Cognitive Neuroscience: Chapter 2: The Methods of Cognitive Neuroscience
What is going on in time as someone performs a task?
- MEG - EEG/ERP - fMRI - PET
MEG
- Magnetoencephalograph (MEG) -Changes in electric current → magnetic field changes • Similar to EEG - Measure voltage change vs. magnetic field change due to same events Like a power line harder to detect than voltage change can see sulci because the magnetic field exits skull high temporal and good spatial bc of more electrodes most expensive
Transcranial magnetic stimulation
-Induction of electric current in nearby cortex -reversible lesion -temporarily impair function -limited to surface of cortex (safety and biology) -mildly invasive -spatial is not great -temporal is pretty okay
extracranial stimulation
-Transcranial magnetic stimulation
Hubel and Wiesel
-discovered feature detectors -static is neuron firing -patterns of light -receptor field -on center unit -can plot response to stimulus -Tuning curve: still respond to stimuli close to desired stimulus
NIRS (near infrared spectroscopy)
-like a remote in the head -infrared in skull -Reflection tells us about blood oxygen content -similar to fMRI signal -help prove the fMRI -limited in spatial but high in temporal -oxygenated and deoxygenated
Diffusion Tensor Imaging (DTI)
-map of water diffusion over time -noninvasive -usually done on a cadaver -usually done cadaver -like a map of complex streets -MRI
Recording single-unit activity in monkey
-neurons always fire -increase rate of firing -respond to faces
Humunculus
The part of the brain in the parietal lobe that maps the sensory input and motor output. It includes both the motor cortex and the somatosensory cortex
Compare and contrast the neuroimaging methods discussed in Chapter 2 in terms of spatial resolution and temporal resolution.
Transcranial magnetic stimulation has not great spatial resolution because it is limited to the surface of the cortex but it's temporal resolution is okay. ERP and EEG has high temporal resolution but low spatial resolution. MEG has high temporal resolution. MEG is also better at spatial resolution when it comes to sulci. The improvement is because it usually has more electrodes. PET have high spatial resolution, but it is limited by temporal resolution because it has to take so long to produce the images and the reagents have a short half-life. MRI has better spatial localization than PET by "several millimeters compared to approximately a centimeter in PET." It also has better temporal resolution i.e. "a few seconds compared to a minute or more in PET."
Optogenetics
Transgenic technique that combines genetics and light to control targeted cells in living tissue . Optogenetics combines genetics with the use of laser light to activate specific neural circuits or neuronal cell types in a variety of experimental settings, including in a freely moving and behaving animal. Thus, it has both high neuronal selectivity and high temporal resolution.
Antagonist
bind and block receptors
antagonist
block receptors
cell recording
Record of action potentials - Extracellular: outside -frequency of action potential -action potential
Midazolam is an example of
-psychoactive drug - Temporary anterograde amnesia - Affects GABA neurotransmitter - Agonist - enhances effect of GABA - used to be jet lag and drug in past - what memories are affected (hippocampus)
Slow event-related design
1st image, wait 15 s, 2nd image
What are the strengths and limitations of studying brain function direct electrophysiological recording from neurons?
A strength is that it can be used to record extracellularly or intracellularly. Extracellular recording monitors "or the electrical activity associated with action potentials generated by one or more neurons near the tip." Depending on size and placement according to the text, extracellular recording can also pick up neurons nearby and "thereby assisting in the gleaning of information about the behavior of small groups of nerve cells in a region." Intracellular recording focuses on a single neuron and "records the electrical activity from within the cell as action potentials and post-synaptic potentials are generated." This provides "more detailed information about how single neurons behave during cognitive or other function." A limitation is that of course it is only done routinely to animals and not humans. To study complex functions, researchers need monkeys which are expensive, ethically questionable, and not completely generalizable to humans.
What are the strengths and limitations of studying brain function using pharmacological perturbations?
