Introduction and Lesion Studies

  

—  The Mind Body Problem

—  How can brain give rise to feelings, thoughts, emotions?

—  Dualism (Descartes): Mind and body are 2 different things; mind is non-material and immortal.  Interact through pineal gland

—  Dual aspect monism (Spinoza): Mind and brain are 2 levels of explanation or experience of the same thing;
analogy = electron – can be described as wave or particle
Thus mind is our subjective experience of brain processes

—  Monism/reductionism (Crick): Psychological concepts (eg memory, feelings) will reduce to biology as we increasingly understand brain processes

—  Historical Context

—  Earliest psychologists interested in studying emotions, consciousness etc

—  No tools in science, neurology, etc for doing this – no objective way to study these subjective states

—  Therefore turned to dubious methods – introspection; free association; interpretation of dreams, etc etc

—  Behaviorists rejected all this – tried drag psychology back to realm of
objective science

—  Cognitive revolution in turn = response to deficiencies inherent in behaviorism

—  After Behaviorism came Cognitive Revolution

—  Have to study the contents of the ‘black box’

—  Coincided with development of computer technology

—  Cognitive psychology came to use ‘computer metaphor’
of mind

—  INPUT à PROCESS à OUTPUT

—  Models of higher fx at fairly abstract level – refer to stages and processes; lots of flow charts

—  Cognitive Psychology

—  Key concept:
Can develop coherent & testable theories of
cognition that do not refer to the brain

—  Key metaphor:
Brain as computer

—  Information processing approach

PERCEPTION      ATTENTION      STM      LTM

—  Can understand cognitive processes as software programs/algorithms – relatively independent
of hardware

—  Cognitive Psychology cont … NB Concepts

—  Serial vs Parallel processing
Early models had serial stages of processing
                       1^st this, then the next step, and so on

Interactivity – idea that stages not independent; later stages can begin before initial stages complete

—  Parallel processing – different processing occurs at same time

—  Top down vs bottom up processing
Bottom up = from sensation upwards
Top down = from concepts/knowledge down
Top down processes can impact on processing of sensory information
 

—  Connectionist Models of Cognition – Neural Networks

—  AKA Computational models of cognition or PDP – parallel distributed processing

—  Very influential models

—  Architecture
Nodes: simple information carrying units
(analogous to neurons)
          Respond to certain inputs; produce certain outputs

—  Nodes interconnect – more strongly with some
than others

This reflected in WEIGHTS

—  Neural Networks

—  Connectionist models; parallel distributed processing

—  (Very) Roughly analogous to actual
neural organization in brain

—  NODES analogs of neurons

—  Inputs and outputs analogous to
neural connectivity

—  Weights of inputs rough equivalent of excitatory vs inhibitory neurotransmitters

—  Neural Networks cont …

—  Claim that computerized neural networks have ‘neural plausibility’ is controversial

—  Networks developed that model dyslexia; semantic dementia, etc

—  Nodes responding to varying weighted inputs; weight changes dependent on learning or training of network; hidden nodes; all similar to what we know of actual neurons & connectivity

—  However, criticisms around actual applicability to neural tissue and networks

— 
 The Return of the Brain … Cognitive Neuroscience

— 
 The Return of the Brain … Cognitive Neuroscience

—  Phrenology
Idea of functional specialization – different areas in brain devoted to specific cognitive functions

—  Modern cognitive neuroscience uses empirical methods
to ascertain different functions

—  It does not assume that each region has one function or
that each function has a discrete location
(unlike phrenology),
but does assume some degree of specialization of neurons
in particular regions

—  Functional specialization or Modularity?

—  Functional specialization =    brain contains regions that specialize in particular types of information processing

—  VERSUS

—  Equipotentiality = idea that any bit of brain can do
any type of processing

¡  Very clear that latter NOT true

¡  BUT how far can we go in localizing function?

—  Fodor – MODULES

—  These are domain specific (process only 1 type information)

—  Allow rapid efficient processing independent of other  cognitive systems

—  May be innate

—  CENTRAL SYSTEMS are domain general (eg memory)

—  Strong version of modularity not supported, but still active debate & research around functional specialization and relative independence of some processes

—  Development of Cognitive Neuroscience cont …

—  Advances in neuroimaging created radical change

—  Allowed study of brain in ways IMPOSSIBLE before

—  This area of investigation went from being peripheral
(physiological psychology; clinical neuropsychology)
to dominant focus internationally

¡  Structural imaging – improvements allowed lesions to be located and measured far more precisely than ever before

¡  Functional imaging – allowed study of living, working
brain tissue

—  1970s: structural imaging methods (CT, MRI) enable precise images of the brain (and brain lesions)

—  1980s: PET adapted to models of cognition developed by psychologists

—  1985: TMS is first used (a non-invasive, safer equivalent of Penfield’s earlier studies)

—  1990: Level of oxygen in blood used as a measure of cognitive function (the principle behind fMRI)

—  Does Cognitive Psychology need the Brain?

