Transcript Pt2Localization - MemoryAndCognition

Memory and Cognition
PSY 324
Topic 2: Cognition and the Brain
Dr. Ellen Campana
Arizona State University
A Brain



Gray & White Matter
Solid tissue
Made up of neurons

Golgi showed by
staining slices with dye
Neurons

Similarities with other cells of the body
Have a nucleus containing DNA
 Surrounded by a cell membrane
 Contain mitochondria and other organelles
 Do basic cell stuff (protein synthesis, energy
production)


Unique characteristics
Do not reproduce
 Structure, function, chemicals (details to come)

Structure of a Neuron
Structure of a Neuron
Varieties of neurons

Function = transmit information to other cells
Sensory / Afferent neurons: info TOWARD CNS
 Motor / Efferent neurons: info AWAY from CNS
 Interneurons: info to other neurons in the CNS

Info Transmission: Simple Story




Neurons are transducers – convert environmental
energy to electrical energy (starting with receptors)
Energy is propagated from the dendrites into the cell
body.
Energy is propagated to the end of the axon. When it
goes above a threshold, it triggers the release of
neurotransmitters into the synapse
The neurotransmitters in the synapse trigger the same
process (or a different one) in the next cell
Synapses
Info Transmission: Deeper Story

Background concepts from physics
Matter composed of molecules (always moving)
 Molecules can have +/- charge
 Like charges repel, opposite charges attract

+
-
+
-
Info Transmission: Deeper Story
Each neuron has a resting potential, the voltage
difference across the cell membrane, caused by
the chemicals inside/outside the cell when the
cell is not firing
+
+
+
+
+
+
+
+
- +
+
+
+
+
+
+
+

Info Transmission: Deeper Story

Axon lined with ion channels (sodium channels,
potassium channels) that open and close during
an action potential to propagate the signal
Depolarization phase: Sodium (Na+) Channels
 Repolarization phase: Potassium (K+) Channels

Depolarization Phase
Depolarization + Repolarization
Info Transmission: Deeper Story

Action potential moves down the axon, as gates
open and close in sequence
Synapses

Action Potential reaches the end of the axon,
triggering release of neurotransmitters
•Excitatory neurotransmitters
increase firing rate in next neuron
•Inhibitory neurotransmitters
decrease firing rate in next neuron
•NOTE: Other neurotransmitters do
other things (less well understood,
less relevant to cognition, especially
to models we will talk about)
Method: Single-Cell Recording



It is possible to record activity of a single cell
Tiny wires (called microelectrodes) stuck into
axon, attached to oscilloscope for data display
Time is a factor
De/Repolarization cycle = 1/1000 S or 1 ms
 Activities of cognition take at least 100ms – at that
resolution action potentials show up as spikes
 Often most useful to talk about firing rate

Method: Single-Cell Recording

Pictures of “spikes” –


http://viperlib.york.ac.uk/ , keyword single cell
recording
Video clip from Hubel & Weisel (they got the
1981 Nobel Prize in physiology and medicine
for this work)

http://viperlib.york.ac.uk/ , keyword single-cell
recording
History of Single-Cell Recording

Participants in experiments:
1880s – People injured by accident with exposed
brains, also patients with epilepsy
 1950s – Fully anesthetized animals (cats, squirrels,
monkeys, apes)
 1980s – Awake, active monkeys and apes



Invasive, destructive procedure
Data: there is a cell in the [animal] that increases
firing under [conditions]
Value of Single-Cell Recording

By itself, the data doesn’t tell us much


Can find cells that do almost anything
The value of single-cell recording for
understanding human cognition depends on:
Functional organization of the brain
 Consistencies of organization within species
 Meaningful mapping from animal models to
organization of human brain


Fortunately, much evidence that these exist
Clarification from last time

Question came up about diffusion and
connection to neuron behavior
Clarification from last time

Background concepts from physics
Matter composed of molecules (always moving)
 Molecules can have +/- charge
 Like charges repel, opposite charges attract

+
-
+
-
Clarification from last time

Background concepts from physics




Matter composed of molecules (always moving)
Molecules can have +/- charge
Like charges repel, opposite charges attract
Cell membrane maintains an imbalance





At rest, negative inside and positive outside (pumps maintain)
Ion channels open / close quickly
Particles rush in / out (like a hole in a boat)
Sodium and Potassium are +, but other chemicals create the
negativity inside the axon at rest
Focus on Sodium and Potassium because of the gates
Clarification from last time
+
-
+
+
-
+
+
+
+
+
- - - - +
+
+
+
+
+ +
+
Axon
Cell membrane (imbalance)
Diffusion and charge drive the
process – cells rush in/out
There are mechanical pumps
but they just restore resting
potential at the end
Clarification from last time

Things to note
All action potentials are the same size, in terms of
voltage (all-or-nothing principle)
 Most useful to think of them as on / off, or to think
about firing rates (spikes per second)

These are from a specific study.
Neuron A responds when the
stimulus is ON. Neuron B
responds when the stimulus is
OFF. Neuron C responds to
changes in the stimulus.
Why study neurons?

