Neuroscience Methods 2

Behaviour is a “window” into which cognitive processes?

Behaviour is a “window” into which cognitive processes?

• perception

• learning

• memory

• problem solving

• language

History of cognitive psychology

• introspection: “thinking aloud”

• behaviourism

• rise of cognitive psychology via the computer era

• psychophysics: performance measures

  • reaction time

  • errors

  • threshold

Introspection

• the examination of one’s own conscious thoughts and feelings

• very subjective

• we are not always aware of things that happen (e.g. anosognosia for hemiplegia → paralyzed movement of the limbs)

• Nisbett & Wilson (1977): people do not really “introspect”, they tend to make up a plausible story based on their ideas about the world

Behaviourism

• Watson (1878-1958)

• we can only objectively report behaviour and environment

• forget about the mind: behaviour is essentially reflexive

• brain is a stimulus response device

• what happens within the brain = black box

  • input coming in via the senses + output

  • we don’t know what happens inside the black box (not going to try and understand)

• systematically study what goes in + what comes out (input & output)

Classical conditioning

1. before conditioning: unconditioned stimulus → unconditioned response

2. before conditioning: neutral stimulus → no conditioned response

3. during conditioning

4. after conditioning: conditioned stimulus → conditioned response

   Example: dog associating/learn relationship between bell and food → dog salivating when bell is rung

   • anticipated food

   • response can extinguish if not reward (i.e. food) is given

   • dog unlearns behaviour

Computers

• comparing computers and the brain

  • both have input and output

  • computers have wires processing input → turning into output

• “look” inside black box

Cognitive Psychology

• create hypotheses of what might be happening with different predictions → inside black box

• observe what happens in different conditions and draw conclusions

Representations

• some internal representation of the outside world needed to process & act upon the world

• representation: behaviour is not based on the “physical” world, but on representations of the world, as generated by the brain

• multiple representations: representations exist at various levels of abstraction, have various purpose/function, and are localized at various places → colours

• inference of representations

• classic paradigm: Stroop (1935)

• stroop test: name colour of letters in 3 different conditions

Stroop test

• name the colour when the text of the colour = colour it represents (RED) = faster

• third test: conflict between two representations of colour

  • representation you viewed with your eyes

  • linguistic representation of the word

    1. representations: colour, word

    2. computation: convert to appropriate action → conflict?

    3. chronometry (Reaction Time Subtraction): ‘Inference’ = Incongruent - Neural RT

       → time tells us something about internal processes

Psychophysics

‘psycho” (of the mind) + “physics” (natural laws)

• uncovering lawful relationships between physical stimulus and their resulting percept

• measuring relation between physical parameters of a stimulus properties and the psychological percept

• physical properties of stimulus → brain → behaviour

Response time → first to measure RT: Wundt & Helmholtz

• studied relationship between intensity of light and how quickly someone could respond

• constancy in responses

• vary parameters

• distribution of response times → median is considered

  • respond faster to sounds or light

  • can deduce differences - maybe the differences are systematically varying with the stimulus intensity or with the stimulus type

• if you combine sensory information = much faster response time (multisensory response enhancement)

• multisensory neurons: integration of A and V

• responds to more than one sense → neural activity of said neuron = stronger

Detection threshold

• how much signal strength do we need to detect a stimulus XX% of the time?

  • can be determined

  • can you see the stimulus (circle)? → different frequencies

  • results depicted: cumulative distributive function or psychometric curve

    • proportion of detection responses over stimulus contrast → S shaped

    • threshold = 50% point (50% of the time stimulus detected)

• studies manipulated findings, also including spatial attention → threshold changes

  • attention modulates sensitivity to visual stimuli be lowering the threshold at which you perceive something!

How to calculate another psychometric function

• horizontal axis: intensity difference between the two stimuli

• vertical axis: proportion of different responses (how often you say they (the circles) are different

• calculate the “Just Noticeable Difference” (JND)

  • difference in intensity corresponding to the 50% and 75% point or the 50% and 25% point

• seems to be some lawful relationships between stimulus properties and perception!!

Illusions

• explain how information is represented in the brain and what kind of rules/principles our perceptual system uses to perceive the world around us

  • example: lines look like different length depending on their arrows

  • possible explanation → depending on the context in which we perceive vertical lines one might look longer that the other because of perspective

• things that are father are often smaller → in terms of retina

Ebbinghaus illusion/Titchener circles

• used to make a distinction between action and perception pathways

• participants asked which orange circle was bigger and grab the stimulus

  • the action pathways used the correct information to open the size of the fingers

  • the perception pathways fooled us into thinking one orange circle was larger than the other

Illusion definition + connection to representations

• illusions: an internal representation that does not accurately reflect the world and the physical properties of stimuli in the world

• illusion is informative about the nature of that representation and how it is constructed