A strength of studying brain function using pharmacological perturbations is that we can look at how certain substance abuse disorders "lead to specific changes in neurotransmission in the brain systems that underlie these function." By studying the symptoms that these substances cause short and long term, we can better understand how neurotransmitters work with our brain to create different cognitive processes. A limitation of studying people with substance abuse disorders is that the process is not controlled. Another way to study pharmacological influences is in an experimental setting. The patients are given the drug and affects are monitored. This of course is limited by the ethics of what drugs can be administered. The external validity is also limited since it is in an experiment setting.
blocked design experiment
Block of objects that are faces vs. block of objects that are clothes • Block of addition problems vs. block of problems are are multiplication problems • Block of reading and rereading vs. block of reading and self-explanations no more than 60s block you don't see the dynamics
Why are MEG and EEG complementarity with respect to the brain areas they are most sensitive to?
According to the text "MEG is sensitive mainly to neuronal activity in the cortical valleys, or sulci; it is relatively insensitive to activity in gyri." EEG can pick up activity in the gyri. EEG are a little distorted by the skull and other tissue while the MEG does not have this problem as much. They each provide their own strengths and weaknesses and together make a better picture of the brain than alone.
Why is a double dissociation more informative than a single dissociation?
According to the text, double dissociation is more informative than single dissociation because "single dissociation could arise from a general factor (such as the relative difficulty of the tasks) rather than from the engagement of separable functions." The text goes on to say however, "when another type of lesion produces the opposite, or at least quite different, characteristics, the case for the engagement of functionally distinct neural systems becomes much stronger."
Describe the BOLD response measured in fMRI. What exactly is being measured in an fMRI study?
BOLD stands for blood oxygenation level-dependent. According to the text, "oxyhemoglobin (hemoglobin carrying a bound oxygen molecule) and deoxyhemoglobin (the oxygen-depleted form of hemoglobin) have different magnetic resonance signals." If an area is using more oxygen, it is more active. Basically, oxygen rich blood rush to an active site where then it becomes deoxygenated. This results in a change in their magnetic resonance signals which can be monitored by an fMRI study. Another limitation is that standard fMRI do not usually take into account the cortex's local organization.
Benefits and limitations to lesions
Benefits -tentative causation -learn from tragedy -recovery Limitations -less generalizable, quasi-experiment, might be form connections
benefits and stimulation to stimulation
Benefits -temporary -non-invasive -controlled Limitations -lower spatial resolution -can't see below cortex
Psychoactive drugs
Caffeine, nicotine, alcohol, etc. Aganoist - acts as neurotransmitter Antagonist - blocks receptors
Wilder Penfield
Canadian neurosurgeon who used electrodes and electrical stimulation techniques to "map" out different parts of the brain during surgery -intracranial stimulation -humunculous -somatosensory cortex - neurosurgery - circa 1950 - mapping of the sensory and motor strips • Precentral and postcentral gyri
Intracranial stimulation
Direct stimulation - Animals - Preparation for surgery - Enhance or disrupt neural function • Wilder Penfield
Functional magnetic resonance imaging (fMRI)
Does NOT measure neural activity - At least not directly • Measures magnetic susceptibility changes due to blood oxygenation - Oxygenated hemoglobin vs. deoxygenated hemoglobin • Neurons fire → signal for increased blood flow → increase in proportion of oxy vs. deoxy hemoglobin → measured signal • Has replaced PET in most cognitive work - Better temporal/spatial resolution, much less invasive -Measure of statistically different compared to baseline
What are the similarities and differences between EEG, ERP, and MEG?
EEG, MEG, and ERP are all noninvasive ways to measure electrophysiological brain activity. According to the text, "EEG recordings measure electrical brain waves that can be detected at the scalp." They are noninvasive and measure the "dendritic field potentials." EEG's pass voltage through electrodes on the scalp. This allows the researcher to assess the overall state of the brain. EEG's are limited in "investigating specific cognitive functions" because it is "linked in time to any particular cognitive process or event." ERP's are "small voltage fluctuations in an ongoing EEG triggered by sensory and cognitive events; they reflect the summed electrical activity of neuronal populations specifically responding to those events." This provides high temporal resolution of neural processes according to the text. ERP's are smaller than the EEG they are embedded with so they must be conducted several times to get the needed data. ERPs are recorded with patients conducting different tasks or having different stimulations to find out about cognitive processes. MEG is the magnetic counterpart of the EEG. Magnetic fields produce the currents used. "MEG is sensitive mainly to neuronal activity in the cortical valleys, or sulci; it is relatively insensitive to activity in gyri." According to the text, "EEG picks up voltage fluctuations from sources in both cortical gyri and sulci, although it tends to be more sensitive to the former." MEG is easier to localize because it is less sensitive to the scull and other tissues than the EEG. The signal is less complex than ERP according to the text.
