Cognitive Neuroscience: Methods & Strategies of Research PDF

Cognitive Neuroscience: Methods & Strategies of Research I. The turn to the brain in Cognitive Science II. Methods of studying functional specialization in brain A. Early research on structure and function of the brain 1. Lesion studies 2. Brodmann areas III.Mapping the brain’s electrical activity A...

Cognitive Neuroscience: Methods & Strategies of Research I. The turn to the brain in Cognitive Science II. Methods of studying functional specialization in brain A. Early research on structure and function of the brain 1. Lesion studies 2. Brodmann areas III.Mapping the brain’s electrical activity A. Communication within a neuron B. Single-cell recordings C. Measuring the activity of populations of neurons 1. EEG 2. MEG IV. Neuroimaging techniques A. CT or CAT scan B. PET scan C. fMRI D. Knife-edge scanning microscope V. Brain stimulation techniques A. Transcranial magnetic stimulation B. Penfield studies C. Other brain stimulation techniques VI. New developments: optogenetics VII. Combining resources VIII. Some problems with neuroimaging data IX. Optimizing use of techniques The Turn to the Brain in Cognitive Science ✧ Early models of cognitive functions, such as visual perception, focused on top-down analysis and included relatively little discussion of neural implementation ✧ Neuroimaging techniques that emerged in the 1980s and 1990s, such as PET and fMRI, allowed neuroscientists to begin establishing large-scale correlations between types of cognitive functioning and specific brain areas ✧ Other techniques, such as single-cell recordings, have made it possible to study brain activity in nonhuman animals at the level of the single neuron Methods for Studying Functional Specialization in Brain Lesion studies: Examination of effects of destruction of brain tissue through stroke, accident, or surgery One of earliest methods of studying functional specialization (functions of particular brain regions) Ø Phineas Gage – effects of frontal lobe damage Ø Man who died laughing Early Research on Structure and Function of the Brain The cerebral cortex is divided into segregated areas with distinct neuronal populations These different regions are distinguished in terms of the types of cell they contain and the density of those cells In 1909, the German neurologist Korbinian Brodmann identified 52 different cortical regions using staining techniques − Ex: Brodmann area 17 in occipital lobe is area V1, also known as the primary visual cortex or striate cortex ➜ This anatomical classification of neural areas can serve as a basis for classifying cortical regions according to their function, but does not provide a map of anatomical connectivity Exploring Anatomical Connectivity Tract tracing can be used to explore anatomical connectivity A marker chemical, such as radioactive amino acids or horseradish peroxidase (HRP), is injected into a particular brain region Looking to see where the marker ends up allows us to identify where the cell projects to However, it is only possible to see where the marker winds up by examining sections of the cortex through the microscope, so this cannot be done on living creatures Using this technique, anatomical wiring diagrams have been developed for monkeys, rats, and cats Limitations Anatomical wiring diagrams do not carry any information about the direction of information flow between and across neural regions Anatomical connectivity is studied almost completely independently of cognitive functioning, such as how different brain regions might form circuits to perform particular information-processing tasks v Techniques for studying human anatomical connectivity in vivo, such as diffusion tractography are being developed, but are still in their infancy Mapping the Brain’s Electrical Activity Cognitive functioning involves the coordinated activity of networks of different brain areas Identifying these networks requires going beyond anatomical activity by studying what goes on in the brain when it is performing particular tasks One way to do this is by mapping the brain’s electrical activity Electrical activity is a good index of activity in neurons When neurons fire, they generate electricity Communication within a Neuron: Electrical Transmission Axons have two basic electrical potentials: resting potential and action potential Resting membrane potential (-70mV): occurs because various ions are located in different concentrations in fluid inside and outside the cell – Extracellular fluid: rich in Na+ and Cl– Intracellular fluid: rich in K+ and various organic anions (A-) Action potential: the brief electrical impulse that provides the basis for conduction of information along an axon – Sodium channels open: Na+ rushes into the cell, driven by force of diffusion and electrostatic pressure – Entry of positively