Bio psych notes - dfdf PDF

Preview

Summary

dfdf...

Description

P S YC 2 0 0 0 6

B I O LO G I C A L P S YC H O LO GY

SUBJECT GUIDE Overall Assessment Structure:

• Assignment 1 : Hurdle requirement. TMS Report will be completed in Assignment 2, Assignment 1 only requires the Methods and Results sections. Mainly for feedback purposes. Review Even though this assignment doesn’t contribute to the final grade, do make sure you put in your best effort because if you do well in these 2 sections, it will save you a lot of time for Assignment 2, which will require quite a lot of work for the Introduction. Make sure to submit as much as you can and treat it like a real graded assignment, because the feedback will also help you when you are finalising Assignment 2.

• Assignment 2 : 40%. TMS Report continued from Assignment 1, should be a complete report in APA Style. Review Overall quite a good assignment, though you will probably get annoyed with learning about TMS in such depth by the end of it. USE YOUR TUTOR! I had so many questions about my assignment, and even though tutors are not supposed to help you directly with the assignment, when you phrase questions in a specific way, you can kind of infer what exactly they want. Also do not underestimate how specific you’ll need to be in following the APA Style, as they’re quite specific about what they want.

• Final Exam : 60%. 2-hour multiple choice exam, with 15 minutes reading time. Consists of 120 multiple choice questions. Lecture, tutorial & required readings are all assessable content. Review Overall quite a breezy exam if you’re good at cramming. Questions are kind of weirdly specific with details like names of people who came up with certain theories, and details of experiments/studies discussed in lectures. Olivia Carter’s section on psychopharmacology is a good section to score high, because her questions are usually directly from her lectures. Neurogenetics is also really good because Patrick Goodbourn highlights all required information with a star at the corner of important slides, and questions are straightforward. Make sure you know the different genes involved with the different diseases mentioned.

PS YC 20006

2

TRANSCRANIAL MAGNETIC STIMUL ATION Overview of TMS TMS is a non-invasive method of creating visual cortical lesions — small, temporary, and localized —> allows for better understanding the function of specific brain regions where TMS is applied. The TMS disrupts the signal of the true-firing, resulting in decreased function. We can observe the effect of doing so by assessing performance in function. If function is affected negatively, we infer that the interference has caused an effect, and thus the area is involved with such function. Why not patient brain resection/lesion study? : insufficient number of live patients to learn about all the parts of the brain, rarely find lesions in a single specialized area, recovery may compensate for lesions & affect observations. Mechanism Uses a “figure-of-eight” coil on the scalp which produces a rapidly changing magnetic field to induce currents in the brain directly below. Current depolarizes & thus masks the true firing from neurons, and will affect the associated function. TMS coil provides a more focal area, and can therefore make more accurate inference about which part of the cortex has been affected. Types of TMS include: • Neural-noise approach • Virtual lesion approach • Probing-excitability approach • Paired-pulse approach NEURA L -N OISE A P P ROA CH Good to determine causality. Use a single-pulse TMS at the time window which the region is required, to disrupt cognitive processing — injecting neural noise. (This is the approach used in the assignment). Critical period for neural-noise of visual stimuli is 40-120ms. Note: Regions don’t stop working, TMS just interferes with the functioning by disrupting the signal transmission process. VIRTUA L L ESION A P P ROAC H Using repetitive TMS (rTMS) to interrupt or enhance cognitive processing. This provides interference for a longer period of time — Able to measure whether the task is impaired by the longer interference, & for how long — usually reduced / slowed function. Note: There are strict guideline for rTMS (don’t need to know specifics)