Complementatrity
EEG/ERP and MEG
Why are analyses across multiple patients (such as that in Figure 2.2) more useful than analysis of a single patient?
Each person had a different range of brain damage. By overlapping their damage and comparing them, researchers could find where they were damaged in common. Then they can compare the symptomology and see if there is overlap. This takes the research from a case study to something more accurate.
EEG
Electroencephalogram (EEG) Action potentials → voltage changes - 1000's of action potentials → voltage change on scalp - μv (mico volts, 0.000001 volts) - not single action potential - skull is in the way and muscle contractions conflict it
ERP
Event-Related Potential (ERP) -high temporal but low spatial magnitude -specific event in EEG -completely noninvasive
Thinking back to the lexical decision experiment from Chapter 1 (copied below), describe how you could use two methods you just read about to understand where and when in the brain each cognitive process (e.g., perceptual processes, memory processes) is occurring. Describe how the two methods you chose complement each other. Lexical decision experiment: Imagine that your task is to sit at a computer and judge whether or not a string of letters presented on the screen is a word or not. If it is a word, then you should press the 'w' key. If it is a nonword, then you press the 'n' key. Create a cognitive model for this task by describing the sequence of cognitive processes that must occur to complete the task. Consider what kinds of perceptual, motor, memory, and other cognitive processes must occur.
I would use an fMRI with high spatial resolution and relatively low temporal resolution. I would also use MEG and EEG with high temporal resolution and pretty good spatial resolution. The fMRI with the MEG/EEG could allow me to see a more complete picture of the brain and its cognitive functions.
blocked design
In this design, the brain activity measured reflects neural activity integrated over an extended period (a block) while the subject performs a task or is presented with a particular stimulus condition. The resulting activation pattern can then be compared to the activity recorded during a different block of time in which the subject is not doing the task or is under a different condition. The need to integrate activity across time in this way thus substantially limits the specificity with which brain activation can be associated with particular cognitive processes.
1. Thinking back to the lexical decision experiment from Chapter 1 (copied below), describe how you could use one of the methods you just read about to understand where in the brain each cognitive process (e.g., perceptual processes, memory processes) is occurring. Lexical decision experiment: Imagine that your task is to sit at a computer and judge whether or not a string of letters presented on the screen is a word or not. If it is a word, then you should press the 'w' key. If it is a nonword, then you press the 'n' key. Create a cognitive model for this task by describing the sequence of cognitive processes that must occur to complete the task. Consider what kinds of perceptual, motor, memory, and other cognitive processes must occur.
In this hypothetical experiment, if I wanted to see if an area of the brain was involved in the lexical decision task, I might use transcranial magnetic stimulation (TMS) to disrupt that region while the subject was preforming the task. I would be able to see if overstimulating the region resulted the subject's worse performance or if certain stimulation improved function. I might try to use repetitive TMS specifically to "impair or improve performance." Since this is noninvasive and temporary, I think it would be a great way to conduct this experiment. Although it does cause discomfort and there is a risk for seizures, I would make sure to have informed consent from all participants.
optical imaging
Invasive - Hole opened in skull - Variety of techniques • Reflectance • Fluorescent dyes • Genetically engineered animals • Noninvasive - NIRS (Near Infrared Spectroscopy)
What are the strengths and limitations of studying brain function by examining brain damage (stroke, trauma, and disease)?
One major strength of studying brain function by examining brain damage is "that if damage to a brain area or system disrupts a cognitive function, it is likely that the damaged region is involved in some critical way in the performance of that function." A major limitation is "that the brain damage is the result of many factors that are not under the control of the experimenter." This means that there could be different parts damaged differently so it hard to draw conclusive evidence. Due to the fact that it would be unethical to intentionally damage a patients brain, these are case studies that occur naturally. This limits the generalizability because each case is not representative of the population.
What are the strengths and limitations of studying brain function using PET?