charged ions causes membrane potential to reverse, so the inside becomes positive – Sodium channels close again – Potassium gates open: depolarization caused by influx of Na+ causes K+ to move outside the cell, quickly bringing the membrane potential back to its resting value – Potassium channels close again – Na+ is pumped back out and K+ pumped back in by sodium-potassium pumps Action potential of a given neuron is an all or none phenomenon Once an action potential is triggered in an axon, it is propagated without decrement to the end of the fiber Neurons represent intensity by their rate of firing There are various techniques for measuring the brain’s electrical activity… Single cell recordings: Microelectrode is placed close to an individual neuron to record the discharge of action potentials in the cell Has been used to identify neurons that respond to particular stimuli Ø Ex: mirror neurons fire both when a monkey performs a specific action and when it observes that action Single-neuron recordings of a mirror being performed by an observer neuron: The neuron fires both when the monkey grasps food and when the monkey observes the experimenter grasping food Measuring the Electrical Activity of Populations of Neurons Electroencephalogram (EEG) Electrodes are attached to the skull and wired up to a computer Technique provides amplified recording of the waves of electrical activity that sweep across the brain's surface − Each electrode is sensitive to the electrical activity of thousands of neurons, with the closest neurons making the largest contribution − The coordinated activity of these neural populations are seen as oscillatory waves at different frequencies o Useful in clinical contexts for diagnosing epilepsy and tumors EEGs also provide way of measuring event-related potentials (ERPs) − ERP activity is provoked by a specific stimulus − ERPs have very fine temporal resolution but low spatial resolution v Most widespread and least expensive technique for studying electrical activity of a large population of neurons Gamma Very high frequency waves of 26-42+ Hz Thought to signal active exchange of information between cortical and other areas of the brain May be seen in conscious state or during REM Beta Irregular, low-amplitude waves of 12 to 25 Hz Most evident in frontal and usually seen on both sides of brain in symmetrical distribution Associated with a state of arousal or alertness Alpha Synchronous waves of 7.5 to 13 Hz Associated with relaxed awake state Theta Delta Synchronous waves of 3.5 to 7.5 Hz Transition between sleep and wakefulness Associated with deep meditation, hynagogic state, creativity, and memory retrieval Low frequency, high-amplitude waves of less than 4 Hz Occurs during deepest stages of sleep and loss of consciousness (e.g., coma) MEG (magnetoencephalography) Large numbers (up to 300) of magnetically sensitive sensors are placed on scalp Measures magnetic fields created by brain’s electrical activity Allows finer spatial resolution than is possible with EEGs Also less susceptible to distortions from skull than EEG However, must be carried out in room specially constructed to block all alien magnetic influences, including the earth’s magnetic field, so very expensive ➜ Currently primarily used in medical diagnosis Neuroimaging Techniques CT (computed tomography) or CAT Scan Series of x-ray photographs is taken from different angles and combined by computer into a composite representation of a slice through the brain or body PET (positron emission tomography) Scan Visual display of brain activity that shows where a radioactive form of glucose goes while the brain performs a given task The radioactive isotope decays into a nonradioactive atom after about a minute There may be all sorts of activity going on in the brain that are not specific to the particular experiment that the participant is performing ➜Ways must be found to filter out potentially irrelevant, background activity Ø Ex: Asking people to look at words flashed on a screen without responding, then asking then to say the words out loud fMRI (functional magnetic resonance imaging) Has superseded PET in many domains Technique uses magnetic fields and radio waves to produce computergenerated images that show level of activity in different parts of the brain Oxygenated and deoxygenated blood respond differently to the magnetic field, so fMRI can detect increases in blood oxygen, providing a measure of blood flow and cognitive activity − Difference between oxygenated and deoxygenated blood is known as the BOLD (blood oxygen level dependent) contrast − fMRI measures the BOLD signal In early 1990s, event-related fMRI emerged: able to measure the BOLD signal associated with individual rapid occurring neural events Ø Ex: Which areas of brain show increased activation when one is viewing a picture that is