Quick viewing(Text Mode)

Cognition: an Overview of Neuroimaging Techniques 1063

Cognition: an Overview of Neuroimaging Techniques 1063

Author's personal copy

Cognition: An Overview of Techniques 1063

Cognition: An Overview of Neuroimaging Techniques

S A Bunge , University of California at Berkeley, standard approach is to compare two task variants Berkeley, CA, USA that differ only in that task 1 but not task 2 is I Kahn, Howard Hughes Medical Institute at Harvard hypothesized to engage a specific cognitive process. University, Cambridge, MA, USA This approach relies on the assumption of ‘pure inser- ã 2009 Elsevier Ltd. All rights reserved. tion’; that is, the assumption that the insertion of the new task requirement affects only the cognitive

process of interest.

Anatomical Techniques Classes of Techniques

Anatomical techniques are used in a variety of ways A variety of noninvasive functional neuroimaging in the service of the study of cognition, for example, techniques are available for use in humans (Figure 1); to localize neuropathies in patients with cognitive these techniques are categorized into two main clas- disabilities or to compare the size of specific ses. The first consists of methods for directly mea- structures between groups of subjects through volu- suring electrical activity associated with neuronal metric analysis. Additionally, anatomical techniques firing, such as (EEG) and are used in conjunction with functional techniques in (MEG). The second main order to localize brain activity. class consists of methods for indirectly measuring The earliest technique for brain structure neuronal activity, which operate under the principle was computed tomography (CT). Today, it has been that neural activity is supported by increased local largely replaced by the much more powerful tech- flow and metabolic activity. These methods nique of magnetic resonance imaging (MRI). MRI include emission tomography (PET), func- provides excellent detailed structural information tional magnetic resonance imaging (fMRI), and near- and enables the naked eye to distinguish gray matter spectroscopy (NIRS). More information on (neuronal cell bodies) from (myelinated each of these methods is provided in the following tracts). New anatomical techniques, such as diffusion sections. tensor imaging, have been developed to specifically visualize myelinated tracts. These methods can be Direct measures of neural activity: EEG and MEG used to track the normal and abnormal development EEG is the oldest functional brain imaging technique, of neural pathways in childhood. dating back to Berger’s discovery in 1929 that brain

electrical activity could be recorded from Functional Techniques placed on the scalp. This technique is still widely used today because of its ability to provide real-time mea-

Functional techniques have been the dominant force surements of brain activity. EEG records transient in cognitive because they enable us to electrical dipoles generated by the net flow of electri- determine when and where neural activity in the cal current across the cellular membrane during neu- brain is associated with the ability to perform a par- ronal depolarization associated with postsynaptic ticular cognitive task. These techniques allow us to potentials. Global EEG is used to measure neural examine the workings of the throughout activity during different brain states, such as sleeping the life span, in sickness and in health. and waking. A more powerful tool for cognitive neu-

It is common to design several task variants that roscience than global EEG measurements consists of differ slightly in terms of task requirements. On the event-related potentials, which refer to EEG activity basis of this approach, differences in brain activation averaged over a series of instances (or trials) triggered between the task variants enable us to isolate brain by the same event (e.g., the presentation of a visual structures that are implicated in hypothesized cogni- ). tive processes. For example, if one wanted to identify Whereas EEG records the electrical activity asso- brain regions involved in actively maintaining infor- ciated with neuronal depolarization, the newer tech- mation in (referred to as ), one nique of MEG records the magnetic field produced by might parametrically vary the number of items (or the this electrical activity. The signal measured by both load) that subjects needed to maintain and then iden- techniques results primarily from the activity of pyra- tify brain regions in which the level of activation midal , which constitute roughly 70% of cells varied according to working memory load. Although in the neocortex and are oriented perpendicular to parametric variation is arguably the most powerful the cortical sheath. EEG records electrical activity functional brain imaging manipulation, a more oriented perpendicular to the surface of the brain,

Encyclopedia of Neuroscience (2009), vol. 2, pp. 1063-1067

Author's personal copy

1064 Cognition: An Overview of Neuroimaging Techniques

0 PET Brain −1 EEG + MEG

Lesions −2 fMRI Micro- Optical lesions −3 imaging Column 2-Deoxyglucose −4 Layer

(m) Log size

Neuron −5 Single unit Light −6 Patch clamp microscopy

Synapse −7 −4 −3 −2 −101234567

ms s min h d

Log time (s)

