light scintillation
scintilating crystal photomultiplier tube
Coincidence detection circuit scintilating crystal photomultiplier tube
Coincidence detection circuit
FIGURE 13.2. Measurement of positron emissions. A) An unstable 150 nucleus emits a positron that collides with an electron, releasing a pair of high-energy 511 keV photons at a 180° angle. The photons are detected by an annihilation photon detector, which is comprised of crystals that scintillate when struck by a photon and photomultiplier tubes that transform the light emitted by the crystals into an electrical impulse. When two detectors 180° apart are activated, the coincidence is registered and sent on for signal processing and image reconstruction. B) Rings of annihilation photon detectors are arrayed around a subject's head. PET scanners have multiple rings arrayed in parallel, allowing multislice data collection. Detection of coincident scintillations at 180° angles within a ring (in 2-D imaging) or across rings (in 3-D imaging) allows identification of the approximate location from which the positron emitting radiotracer is located. Figure from pages 62 and 63 of Images of Mind by Michael I. Posner and Marcus E. Raichle. Copyright 1994, 1997 by Scientific American Library. Reprinted by permission of Henry Holt and Company, LLC.
when penetrated by photons (Raichle, 1983). This scintillation is then converted to electrical impulses that can be amplified and analyzed. Within the crystals' range of sensitivity, there exists a direct relationship between the concentration of radiotracer present in a brain region and the level of photon detections arising from that region.
PET allows assessment of multiple aspects of brain functioning depending on the radiotracer used. Importantly, PET can be used to measure regional cerebral blood flow (rCBF). Recall that when neurons in a brain region become active, they increase their oxygen consumption, which is compensated for by a substantial increase in rCBF (Figure 13.1a). This increase in blood flow exceeds the oxygen consumption demanded by the neurons, making rCBF a particularly robust index of regional neural activity (Fox & Raichle, 1986; Fox, Raichle, Mintun, & Dence, 1988). When unstable 150 is attached to H2, it can be injected directly into the bloodstream. Once in the bloodstream, H2150 will travel wherever the blood travels, such that areas with the highest levels of rCBF will emit the most positrons (Herscovitch, Markham, & Raichle, 1983). Thus, by measuring 150 positron emissions, we can index neural activity. Indeed, the measurement of rCBF with 150 PET represents a far more simple and direct index of neural activity than the BOLD response, which is influenced by several different features of the hemodynamic response (i.e., blood volume, flow rate, and oxyhemoglobin-deoxyhemoglobin ratios). The directness of the relationship also makes PET less sensitive to some of the artifacts associated with fMRI discussed at greater length later.
The temporal window measured in PET studies is directly linked to the speed at which the radiotracer decays. To get adequate signal-to-noise ratios, the detected positron annihilations are aggregated over time. With 150, which decays rapidly, one typically scans for 30 seconds to 90 seconds to achieve adequate signal-to-noise ratios. The data from these scans thus represents the aggregate of activity during this window, with the largest weighting occurring earlier in the scan when positron emissions are highest (Sil-bersweig et al., 1993). Therefore, the minimum temporal resolution of 150 PET is on the level of about 30 seconds. In cognitive studies, this dictates that tasks need to Engage Brain regions for a substantial portion of a 30-second to 90-second scan window if they are to produce robust changes in rCBF measurements.
In addition to measuring rCBF, PET can also be used to measure glucose metabolism in the brain, which provides an even more direct index of neural activity. Glucose metabolism is assessed by labeling a deoxygenated form of glucose with 18E 18F-deoxyglucose (FDG) is injected into the bloodstream, and the FDG is taken up by brain regions in direct proportion to their metabolic demands (Raichle, 1988). In contrast to 150, the slower decay of 18F requires imaging over substantially longer temporal windows, requiring tasks to be carried on for 20 minutes or longer. Because of this, 150 provides the primary tool for studying brain activations, whereas FDG is more frequently used to make baseline (resting) comparisons between different subject populations.
Radiotracers can also be created by tagging lig-ands or precursors for various neurotransmitter systems with 18F or UC (Fowler, Ding, & Volkow, 2003). These radiotracers allow for the assessment of many of the major neurotransmitters systems, providing the ability to detect individual or group differences in neuroreceptor density, transporter density, and even neurotransmitter synthesis (see Table 13.1). This has proved highly useful both for research and, in some cases, clinical diagnosis such
Was this article helpful?