  • adaptation of colour

    • receptors sensitive to different wavelengths

    • long stimulus of channel reduces sensitivity causing the complimentary colour to be perceived

Representations part 2

• the representations in our brain can be very different than expected from the outside world (stimulus)

• we might have an idea about what a representation should look like based on the stimulus properties, but maybe the brain represents this information in a very different way

• we can describe these (lawful) relations but we also want to know how the brain operates to create these representations

Mathematical models

• utilized to try and understand what occurs with the brain & how one perceives the environment

• example: ventriloquist effect

• sound in localized towards a visual source (mouth of the puppet)

• effect described by mathematical models

Bayesian cue integration

• accurately predicts performance

• if our brain uses the reliability of the sensory input to also weigh how important that input should be in perceiving the location of a certain stimulus → should follow a mathematical rule

• mathematical formula describes the weight by using variability in responses to a certain stimulus to estimate the weight

• variability of responses of two stimuli = the same → equally precise and reliable

Optimal cue integration

• distribution to sound are much wider → less sure of where the sound is coming from

• sure of localization of stimulus = small distribution

• difference in variability between the two distributions (visual and auditory) reflects something about the reliability of your perception & responses to stimuli

• mathematical formula would based on this variability assign higher weight to visual input

Neural respresentations

• neurons respond to specific stimuli

• neuron can be tuned to specific orientation

Visual spatial representations

• brain maps of space

• receptive fields are organized neatly into maps

• map spatial relations are similar to relations on retina (which has a direct relation with the visual field)

Auditory spatial representations

• frequencies (high and low) organized neatly into maps in the auditory cortex

  • something similar can be found for the spatial metrics obtained via hearing

• spatial localizations of sound are also neatly organized

  • spatial localizations of sound are transformed into the spatial reference frame of the eyes into the brainstem

    • make eye movements towards sound locations

Another way to learn about behaviour

• lesions in animals (brain damaging animals)

• either irreversible (destroyed or removed) or reversible (e.g. coding)

• conceptual advantage of lesions over recording: investigate causal relation

• causality between the functional integrity of a certain area and the behaviour of interest

Lesions in humans

• infer whether damage to a certain area of the brain leads to certain behavioural functions

• example: brain damage to certain area → affect attentional processing → causes people to ignore information that they do perceive → not in their visual system (do not register visual information in a certain part of their visual field)

  • neglect

EEG → Electroencephalography

• less invasive that cortical cooling with cats or neuro recordings in animals

• measure electrical brain activity using EEG’s in humans by picking up the tiny neural activity with someone’s brain on the outside of the skull

• measures large scale activity using electrodes placed on the head

  • activity can be amplified

• whole brain is active

  • what you measure on the outside (EEG) is a sum of all the activity that is occurring on the inside

  • most likely, the neurons close to the recording site contributes most to the observed measurements

    • depending on the orientation of the neuron some activity might not be picked up → limitations of what is measured

EEG → great what?

• great temporal sensitivity

  • measure every millisecond

• raw data: see neural ongoing data, can only observe some patterns → like alpha oscillations (certain waveforms that occurs when you close your eyes/are tired)

• more advanced analysis: look at frequency of raw data or even related potentials

  • very specific characteristic electrical brain responses to a certain stimulus

What does EEG measure

• not action potentials

• not summation of action potentials

• summation of graded post synaptic potentials (PSPs)!!

• event related potentials: average brain responses within certain amout of milliseconds after presenting a stimulus to a certain participant/patient

Limitations of EEG

• electrodes are near the scalp → lots between neural activity and what the electrode is measuring

• where is my signal coming from?

• conductance, blurring, signal loss with depth all decrease with spatial resolution

• we want 3D answer from 2D data

Source localization

• try and calculate where a certain signal is coming from

• forward problem: modelling (relatively straightforward)

• inverse problem: estimation of the model parameters (unconstrained)

The signal (EEG)

• raw EEG: difficult to interpret, but some clear patterns

• Event related potentials

• small changes in electrical activity in response to stimuli

• this specific brain activity comes on top of ongoing background EEG (relatively large amplitudes)

• rhythms in background EEG are not necessarily correlated with stimulus onsets, but with stimulus evoked responses are

• by averaging across a large number of trails the background EEG averages out and the ERP (remains??)

• random activity (not related to stimulus presentation) becomes closer to 0

• start seeing characteristic response to that stimulus

• certain stimulus represented from raw data (don’t see anything) → average enough times → filter out “noise” and get characteristic of response to a stimulus

Event related potentials

• see whether certain manipulations affect the electrical response

• attending a certain stimulus alters the amplitude of P1 (first positive peak) to a visual stimulus

Strength and limitations of EEG’s and ERP’s

strength: high temporal resolution

weakness: source difficult to determine

• what is the time of interhemispheric transfer? → EEG = best method

  • sensitive to temporal resolution

  • amplitude

  • latency