Table 1 Summary of PET and fmri Methods What is imaged PET fmri Brain structure Regional brain activation Anatomical connectivity Receptor binding and regional chemical distribution Blood flow ( 15 O) Glucose metabolism ( 18 FDG) Oxygen consumption Benzodiazapines, dopamine, acetylcholine, many others Kinetic modeling Structural T1 and T2 scans BOLD (T2*) Arterial spin tagging (AST) FAIR Diffusion tensor imaging MR spectroscopy Gene expression Various radiolabeling compounds MR spectroscopy with kinetic modeling
Table 2 Relative advantages of PET and fmri PET Mapping of receptors and other neuroactive agents Direct measurement of glucose metabolism No magnetic susceptibility artifacts Quiet environment for auditory tasks Imaging near fluid spaces FMRI Repeated scanning Single subject analyses possible Higher spatial resolution Higher temporal resolution Single trial designs Easily combined with ERP and other measurements because there is no magnetic field Estimation of hemodynamic response and separation of stimulus and task set related variables Lower cost
Figure Captions Figure 1. A graph showing the results of a search of the Medline database for articles with the words functional Magnetic Resonance Imaging (fmri) in the title. Figure 2. A spatial and a verbal task used to study item-recognition performance in working memory. Note that the two tasks are similar in structure except for the material that must be retained and retrieved. Figure 3. Lateral and superior images revealing activations in spatial and verbal working memory tasks. In each row, three views of the brain in grey-scale renderings of a composite MRI have superimposed on them activations from a PET experiment, where the activations are shown in a color scale with blue the least active and red the most active. Figure 4. A schematic of a task used to study processes required to switch between two tasks. The figure shows that the task entails two types of switches, between different internal counters or between different operations on the contents of those counters. Figure 5. Behavioral data from the dual-switching task. Note that there are main effects of both types of switch, and that there is no interaction between these two separate effects. Figure 6. One contrast in brain activations between the two types of switches in the dualswitching task. The top panel shows activation in a ventromedial prefrontal site and the bottom panel shows activation in a lateral prefrontal site for each type of switch. Note the double dissociation in patterns of activations in these two sites for the two types of switch. Figure 7. Measured rcbf responses in three areas across six conditions, one rest condition and five levels of increasing difficulty in the Tower of London task. These areas showed linear
increases in rcbf with increasing difficulty, whereas other areas (such as visual cortex) showed a response to task versus rest but no changes among the five difficulty levels. Reproduced from Dagher et al. (1999). Figure 8. A schematic diagram of the main components of a PET scanner. Figure 9. The PET Scanner. Each detector counts the number of anhilation events that take place in a column of tissue. The column can be subdivided into smaller units that represent the image pixels. The detector counts the sum of the events in each of the elements in the column. Figure 10. PET image reconstruction. The raw data are a set of projections (sums) at different angles as shown in A. Backprojecting the raw data onto the image means adding the numbers of counts in the projection to the pixels that are aligned with each point in the projection, as shown in B. An image can be obtained after the data from all the projections has been added as shown in C. Figure 11. Representation of the proton s magnetization Figure 12. The Spin Ensemble Figure 13. Tipping the magnetization vector from the z-axis onto the xy-plane. The duration and strength of the B1 field determine how far the vector is tipped (ie the flip angle ) Figure 14. The dephasing process occurs because all the spins in the ensemble do not precess at the exact same rate. Some of them get a ahead, and some of them lag behind. The net effect is that they start canceling each other out, shortening the length of the magnetization vector.
Figure 15. The same slice of brain tissue can appear very different, depending on which relaxation mechanism is emphasized as the source of the contrast in the pulse sequence. Using long echo times emphasizes T2 differences between tissues, and shortening the repetition time emphasizes T1 differences in tissue. Left: one slice of a T1 image. Right: the same slice acquired as a T2 image. Figure 16. Refocusing of the spins by Gradient Echoes and Spin Echoes. Figure 17. The BOLD response to a single event is shown in the top portion of the figure. This is commonly referred to as the hemodynamic response function, or HRF. A train of events, like the one shown in the middle figure, would produce a BOLD response like the one shown in the bottom part of the figure. Figure 18. The Design Matrix should include all the significant effects that are present in the experiment. Each effect is represented by a column of data containing the expected time series that one would see if that were the only effect present.
Figure 1
Figure 2
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10 A C B
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Figure 16
Figure 17
Figure 18