PS YC 20006

3

P R O B I N G E X C I TA B I L I T Y A P P R OA C H Tests how excitable / responsive the motor cortex is during the cognitive task. If it’s involved with the cognitive function, should already be activated when the single-pulse TMS is applied. Doesn’t actually disrupt the performance of cognitive function. Measure motor-evoked potentials (MEPs) — electrical activity of the muscles — and compare them between different conditions. Higher MEP = more excitable during the task ; Suggests region might already be activated for the task & is involved in the cognitive process. BUT this approach is susceptible to spillover effects (activation of nearby areas excites motor cortex). PA IRED -P UL SE A P P ROAC H 2 pulses delivered with a short interval (1 sub-threshold, 1 supra-threshold). Looking at how strongly the first pulse influences the effect of the second. Example: Schizophrenia — abnormalities in motor cortex inhibition suggested ; evidence of reduced cortical silence period (CSP) (suppression of tonic motor activity after descending excitatory activity). Compared to control, increased response to the 2nd pulse in both medicated and nonmedicated participants — indicates a general deficit in motor inhibition. Clinical Applications of TMS Note: In Victoria, TMS is an endorsed treatment option for depression (imbalance of prefrontal cortex activity between hemispheres) ; Treatment of last resort only.

Neural Noise Approach

Virtual Lesion Approach

Probing Excitability Approach

Paired-pulse Approach

TMS type

Single-pulse

rTMS

Single-pulse

2 pulses in brief succession

Measurement

Effect of TMS on task performance

Effect of longer-lasting TMS on task performance

Responsiveness of motor cortex to TMS (MEP measurement)

How strongly first pulse affects second pulse

Causality?

✓

✗

✗

✗

Affect function?

✓

✓

✗

✗

PS YC 20006

4

STATI STI C AL HYPOTHESIS TESTI NG Overview! We need to ensure the observed difference in a study is significant enough for us to make an inference — not just due to chance or measurement error. Statistical testing shows us the likelihood that our results are by chance vs due to a real difference, by comparing results to a probability distribution representing chance!

H0 (Null Hypothesis): Our empirical result was a result of chance. If it’s very unlikely, p < 0.05, then we reject the H0 . BUT there is always still a probability that we’re wrong, depending on the ⍺ level set. Note: we assume H0 is true until proven otherwise. Note: we don’t use the z-distribution (normal) as that requires knowing population μ (mean) & sd. T-DISTRIBUTION t - distribution is dependent on degrees of freedom (df = n-1) — as df increases, distribution approaches normal distribution as it’s more representative of the population. t-distribution takes into account expected mean (M) and standard error of the mean ( SM).

t=

M−μ SM

ONE-SAMPLE DESIGN def. A group of people with results from different people; compare results to a single value.

t=

M−μ , where (M - μ) is the mean diff. & SM is the estimated SE of the sample mean SM

SM =

s n

SS , where s 2 = , since we don’t know the population s, & SS = (ΣXi − M )2 n

where

Example:

SM = 20

18

12

14

M = 16, SS = 40 13.333 4

= 1.826

s= t =

40 = 13.333 3

16 − 10 = 3.286 1.826

3.286 is our empirical t-value. Now, we compare to critical t-value in t-test tables with ⍺=0.05. It empirical(t) > critical (t), then we reject the H0 , because there is significant evidence to show an effect. Note: Directional Hypothesis — look at ⍺=0.05 Non-directional, 2-tailed test — look at α=0.025 (more conservative)

PS YC 20006

5

Can compare group data to known values

May not know population values Doesn’t record over time Cannot make comparisons

INDEPENDENT MEASURES DESIGN def. 2 groups of diff. exposures (e.g. control & experimental); compare values

t=

(MA − MB ) − (μA − μB ) , where MA − MB is diff. in group mean, μA − μB is H0 S(MA−MB)

SMA−MB =

Sp2 =

s p2

+

nA

SSA + SSB d fA + d fB

sp2 nB

, where the S(mean difference) is the sqrt of the pooled variance

, where the pooled variance is an average of the sample variances.

Note: this only works when the 2 groups have the same sample size, n. Group A:

sp2 = 648

540

689

19754 + 14418 = 5695.33 3+3

523

Group B:

t =

600 − 500 5695.33 4

552

449

567

+

= 1.873

5695.33 4

432

Measurements are independent

Confounding may impact results, needs a large sample size Cannot make inferences over time

REPEATED-MEASURES DESIGN def. Single group which provides data for both conditions, a.k.a. paired samples

t=

MD − μD

SMD =

sMD SD n

, where μD=0 under H0, and standard error of the mean difference is used.