Noninvasive Invasive Figure 1 Relative spatial and temporal sensitivities of different functional brain imaging techniques. The level of invasiveness of each technique is indicated by a grayscale; more highly invasive techniques are shown in darker gray. Techniques used in animals only are outlined with a dashed line. Adapted from Cohen MS and Bookheimer SY (1994) Localization of brain function using magnetic resonance Trends in Neuroscience imaging. 17(7): 268–277. whereas MEG records activity oriented parallel to the from the atom to give an unstable form that will 16 surface of the brain. Thus, EEG measures the activity emit one positron to revert to the stable form O2. of pyramidal cells in cortical gyri and the depths of In a PET scan, a small amount of radioactive tracer is the sulci, whereas MEG is sensitive primarily to the injected into a vein. The tracer enters the brain after activity of pyramidal cells in the superficial parts of approximately 30 s, and in the following 30 s radia- the sulci, and it is therefore more limited in its scope. tion in the brain rises to its maximal value; a picture of The major challenge with both EEG and MEG is the rCBF is taken during this time frame. referred to as the ‘,’ which is the The greatest advantage of PETover the more recent challenge of identifying the source of the underlying method of fMRI is the choice of radioactive tracer. signal. This source can be a great distance from the Researchers can synthesize radiopharmaceutical point on the scalp at which it is measured, and it is compounds that bind to dopamine or serotonin affected by factors such as head shape and dipole receptors (C-11 or F-18 N-methylspiperone), opiate location and orientation. Thus, it is necessary to receptors (C-11 carfentanil), etc. PET is likely to con- build source localization algorithms to determine tinue to be important for understanding the role of the likely source of a signal. One advantage of MEG various in cognition. However, over EEG is that the signal is less sensitive to factors there are several disadvantages of PET that have led such as head shape. Thus, the two techniques are to its being surpassed by fMRI as the most widely complementary in that EEG samples from more neu- used indirect measure of brain activity. The first of rons, but MEG is less sensitive to signal distortion. As these is cost; PET facilities require not only a PET such, some researchers take the approach of measur- camera but also a , which is used to produce ing EEG and MEG simultaneously to take advantage the radioactive tracers. The second disadvantage is of what each technique has to offer. the poor temporal resolution – on the order of 1 min compared to 6–8 s for fMRI. Finally, PET is more Indirect measures of neural activity: PET Roy and invasive than fMRI, requiring the injection of a radio- Sherrington first showed in 1890 that brain stimula- active tracer (albeit one with a half-life of a few tion led to a local increase in blood flow to active minutes), and thus it is not suitable for use in children populations of neurons. Landau and others sub- or other special populations. sequently used radioactive tracers to measure regional cerebral blood flow (rCBF) in animals (1955), and in Indirect measures of neural activity: fMRI fMRI is 1963 this technique was first applied to humans. The currently the most widely used brain imaging tech- 15 most widely used radioactive tracer is O2, an oxygen nique for a number of reasons, including (1) wide- molecule in which one electron has been removed spread availability of MRI scanners and technology,

Encyclopedia of Neuroscience (2009), vol. 2, pp. 1063-1067

Author's personal copy

Cognition: An Overview of Neuroimaging Techniques 1065

(2) relatively low cost per scan, (3) the lack of recog- changes in these optical properties are likely to be due nized risks for properly screened subjects, (4) good to changes in light scattering as a function of neuronal spatial resolution, and (5) better temporal resolution activity. Because this type of optical signal is directly than other indirect neuroimaging methods. linked to neural activity, rather than blood flow, it

In addition to observing that brain activity was provides a much higher temporal resolution than the associated with local changes in blood flow, Roy signal typically analyzed in NIRS studies. Addition- and Sherrington noted that the increase in total oxy- ally, the EROS signal has a high degree of spatial genated blood delivered far exceeded the demand. It resolution, in that it can be localized to an area of 3 is this surplus of oxygen that is detected with fMRI, less than 1 cm . with what is known as the blood oxygen-level depen- A drawback of all three optical imaging techniques dent (BOLD) contrast. Oxygenated , or noted previously is that they can only be used to oxyhemoglobin, has magnetic properties different measure activity on or near the surface of the brain from those of deoxyhemoglobin. Thus, the deployment (within 3–5 cm of the surface). However, these tech- of a strong magnetic field (typically 1.5–4 tesla when niques are advantageous in that they are relatively used in humans) enables one to measure changes in inexpensive and completely noninvasive. Further- the ratio of oxy- to deoxyhemoglobin in local drain- more, because these techniques are not as sensitive ing venules and veins. An increase in blood flow in to head movement as fMRI, they hold much promise responseto a specific stimulus (be it brain stimulation for the study of clinical populations and children. or the presentation of a visual or auditory stimulus) is referred to as a hemodynamic response. This response Trade-Offs between Temporal and is sluggish with respect to the underlying neural activ- Spatial Resolution ity; for example, the hemodynamic response in visual cortex to a brief (e.g., 30 ms) visual stimulus peaks Direct measures of neural activity, such as EEG and roughly 4–6 s after the onset of the stimulus. MEG, have exquisite temporal resolution. They are

sensitive to changes in neural activity on a millisecond resolution. However, these methods have poor spatial Related Techniques resolution in that it is difficult to pinpoint the precise