—  Information processing models do not refer to brain

—  Opinions divided:

¡  Some think cognitive psych can inform theories and research in neuroscience

¡  Others want to keep cognitive psychology distinct; are very sceptical abt cognitive neuroscience, esp neuroimaging
(‘new phrenology’)

—  NB to note that computer software analogy misleading: cognitive processes are NOT independent of the hardware – brain constrains how and where information processing happens

—  Does Neuroscience need Cognitive Psychology?

—  Memories, thoughts, perceptions etc
= stuff of cognitive psychology

—  Grey matter, white matter, blood supply
= stuff of neuroscience

—  Framework linking two has to deal with
mind-body problem

—  Does Neuroscience need Cognitive Psychology? – cont …

—  Neuroimaging revolutionized brain sciences

—  Is it the new phrenology?

—  Problems with phrenology:

¡  No real scientific grounding

¡  Used naïve psychological concepts

¡  Insights from cognitive psychology thus vital to inform theory and develop appropriate
research questions

§  PSY2010
cognition and neuroscience

§  Lesion Studies

§  Key Concepts

§  See 10 interesting facts about the brain
(Ch 2, p. 17)

§  Remember to think of brain in evolutionary terms; also in hierarchical terms

§  Newer structures added to older structures

§  Often newer structures allow more complex processing; but older structures, processes can be hugely influential
(eg emotion over-riding reason)

§  Lesion studies

§  = Reverse engineering

§  Infer function of region by examining what rest of system can and cannot do when particular part is damaged

§  Oldest method in brain studies

ú   Broca – patient with severe language deficit

ú   Wernicke – clarified 2 (at least) components of processing language; and 2 discrete locations implicated

§  Distinct Research Traditions

§  Classical neuropsychology

§  Group studies of individuals with damage to particular region, contrasted to normals

§  Helped enormously by improved structural imaging techniques

§  Constrains and informs interpretation of functional neuroimaging data

§  Best suited to establishing lesion-deficit associations

§  Cognitive neuropsychology

§  Single cases, patterns of deficits and spared abilities used to infer structure of cognition, no reference to brain location

§  Guides development of detailed information processing models

§  Provides theoretical framework that underlies much neuroimaging research

§  Best suited to finding out how cognitive processes are subdivided

§  Single case studies

§  This section of chapter particularly difficult

§  Trying to explain some very high level concepts

§  Critical to remember single case study =
method of COGNITIVE NEUROPSYCHOLOGY

§  Therefore not interested in BRAIN or LOCATION

§  When talk of LESION referring to damage to abstract cognitive information processing system – NOT particular part of brain

§  Single case studies cont …
Assumptions underlying theorizing in
this area (Caramazza)

§  1) Fractionation:
     Brain damage can result in a selective
     cognitive lesion

ú   Eg inability to perceive color, although
perception of form, movement etc intact

§  2) Transparency:

§  Lesions affect component(s) of the cognitive system, but DO NOT result in entirely new
system formation

§  Most difficult to defend

§  See cases of plasticity (esp in children); rehabilitation; recovery of function
                                à evidence that brain reorganizes itself

§  C responds that NOT INTERESTED IN THE BRAIN – the system is organized the same

§  3) Universality:

§  Cognitive systems in all people are basically
the same (ie we all process language the same, ..)

§  Again tend to get confusion around level of analysis: the neural (ie location) & the abstract cognitive (ie information processing system)

§  Clearly not every brain is wired the same

§  But are our cognitive systems basically the same? Surely qualitative individual differences exist …
Generally, we are remarkably similar

§  Group studies

§  Method of classical neuropsychology

§  Aimed at establishing anatomical lesion – deficit associations

§  3 approaches to grouping patients

            By syndrome

            By symptom

            By location of
anatomical lesion

§  Group studies cont …

§  Grouping by syndrome à most coarse analysis

§  Grouping by symptom = newer method
Allows finer analysis; also allows discovery
of multiple sites’ involvement in
cognitive process

§  Grouping by anatomical lesion – good for confirming predictions regarding which
areas are critical for particular functions

EG: Grouping by syndrome

§  Right hemisphere syndrome includes anosognosia; unilateral neglect; deficits in visuospatial cognition