, where standard error of the mean is used, and SSD =

PS YC 20006

(Di − MD )2 Σ

6

MD = 100

Example:

Condition A: Condition B: Difference (D):

648

540

552

449

96

91

689 567

122

SSD = 662

662 = 14.855 3 14.855 = 7.427 SMD = 2 100 = 13.464 t= 7.427

523

SD =

432 91

No baseline/confounding effects

Not independent, variance different

Can study changes over time

Individuals not naive in 2nd round

Need fewer subjects for test EFFECT SIZE MEASURES Cohen’s d def. Gives an estimate of the effect size, independent of the sample size, n. Independent measures:

d=

MA − MB sp2

Repeated measures:

d=

MD

d = 0.2 (small effect) d = 0.5 (medium effect) d = 0.8 (large effect)

SD2 Percentage of variation explained my experimental manipulation (r 2)

t2 2 r = 2 s + df

r 2 = 0.01 (small effect) r 2 = 0.09 (medium effect) r 2 = 0.25 (large effect)

Note: r 2 measurement is not independent of the sample size!

TIP: Not a statistics subject, so not necessary to learn all of these formulae for the final exam, but make sure to understand the concept of each terminology & study design. Good to know numbers that define effect size of Cohen’s d and r 2

PS YC 20006

7

ELECTROENCEPHALOGRAPHY (EEG) Overview Electroencephalography is a non-invasive method of detecting neural activity with sensitive electrodes and conducting gel. Electrodes detect small fluctuations of electrical signals from cortical neurons. Raw signal needs to be adjusted (lots of noise) before we can analyze. It allows us to learn about cognition when people perform tasks. Note: Most accurate way of measuring activity uses intra-cranial EEG to directly measure at the exposed cortex, but this is not practical. METHODS Recording Reference electrode placed on the nose/behind ears. Conductor gel applied to optimize the conduction & signal detection. Then, participant will wear an electrode cap — each electrode is for a specific cortical region, so signal can be traced easily to the correct region of the brain. Letters of electrode encode for the brain region that it’s observing: • F = Frontal

• P = Parietal

• O = Occipital

• T = Temporal

• C = Central

Neurophysiology of EEG The EEG signals recorded are post-synaptic potentials — voltage that rise when neurotransmitters bind to receptors on post-synaptic neurons. This voltage is what causes the ion channels in the postsynaptic neuron to open / close & lead to the change in potential that transmits the signal through the axon etc. Neurons are charged, acting like a small dipole — there is current flow, but we cannot pick up charges from individual neurons, only when they’re in a cluster (>10,000 simultaneous activated neurons) and are all facing the same orientation, and are in a gyrus (Sulci neuron activity cannot be picked up by EEG). Signal Signals are measured in relation to a reference electrode. Amplitude: 10 μV - 100μV (very small, needs amplification) Frequency: 256 - 104 Hz Limitations • EEG is biased to pickup information from cerebral cortex gyri, but not from the sulci • Meninges, cerebrospinal fluid, & thickness of the skull smears the EEG signal & makes it harder to identify the specific source of the signal. “Inverse problem” : one configuration of

PS YC 20006

8

signals can have multiple dipole solutions — always more than 1 solution, because more than 1 spot of the brain being activated at the same time. • Activation of multiple regions at the same time also causes noise in our results — requires amplification and processing. PROCESSING The raw signal we receive has a lot of noise. Processing the signals is about identifying the dominant signal, making sense of the regularity in the noise — identify the source, relate them to cognition, & isolate the correct signal — done through mathematical algorithms! There are many signals but we only need to remember 2: • Artefact rejection: Finding all alternative sources of non-brain signals (artefacts) e.g. sweating, Electrical noise (EEG machine) • Ocular correction: Removing the signal from blinking. Complex muscles in the eye, when blink or moving, the eye will emit strong signals. Averaging: Underlying structure of our result is true, but random signal fluctuations produce noise & distort the signal. Repetition removes the random confounding effects. So, each participant repeats many trials & averaging them produces a more meaningful result. EVENT-REL ATED POTENTIALS (ERPS) def. A method of observing the difference in potentials when different stimuli are presented. This allows us to investigate the brain activity related to the cognitive process. Measurement methods: • Peak-Amplitude: Looks at the amplitude of peaks & comparing them • Area under the curve • Peak-to-peak: distance between 2 peaks of each signal Note: results may differ between measures, so we use the method which can find something that always happens then the task is executed (consistency) Reading the signals: In EEG signals, peaks can be named in many ways (time response recorded, location, stimuli, etc.) but most common is the polarity of the signal (remember this is caused by the polarity of the neurons observed). KEY: Negative signal = Upwards ; Positive signal = Downwards Interpreting signals is tricky and requires experience & lots of participants • Signal depends on how you look at it — if you consider red dip to be significant, the initiation occurs at the same time. If considered as noise, then blue initiates first.