Everything we see, hear, do, smell, remember,
taste, touch, pay attention to, and think about is
represented physiologically by neurons firing
All sensations, perceptions and thoughts are neural
activation
 All of our actions arise from neural signals


Study of cognition is about both physiological
and functional models

Increasingly uses brain imaging and neuroscience
methods (later today)
Brain Organization

Hierarchical Structure
Smallest unit: Neuron
 Neurons form Circuits (many levels)



Convergence, Inhibition, Excitation
Related circuits contribute to localized function
Brain areas for different functions
 Hemispheric specialization

Neurons as part of circuits

Neural processing occurs when neurons
synapse together to form a neural circuit
Convergence
 Interaction of excitation and inhibition

Neurons as part of circuits

Neural processing occurs when neurons
synapse together to form a neural circuit
Convergence
 Interaction of excitation and inhibition

Hubel & Weisel

Single-cell recording of feature detectors

Simple neurons (from the video last time)


Complex neurons


Orientation (thickness, location of line)
Orientation, direction of motion
End-stopped

Length, direction of motion
Feature Detectors



Lines (shapes and orientations)
Directed Motion
Complex Stimuli
Geometrical figures
 Common objects in the environment (houses, manmade objects, birds)
 Faces


Depend on selectivity – neurons firing at some
times and not at others
Neural Codes in Daily Life


Consider the case of recognizing the face of a
specific person – how could that happen?
Hypothesis 1: specificity tuning – a particular
neuron could selectively fire when you see that
person
Specificity Coding

Difficulties with Specificity Coding Hypothesis




Related idea: Grandmother cell (coined by Lettvin)



Too many different faces, concepts, etc. to have a neuron for
each one
Depends on experience – would have to learn each face
(because neurons don’t reproduce)
Neurons selective for faces are active for many different faces
Cell responds to image of a grandmother, general concept of
grandmothers, your own grandmother
Some evidence that these might exist in Hippocampus –
associated with memory storage, not vision
For recognition (and many other types of cognition),
specificity coding is not enough
Neural Codes in Daily Life

Hypothesis 2: Distibuted Coding – code for a
specific face is distributed across a set of neurons
Distributed Coding

Advantages
Efficient -- firing of fewer neurons can represent
many more different stimuli
 Similar items can have similar neural codes



Helps with learning
Graceful degradation -- if one or two neurons do
not fire, it is still possible to recognize a face
Reconciling types of coding

Evidence for Specificity Coding
Feature detectors
 Concept cells in hippocampus (memory area)


Argument for Distributed Coding in recognition
Clear theoretical advantages in recognition
 Will see a lot of evidence later


Both are happening in the brain – in different
areas at the same time (parallel processing)

Pattern across (+ interaction btwn) areas=cognition
The Whole Brain


Localization of function - Different parts of
the brain serve different functions
Many, many ways to divide the brain
Like an onion, many layers
 Like a fractal, the closer you look the more complex
it seems


Descriptions may seem contradictory and/or
overlapping because of this
Cerebral Cortex

Most important for Cognition
Cerebral Cortex

Temporal Lobe
Language
 Memory
 Hearing
 Perceiving forms


Occipital Lobe

Visual information (early processing) – feature
detectors
Cerebral Cortex

Parietal Lobe
Touch
 Vision
 Attention


Frontal Lobe
Proportionately larger in humans than in other
species
 Language
 Thought
 Memory
 Motor functioning

Subcortical Structures
Subcortical Structures

Hippocampus


Amygdala


Forming memories
Emotions, emotional memories
Thalamus

Processing sensory information (vision, hearing,
touch)
Hemispheres


Brain separated into sides (hemispheres)
Corpus Collosum connects them


Structurally and
functionally very similar
Lateralization – specific
functions occurring in one
hemisphere or the other
Note: Sperry studied “split-brain patients”, who had had their corpus collosum severed as a
treatment for epilepsy. He shared Hubel & Weisel’s Nobel Prize for this work.
Lateralization



Vision of left part of the world lateralized to the
right side (opposite also true)
Motor Control of left side of body lateralized to
the right side (opposite also true)
Touch on left side of body lateralized to the
right side (opposite also true)
Lateralization

Are there “right-brained” and “left-brained”
people?
Analytical/Logical processing (syntax of language)
usually on the left side (not always)
 Analogy and Broad Thinking usually on the right
side (not always)



Everyone has (and uses) both
Patients who have had a hemispherectomy


Other side usually takes over missing functionality
Coglab “Brain Assymetry”
Localized Function