subsequently well remembered vs. one that is simply judged familiar or not remembered at all? Participants viewed 96 color pictures of indoor and outdoor scenes while in an fMRI scanner 30 minutes later, they were given an unanticipated memory test in which they were shown 128 pictures, including the 96 previously viewed, and asked to identify which they had seen before and how confident they were Pictures were classified as – Well remembered – Familiar – Forgotten Which brain areas were correlated with levels of memory performance for individual events? Parahippocampus and the right dorsolateral prefrontal cortex (Brewer, Zhao, Desmond et al., 1998) More recent research compared human fMRI data with single electrode recordings of individual neurons from monkeys to provide information about what fMRI actually measures Ø Researchers found that the BOLD response from fMRI directly reflects the average firing rate of neurons in the relevant brain area (Rees, Friston & Koch, 2000) Ø However, follow-up research indicated that the BOLD signal that is measured by fMRI correlates more with the input to neurons (as indicated by local field potential or LFP) than with the output (Logothetis, 2001) Knife-Edge Scanning Microscope (KESM) Diamond-tipped knife slices through tissue White light source illuminates the strip of tissue as it comes off the blade, reflecting an image to a camera Computer system produces 3-D reconstruction down to the cellular level ➜KESM is capable of digitizing the entire volume of a mouse brain Blood vessels in mouse brain Brain Stimulation Techniques Repetitive transcranial magnetic stimulation (rTMS): An intense pulse of magnetic energy is sent through a coil placed on the surface of the skull, resulting in the electrical firing of neurons beneath the scalp Can briefly enhance or disrupt neural activity rTMS applied to left (and sometimes right) prefrontal cortex reduces symptoms of depression without any side effects in about 30-40% of patients with depression (Becker, Maley, Shultz et al., 2016; Brunoni, Chaimani, Moffa et al., 2017; Taylor, Bhati, Dubin et al., 2017) − Treatment is performed on wide-awake patients for 20-30 min daily for 2-4 weeks − Stimulation may energize the brain’s left frontal lobe, which is relatively inactive during depression and cause nerve cells to form new functioning circuits Penfield studies Electrical stimulation of association areas of brain during open brain surgery while patient is fully conscious Electroconvulsive therapy (ECT) Person is anesthetized and given a drug which paralyzes the muscles Electrodes are placed on patient’s scalp (usually to non-speech-dominant hemisphere) A jolt of electricity is applied for one twentyfifth of a second, triggering a brief seizure Usually consists of 3 treatments per week for 2-4 weeks May work by increasing release of norepinephrine or by calming neural centers that are overactive in depression Some research indicates that ECT stimulates neurogenesis and new synaptic connections within the hippocampus and amygdala (Joshi, Espinoza, Pirnia et al., 2016; Rotheneichner, Lange, O’Sullivan et al., 2014) Advantages: – 70% people with depression who don’t respond to other treatments get relief with ECT – Improvement in symptoms much more rapid (within a few days) than with antidepressants – Credited with saving many from suicide o ECT is sometimes used in interim period before antidepressant drugs become effective in suicidal patients o ECT may be court-mandated in certain cases Disadvantages: – Prolonged and excessive use causes brain damage, resulting in long-lasting impairments in memory – High relapse rate Other potential new treatments for depression: Transcranial direct current stimulation (tDCS): − Weak current is applied to the scalp on the left side of dorsolateral prefrontal − Safer than ECT (Brunoni, Moffa, Fregni et al., 2016; Mutz, Vipulananthan, Carter et al., 2017) Deep brain stimulation: − Neurosurgical procedure involving placement of a neurostimulator (sometimes referred to as a “brain pacemaker”), which sends electrical impulses, through implanted electrodes, to specific targets in the brain (Bergfeld, Mantione, Hoogendoorn et al., 2016) Optogenetics Involves inserting opsin genes into neurons This causes neuron to manufacture light-sensitive opsin proteins and incorporate them into the membrane These opsin proteins become part of the ion channels in cell membrane that control whether neuron fires or not The neuron can then be activated by a particular wavelength of light (say red) Other opsin proteins can cause the neuron to produce a flash of light of a particular wavelength (say green) when it is activated Ø Producing hallucinations in mice using optogenetics (Marshel, Kim, & Machado, 2019) – Mice were shown pictures of vertical and horizontal stripes on a monitor – Trained to lick