One strength of using PET to study the brain is that it allows one to "produce images that localize brain activity with high spatial resolution." Basically, we can use MRI scans with PET scans to find out "spatial information about specific brain areas involved in cognitive and other functions." A limitation of PET scans is that the "theoretical resolution of PET scanning to several millimeters, although true resolution is typically considerably lower because of spatial smoothing and other steps in the analysis." According to the text, a bigger limitation "is the time required to accumulate an adequate signal, which precludes its use in many cognitive paradigms." Of course, this means that temporal resolution is non-existent unless they use a blocked design. PET limitations in summation include "the short half-life of the reagents, the need for a nearby cyclotron to create the reagents, the use of radioactivity, and the very poor temporal resolution."
What are the strengths and limitations of studying brain function using fMRI?
One strength of using fMRI to study brain function is that it "offers better spatial localization than PET (several millimeters compared to approximately a centimeter in PET), as well as much better temporal resolution (a few seconds compared to a minute or more in PET)." It also does not require radioactive isotopes to map like PET. According to the text, its best advantage over PET in "the study of cognitive processes is its much higher temporal resolution." A limitation of fMRI is it also had a blocked design even though the blocks where shorter than PET, i.e. "around 20 to 30 seconds, as compared to a minute." Later, researchers found that they could have a event-related design. It is fast enough to be "applied in an event-related manner, at least when stimuli are presented at a moderately slow rate." According to the text, the most "significant shortcoming of fMRI (and indeed most other neural activity monitoring methods) is its difficulty in establishing causality in the interactions between brain areas —that is, whether activity in one brain area caused activity in another."
fMRI advantages or disadvantages
PROS -better than PET -noninvasive -high spatial -versatile CONS -low temporal -indirectly measures
What are the advantages and disadvantages of PET?
PROS -high in spatial -pretty good at seeing neural activity with pretty high correlation -Medical purposes about metabolic processes CONS -low in temporal -radioactive -indirect measure of neuronal activity -worse than fMRI in spatial, temporal, and invasiveness
Why are techniques like DTI which examine brain connectivity important for helping us to understand brain function?
Since axons tend to be hydrophobic, water "tends to diffuse more along the fiber tracts." This can be mapped to show how much water is flowing at what rate. This can be used to map "of structural connectivity in the brain." This is called the "connectome" like the genome. Some researchers want to map all the connections in the brain.
squids
Superconducting Quantum Interference Device immerged in the sea of helium too keep them cool why it is so expensive for MEG. Must be magnetically shielded to protect from earth magnetic field
BOLD response
The change in blood oxygenation that accompanies neural activity (BOLD = Blood Oxygen-Level Dependent)
Positron Emission Tomography (PET)
a method of brain imaging that assesses metabolic activity by using a radioactive substance injected into the bloodstream -positron and electron meet and kill each other and they release gamma rays -Radioactive isotope of oxygen - More complex molecules -Radioactively labeled glucose
agonist
a molecule that, by binding to a receptor site, stimulates a response
Aganoist
acts as a neurotransmitter
fast event-related design
additive response -can pull apart one every 4-6 secs
Why can MEG see sulci better?
can see sulci because the magnetic field exits skull
What is a weakness of brain lesions?
connectivity may be defined as function A -C-> B If C is knocked out, it could impede the function of B
Diaschisis
decreased activity of surviving neurons after damage to other neurons
event-related design
stimuli from two or more conditions are presented randomly or interleaved -slow or fast
Structural vs. Functional MRI
structural- anatomy functional- brain function
Neurons
• Action potential - Always the same amplitude voltage change • Never larger • Never smaller - All or none Neural coding - How is information coded into action potentials? frequency can change
Multi-voxel Pattern Analysis
• Can we guess at what people are seeing/thinking from brain data alone? -cubes don't work independently -train computers to recognize patterns -memory research and pared associates task with grand canyon and al gore
Multimodal Neuroimaging
• Multiple methodologies - Sometimes at the same time - Multimodal Neuroimaging • Current trend • Combine strengths from multiple methods - EEG/MEG/fMRI - EEG/fMRI - NIRS/fMRI
Clinical patients/Brain lesions
• Trauma - Stroke, disease • Lesions - Animals - damage specific portions • Pre/post tests - Determine deficit (hard to get pre-test -longest standing method -causal relationship -stroke caused widespread