Parietal Lobe
Sensory Homunculus (near the front,
somatosensory cortex)
 Motor Homunculus (near the back, motor cortex)

These guys aren’t actually
IN your brain, they’re
representations of how
much cortex area is devoted
to different body parts
Sensory Homunculus
Each side of the
brain has a copy,
which processes
touch from the
other side
Localized Function

Parietal Lobe
Sensory Homunculus (near the front,
somatosensory cortex)
 Motor Homunculus (near the back, motor cortex)


Temporal Lobe
Wernicke’s Area – metaphor, meaning in language
 Broca’s Area – logical structure of language
 Fusiform Face Area (FFA) – specialized for faces
(or is it things we’re experts at recognizing???)

Researching Localized Function

Neuropsychology – comparing patients with
localized brain damage

Single dissociation– single patient has some things
impaired, other things not impaired
Single Dissociation: Phineas Gage




Construction accident –
1848
Harlow (doctor) wrote a
lot about his condition
Gage lived, could talk, act, and do all “normal”
activities, but suffered impairment of
emotional, social, and personal traits
Evidence for some separation of language and
social traits, etc.
Researching Localized Function

Neuropsychology – comparing patients with
localized brain damage

Single dissociation – single patient has some things
impaired, other things not impaired


Alice: Short-Term Memory OK, Long-Term Memory
impaired (like in Memento)
Double dissociation -- two (or more) patients
show opposite single impairments

Bert: Long-Term Memory OK, Short-Term Memory
impaired
Double Dissociation
Alice (temporal
lobe damage)
Short-term
memory
Long-term
memory
OK
Impaired
Bert (frontal lobe Impaired
damage
OK
Double Dissociation
Naming Living
Things
Naming
Nonliving Things
Group 1
OK
(damage to area 1)
Impaired
Group 2
Impaired
(damage to area 2)
OK
What can we conclude?

Double dissociation
Two functions involve different mechanisms
 Two functions involve different brain areas
 Mechanisms are independent


Single dissociation
Two functions involve different mechanisms
 Two functions involve different brain areas
 Mechanisms may not be independent

Limitations of Neuropsychology

At least for human processing, brain damage
comes about from natural means (accident, etc.)
Members of groups rarely have exactly the same
damage (location or extent)
 No record of processing or brain organization
before the damage
 Difficult to assess all possible types of functional
impairment


Damage may cause reorganization (plasticity)
Imaging Methods






EEG- Electrodes on outside of head continuously
measure electrical activity
PET- Radioactive dye injected, accumulates in different
regions over time and can be read by a scanner. Essentially
measures metabolism of neurons
fMRI- Brief magnetic pulses used to give a snapshot of
ratio of oxygenated to deoxygenated blood (metabolism)
TMS- New measure. Magnetic field can disable specific
portions of the brain for a short time, simulating damage.
Temporal resolution: Detail with respect to time
Spatial Resolution: Detail with respect to physiology
Image from an
fMRI scan
Image from
a PET scan
Imaging Methods
EEG
Spatial Resolution Temporal
Resolution
Poor
Good
PET
Excellent
Poor
fMRI
Good
Good
TMS
Good
Good
Subtraction Technique



Used for fMRI studies
Method similar to
Donder’s study
Compared two situations
that included different
cognitive processes
Data = blood glucose level
 Relative measure
Visual Stimulus
(light flashing)
Visual Stimulus
(light flashing)
Perception
of the light
Perception
of the light
Response
DECISION

Response
Subtraction Technique

Activation in
control condition
is subtracted from
experimental
condition to get
activity due to
stimulation in the
experimental
condition
Effects of experience

Experience-dependant plasticity

Developmental environment can affect neuron
specialization


Kittens raised in environment with only vertical lines had
more of their brain devoted to recognizing vertical lines in
adulthood (and none devoted to horizontal)
Learning happens through changes in connections
and relationships between neurons, even in
adulthood

Greebles study (back to the FFA)
Greebles

Recall discussion of localization of function
Fusiform Face Area (FFA) was an area in the
Temporal lobe devoted to recognizing faces… or
was it things we’re experts at recognizing???
 Kanwisher has demonstrated, using fMRI, that the
area does selectively respond to faces
 Gauthier and colleagues showed fMRI evidence for
experience-based plasticity in this area (Greebles
study)

Greebles Study




Step 1: measure brain activity in
FFA when viewing Greebles
Step 2: train people to
recognize individual Greebles
and families of Greebles
Step 3: measure brain activity in
FFA when viewing Greebles
Analysis: compare activity in
FFA before and after training
Greebles Study

Conclusions
Plasticity of FFA
 FFA selects for
things we’re experts
about
 Faces are things
we’re experts about