Perfusion fMRI enables direct measurement of the origin of the signal. In contrast, indirect measures of hemodynamic response, unlike BOLD fMRI, which neural activity have better spatial resolution than EEG or MEG but poor temporal resolution. The relies on the comparison of hemodynamic responses under different conditions. This technique is less wi- spatial and temporal characteristics of these various dely used than BOLD fMRI because of its much lower methods are shown in Figure 1. Thus, there is cur- sensitivity. Magnetic resonance spectroscopy is used rently a trade-off between high temporal precision and high spatial precision. One solution to this prob- to measure the relative concentration and distribution of specific or compounds, such as hydrogen, lem is to use fMRI data to limit the possible sources of phosphorus, or carbon. neural activity in EEG and MEG data; another – and even more challenging – solution is multimodal imag-

Optical Brain Imaging ing, for example, the concurrent acquisition of fMRI and EEG data. NIRS operates according to the same principle as optical brain imaging techniques that have been used Advantages and Limitations of in animals. Both techniques capitalize on the fact that Neuroimaging Techniques changes in hemoglobin concentration in cortical tissue affect the absorption of infrared light by the tissue. To date, much of what we know about the functions

However, NIRS is noninvasive, whereas optical imag- of different brain regions comes from neuro- ing in animals is accomplished by exposing the surface psychological studies in humans and lesion studies in of the brain. Because the cortical surface is not animals. A distinct drawback of neuroimaging tech- exposed in NIRS, light scattering by the leads niques is that, unlike neuropsychological and lesion to reduced spatial resolution. techniques, they cannot demonstrate the necessity of During the past few years, Gratton, Fabiani, and a brain region for a specific cognitive process. How- colleagues have developed a new analytic approach to ever, neuroimaging techniques can demonstrate that the analysis of optical data, known as the event- humans without neurologic damage routinely recruit related optical signal (EROS). The EROS signal is this region to perform the task. Moreover, it is possible based on measures of the optical properties of cortical to compare levels of activation across individuals tissue, which change when the tissue is active. The and to state that greater activation of this region is

Encyclopedia of Neuroscience (2009), vol. 2, pp. 1063-1067

Author's personal copy

1066 Cognition: An Overview of Neuroimaging Techniques associated with better performance on a task. Better Finally, neuroimaging techniques with high tempo- yet, itis possible to compare activation between ral resolution, such as EEG and MEG, provide us trials within an individual and to state that on trials in with important clues about brain mechanisms. It is which the subject made an error, activation was lower helpful to say that brain regions X and Y are involved in this region than on trials in which the subject in memory retrieval, but if we can say that brain performed correctly. Thus, functional brain imaging region X is active a short time after region Y, we can techniques can be used to characterize a region’s con- begin to gain insight into the cascade of events leading tribution to specific cognitive processes. Moreover, to memory retrieval. brain imaging techniques can be paired with techniques A common criticism of neuroimaging studies is that for temporarily disrupting neural activity in a tempo- they can be highly unconstrained and atheoretical. rally and spatially precise manner (transcranial mag- Like any scientific tool, however, brain imaging netic stimulation). can be used wisely or foolishly. Certainly, many There are several important advantages of neuro- exploratory brain imaging studies have been con- imaging techniques over neuropsychological ones. ducted, particularly in the initial phase of brain imag-