§  Not all patients with lesions to right hemisphere
have all features

§  Most often result from lesions in temporo-parietal regions, but can result from lesion anywhere

§  à Gross level of understanding; useful clinically,
but not for fine theoretical understanding either
of cognitive systems OR function of discrete
areas of RH

§  Grouping by symptom

§  Establish that all patients have a particular symptom – have to be able to specify this very clearly

§  Then check what lesion(s) they have

§  Helpful in allowing broader investigation of component parts

§  EG Grouping by symptom

§  EXECUTIVE FUNCTION

§  Used to seen as single area of processing – central executive responsible for planning & carrying out goal directed behavior

§  However, now see separate components

            Stuss et al specified symptoms

            Then looked for lesion location

ú   Able to differentiate function for different prefrontal regions

§  Grouping by anatomical region

§  Useful if have clear prediction

§  EG neuroimaging studies have indicated that the right temporoparietal junction critical for
Theory of Mind

ú   Study of patients with lesion to this region can
confirm or disprove the neuroimaging results

§  Critique of group studies

§  Caramazzo:

§  Cannot assume that just because patients have lesion in same anatomical region à same cognitive lesion

§  He argues that single case study only valid method in neuropsychology

§  Have to carefully assess for each patient exactly what cognitive lesion is present

§  Critique of group studies cont …

§  Problems with imaging

§  Not all scans give equally accurate information

§  CT not as good as MRI

§  Scans done in acute phase don’t show final pattern of damage

§  Therefore in group lesion studies, scans should be done 3 months post injury (seldom done)

§  Critique of group studies cont …

§  Neophrenology

§  If you find that function X is impaired after lesion to region Y, can you conclude that Y is the region for X,
and that it specializes in X?

§  Unfortunately not that simple; other regions may
also be involved

§  Damage to Y is perhaps just disrupting the neural network responsible for X

§  Need to consider: does the region perform other functions? Do other regions support this function?

§  Dissociations

§  Single dissociation = patient impaired on one task, but spared on another

§  Double dissociation = 2 opposite single dissociations; gold standard in classical neuropsychology research

ú   Eg 1 – are REM-sleep and dreaming associated?

            Patients who have no REM yet do dream and
Patients who have REM yet do not dream

ú   This double dissociation disproves the old model of dreaming

ú   Eg 2 – Face recognition separate for animal
vs human faces?

            Patient who could recognize wife but not sheep and

            Patient who could recognize sheep but not wife

§  Dissociations cont …

§  EG 1 shows …

§  Brainstem damage associated with loss of REM, however dissociation shows this region not sufficient to account
for dreaming

§  EG 2 shows …

§  Patients had damage to slightly different areas, therefore dissociation suggests that these different regions responsible for processing human vs animal faces

§  Another eg of double dissociation

§  Broca’s aphasia:

§  Can understand speech; cannot produce it

§  Wernicke’s aphasia:

§  Cannot understand speech; can produce it

§  à conclusion re function of the areas

ú   Broca’s area implicated in production

ú   Wernicke’s area implicated in comprehension

§  Problems to watch out for when noting single dissociations …

§  Task resource artefact

§  You assume the patient is doing much better on 1 task than the other because these tap 2 different capacities, when in fact the one just requires much more of the same capacity

§  Task demand artefact

§  You assume the patient is impaired on one task and not the other because they tap different abilities, but in fact the patient just performed poorly on the one task due to factors other than brain damage (eg didn’t understand instructions)

§  Associations

§  Not really useful in building understanding of components of cognition or functional specialization

§  Most useful clinically

§  à Syndromes in clinical neuropsychology

§  à Clusters of deficits that tend to occur following           particular types of damage

§  Lesion studies in animals …

§  HUGE body of work

§  Has yielded enormous amount knowledge

§  Clearly able to control lesion site/ size

§  Compare performance pre and post

§  Schedule post-mortem analysis whenever you like

§  Is this ethically defensible?

§  Transcranial magnetic stimulation

§  Potential for doing similar work as that
done in animals

§  But on humans, and without lasting damage

§  Very strong magnetic
field stimulates
electrical charge in
brain – very focused

§  Comparing TMS and organic lesions

§    + TMS

§  No reorganization

§  Can determine timing
of cognition

§  Lesion is focal (?)

§  Lesion can be moved within same subject

§  Can do pre and post tests on same brain

§  Can study functional connectivity

§     + Organic Lesions

§  Can study subcortical lesions

§  Lesions can be accurately localized via scanning – TMS may have distributed effects (thru connectivity)

§  Changes in behavior and cognition far more marked;
sometimes totally unexpected