PS YC 20006

9

• If dip is noise, then it would cancel out through repetition with other participants RESEARCH APPLIC ATIONS

1. Virtual Search Experiment (Woodman & Luck, 1999) Although hard to interpret, many ERPs are well-studied, so we can use an existing one, analyze the effects of altering the conditions on the known ERP & its associated cognitive function. Virtual Search Experiment used N2pc (second negativity component found on posterior contralateral) which is related to index attention. Discovered that the potential was greatest at the posterior cortex, contralateral (opposite) to where observer is attending.Participants had to search for a coloured square which was open to the left (i.e. blue square) • Participants had to search for a coloured square which was open to the left (i.e. blue square) • To ensure that they attend to one hemifield first, they manipulated the probability that a specific colour was the target • C75 – colour with 75% probability that is was the target • C25 – 25% probability • This prompted participants to search for the C75 first • They also tested what happens when there is no target present • If people search is serial, then attention should switch from one hemifield to the other and this should be demonstrated by N2pc difference • If people search in parallel, then there shouldn’t be a different in N2pc between the 2 hemispheres Conclusion: Participants use serial search, as the N2pc was observed when switching of hemispheres was required

2. ERN: Error-related negativity (Gehring et al., 1993) def. ERN is a negative deflection of up to 10μV in amplitude, observed 80-100ms after erroneous response (Flanker-task: identify presence of error in letter sequence) • ERN was strongest when people emphasized accuracy & weakest for speed • The greater the ERN, the higher chance of getting it right the next time • The greater the ERN, the slower the response next time

PS YC 20006

10

METHODS: FMRI Functional Magnetic Resonance Imaging METHOD When you apply a cognitive process, neurons in the brain responsible for this function would fire. This firing can be measured with a BOLD signal (Blood Oxygen-Level Dependent) produced by the fMRI. The fMRI also produces images of the brain that are active. We can make a reverse inference (making conclusions about the cognitive processes based on the presence of an activation). The large chamber produces a very strong magnetic field of about 3 T (Range: 1.5 - 9 Tesla). The machine can be used for other parts, not just brain imaging. Magnetic field is very strong, so the room must be metal-free. A head coil is used to: • Fixed position of the head to avoid inaccurate imaging • Send radio frequency pulses into the brain • Receive brain signals produced Note: Radio frequency pulses are not felt because we are exposed to them everyday Basic Physics of MRI Our brains compose of >70% water, with lots of Hydrogen atoms. H+ protons are constantly precessing about an axis in random directions. The induced strong magnetic field causes the H+ protons to be aligned parallel along the Z-axis (towards back of the machine) & the precessing frequency is dependent on the magnetic field applied. However, the precessing is still in a random fashion, as they are at different points of the circulation. A radio frequency pulse is applied perpendicular to the magnetic field by the head coil at the same frequency of the precession. This causes the protons to absorb energy and this brings about 2 main effects: • tilts the net magnetization vector to the transversal plane (creates a vector in the y direction) • aligns precession of the spins (rotations are synchronized) Note: when RF switched on, some H+ move to higher energy state (anti-parallel) The image is produced when the RF pulse is switched off. This causes the H+ to revert to their low energy (parallel) state before the pulse — relaxation. Excess energy is emitted in the process. Transversal magnetization decays with differing speeds depending on the density of protons (water) in the tissue — relaxation occurs at different speed. RULE:

PS YC 20006

11

• White regions = active signal, region is less water-dense • Dark regions = decayed signal, region is more water-de...