First, neuropsychological studies necessarily rely on ing, when it was important to validate the new the output of behavior as the critical dependent mea- techniques. However, the second generation of imag- sure, whereas neuroimaging studies can focus on ing studies, starting in the late 1990s, has been much cognitive processes that take place prior to – or are more focused. On the whole, cognitive neuro- not associated with – a behavioral response. For scientists carefully devise experimental manipulations example, it is impossible to determine whether lesions designed to test psychological theories or specific pre- that result in a long-term memory impairment are dictions about brain function. Brain imaging techni- associated with a deficit at encoding and/or retrieval. ques are likely to be an important force in cognitive With brain imaging, however, one can identify brain neuroscience for the foreseeable future. regions associated with the effective encoding of memories separately from brain regions associated with effective retrieval. Contributions to the Study of Cognition Second, neuroimaging techniques enable us to identify the entire underlying a cogni- Neuroimaging research has been used to enrich our tive process. Lesion studies in animals can accomplish understanding of the neural basis of a wide variety this only piecemeal by lesioning each area in turn. of cognitive abilities, including , language, Neuropsychological studies in humans would be and memory. In addition, neuroimaging techniques hard-pressed to accomplish this at all, given the lim- have been used to gain insight into the etiology ited availability of patients with specific brain lesions. of neurobehavioral disorders, for surgical plann- This point is an important one because with lesion ing, and to assess functional recovery after brain studies it is possible to completely overlook – or damage. instead to overestimate – the contributions of a spe- One area in which neuroimaging studies are cific region to cognitive function. For example, the making important contributions is that of memory. has long been thought to be the primary For example, studies of long-term structure contributing to the encoding and retrieval of have begun to uncover the ways in which stored memories. However, brain imaging has demonstrated information is organized in the brain. These studies that in the healthy brain, routinely show that different attributes of an object are stored contributes to memory encoding and retrieval. in a distributed manner across several brain regions, The third point is also related to the limited avail- with visual form information stored in a region that ability of neuropsychological patients: neuroimaging processes form, and functional information stored techniques enable one to examine the role of a given near a region that processes motion. Furthermore, human brain region in cognition even if one does not studies of memory encoding have been able to pin- have access to specific patient groups. Moreover, it point the precise regions for which level of activation can be very difficult (or next to impossible, depending during stimulus processing predicts subsequent mem- on the brain region) to find a patient with a lesion ory for that stimulus. Additionally, studies of memory limited to the region in which one is interested retrieval are being used to adjudicate between models because most brain lesions are rather coarse. Further- of episodic memory positing that recollection (i.e., more, cortical reorganization can lead to recovery distinct remembrance of an item and the context in of function over time, which could lead one to assume which it was previously encountered) and familiarity mistakenly that a lesioned brain region is not (a vague sense that the item has previously been normally involved in a specific cognitive process. encountered) are either distinct processes or merely

Encyclopedia of Neuroscience (2009), vol. 2, pp. 1063-1067

Author's personal copy

Cognition: An Overview of Neuroimaging Techniques 1067 on a continuum of memory retrieval. These examples Further Reading show how brain imaging studies can be used to test or Andreassi JL (1995) : Human Behavior & Physi- adjudicate between psychological models. ological Response, 3rd edn. Hillsdale, NJ: Lawrence Erlbaum. Perhaps the most important contribution of neuro- Cohen MS and Bookheimer SY (1994) Localization of brain func- imaging to the field of cognition will be in the study of tion using magnetic resonance imaging. Trends in Neuroscience higher cognitive functions, which are not highly 17(7): 268–277. developed in nonhuman species and are therefore Frith CD and Friston KJ (1997) Studying brain function with best studied in humans. Carefully designed brain neuroimaging. In: Rugg M (ed.) , pp. 169–195. Cambridge, MA: MIT Press. imaging experiments have begun to fractionate the Kutas M and Dale A (1997) Electrical and magnetic readings cognitive processes that underlie language, reasoning, of mental functions. In: Rugg M (ed.) Cognitive Neuroscience, problem solving, and other high-level mental func- pp. 197–242. Cambridge, MA: MIT Press.

Levy I, Hasson U, Avidan G, Hendler T, and Malach R (2001) tions, but further investigation is necessary. Center-periphery organization of human object areas. Nature

Neuroscience 4(5): 533–539. See also: Cognitive Neuroscience: An Overview; Cognitive Raichle ME (1998) Behind the scenes of functional brain imaging: Control and Development; Electroencephalography A historical and physiological . Proceedings of the (EEG); Event-Related Potentials (ERPs) and Cognitive National Academy of Sciences of the United States of America Processing; Executive Function and Higher-Order 95(3): 765–772. Rorden C and Karnath HO (2004) Using human brain lesions to Cognition: Neuroimaging; fMRI: BOLD Contrast; infer function: A relic from a past era in the fMRI age? Nature Magnetic Resonance Spectroscopy; Neuroimaging; Reviews Neuroscience 5(10): 813–819. Perfusion MRI; Single Photon Emission Computed Wilkinson D and Halligan P (2004) The relevance of behavioural Tomography (SPECT): Technique; Based measures for functional-imaging studies of cognition. Nature

Morphometry. Reviews Neuroscience 5(1): 67–73.

Encyclopedia of Neuroscience (2009), vol. 2, pp. 1063-1067

Top View