pipe only if they saw vertical stripe – Using light, researchers identified neurons in visual cortex that switched on in response to vertical stripes vs. those that responded to horizontal stripes – They then turned off the monitor, leaving mice in darkness – Scientists than used light to switch on the neurons for vertical stripes ➜ Mice responded by licking the pipe, as if they were actually seeing vertical stripes! Ø Implanting false memories in mice using optogenetics (Liu, Ramirez & Tonegawa, 2014) – Neuroscientists tagged neurons associated with a certain memory (i.e., fear of a certain location where they had received electric shock) – Then, using light, they artificially induced those neurons to fire to make new associations between events and environments with no ties to reality (i.e., fear of a different location) Ø Scientists have also changed the songs that young zebra finches sing by implanting new memories in their brains (Zhao, Garcia-Oscos, Dinh et al., 2019) v Technique can be used for precise identification of specific neurons and neural networks v May also potentially be used for treatment of depression, chronic pain, seizures, or restoration of vision in the blind in future J Combining Resources ✧ PET and fMRI neuroimaging are better at providing information about brain regions involved in cognitive activity ✧ EEG is better at providing info about the precise sequence of events as information is processed ✧ MEG can measure both time course and underlying neural substrate directly ➜ There is no technique for studying large populations of neurons that has both high temporal resolution and high spatial resolution – So neuroscientists have combined techniques in order to gain a more comprehensive perspective Some Problems with Neuroimaging Data There can be noise in the system, especially when the voxel size is large Voxel is a three-dimensional version of a pixel (from “volume” and “pixel”) The smaller the voxel, the higher the spatial resolution, but the lower the signal strength − It is often necessary to increase the voxel size to capture small fluctuations in the BOLD signal − However, increasing the voxel size increases the range of different types of brain tissue occurring in each voxel, e.g., white matter or cerebrospinal fluid, and this can distort the signal Everyone’s brain is slightly different Distortions can occur when data are being normalized to allow comparison across participants Optimizing Use of Techniques Neuroimaging techniques are Good at telling us about functional connectivity or which brain areas are simultaneously active while participants are performing a particular task Not so good at telling us about effective connectivity or order in which brain regions are activated how they influence each other However, models of effective connectivity can be developed by combining different techniques, such as neuroimaging with EEG It is also possible to design a series of experiments in a way that they yield information about flow of information Ø Ex: PET experiments that use paired-subtraction paradigm can reveal information about stages of lexical processing – Look at areas that are active when participants are looking at a word vs. saying a word out loud Network Neuroscience There has been a shift in recent years to studying networks or functional connectivity, that is, how different brain regions work together, rather than just brain regions themselves Traditional, localizationist research almost always involves watching how brain activity changes while a person is engaged in a particular task Network research, in contrast, can be done when people are doing nothing at all This gets closer to a person’s natural state ➜ Ex: Someone with a psychological disorder will have the disorder even when they are not doing a working memory task Ø The network approach has proven to be particularly well suited to understanding schizophrenia − Researchers have found that in schizophrenia, the different regions of the brain aren’t as densely connected (Liu, Lian, Zhou et al., 2008; Lynall, Bassett, Kerwin et al., 2010) Other important applications of network neuroscience: Networks can be used not just for diagnosis, but to identify PTSD patients who are unlikely to respond to psychotherapy (Etkin, Maron-Katz, Wu et al., 2019) If scientists can determine the circuits that a highly invasive technique like deep brain stimulation is acting upon, they might be able to achieve similar results with a nonsurgical approach like TMS − Clinicians can access regions buried in the brain, like those targeted in DBS treatments for Parkinson’s, through areas closer to the surface − Also, it might be that the best way to help a symptom that maps to a circuit is actually multiple electrodes, or multiple stimulation sites Ø Tumor problem (Michael Fox, neurologist at Harvard Medical School)