neuropsychology-method

Neuropsychology as a research method

  • Strengths & weaknesses of studying people with brain damage
  • Transcranial magnetic stimulation (TMS) – relationship to traditional neuropsychology
  • Neuropsychology: The study of brain-damaged patients

Flavors of neuropsychology

  • Clinical: Devoted to diagnosing troubles of a given individual, & devising a rehabilitation strategy for that patient (medical)
  • Classical localizationist: Map cognitive deficits onto location of brain damage to try to localize cognitive functions
  • Cognitive: understand what basic cognitive functions ARE, by seeing how different abilities hang together or separate after brain damage

Reverse engineering notion

  • Infer function of a region (or cognitive mechanism) by removing it & measuring effect on the rest of the system
  • Ex: if damage to a region disrupts reading, but not understanding speech, not identifying faces (etc), then one might conclude that the region is specialized for some aspect of processing text

Ways to get brain damage

  • Stroke - disruption of blood flow

  • Brain infection - usually viral

  • Tumors

  • Brain surgery to correct some other problem (remove tumor, remove epileptic tissue, etc)

  • Head injury (car wreck, gunshot, etc)

  • Degenerative disorders (Alzheimer’s disease, etc)

  • different sources of damage can be categorized to 2 factors:

  • Specificity of damage: Circumscribed brain region versus widespread

  • Onset: Sudden versus gradual

Typical steps in neuropsychological research

  1. Medical evaluation by neurologist to diagnose CAUSE (stroke, infection, etc) & start any possible treatment

  2. Part of medical evaluation with some sort of brain imaging scan: Magnetic resonance image and/or CT scan: may or may not show an obvious lesion

  3. When patient is stable, standard neuropsychological testing to determine general nature of problems: are they in language, memory, perception? Start any rehabilitation program that is suitable.

  4. Only a handful of patients then join research studies.

  5. Overall idea: selective in who is assessed in studies

Localizationist neuropsychology: Map deficits onto lesion locations

  • Old version: Wait for patient to die, examine brain visually for general location of damage
  • Modern version: Do structural (anatomical) MRI scan while patient is still alive. Classify each portion of brain image as “intact” or “damaged”

 

  • All MR images consist of individual image elements: voxels (3D pixels).
  • Smaller voxels = crisper images (higher spatial resolution), look less blurry

Localizationist neuropsychology: Map deficits onto lesion locations (STEPS)

  1. Get a bunch of patients & do MR scans of their brains.
  2. Classify each voxel in each brain as damaged or intact. For a given voxel, some patients will have damage there and some will not.
  3. Give patients some task and get performance scores.
  4. For each voxel, do a correlation test between intact/damaged & performance score in task.

 

  • Set up for correction b/w voxels & performance scores

Dronkers et al, 2004

 

  • 100 patients w/ stroke in left hemisphere.

  • All got test of sentence comprehension.

  • Colored areas on brain images mean that some patients had damage there (at least 1)

  • exact color shows correlation b/w having damage in THAT spot & sentence comprehension. (statistical significance of performance & damage in brain)

  • Blue: unrelated to language test

  • Red: deficiency part of language processing

Conceptual issues in neuropsychology

  • Most complex behaviors served by brain circuits rather than single areas

  • A lesion in 1 location might disconnect two other regions that are critical for the task (relevant to localizationist flavor)

  • transparency assumption: damaged brain is like a normal w/ something missing

  • Brain re-organization after damage. If undamaged area takes over an old function, this assumption will be violated (relevant to localizationist flavor).

  • Development of new strategies for doing old jobs can make cognitive neuropsychology difficult too, but good experimental design can detect this.

  • Universality assumption: All people have more or less same brains, & same functional organization – so that a small # of patients tells us about humans in general.

Critical role for dissociations in neuropsychology

  • Single dissocation: A patient is normal (or at least decent) in performing Task A, but impaired in performing Task B

  • hard to intepret b/c of possible role of task difficulty – brain damage will impair a harder task before it influences an easier task.

  • Double dissociation:

  • Patient 1 can do Task A but not B

  • Patient 2 can do Task B but not A

  • This pattern cannot be attributed to task difficulty.

  • Tells us that Tasks A and B rely on different brain circuits, and draw on fundamentally different cognitive abilities

Comparing methods in cognitive neuroscience (Desirable properties)

  • Safety: Noninvasive better
  • Cost: Cheaper better
  • Sample size: How much data is sufficient, and how easy is it to collect that much data?
  • Coverage: How much of the brain can we investigate?
  • Causality vs correlation in brain/cognition relationships. Which does the method show?
  • Spatial resolution
  • Temporal resolution

Spatial resolution

  • Where do you live?

  • “In New York State” - Low spatial resolution answer

  • “111 Chestnut St Binghamton, 13905” - High spatial resolution answer

Spatial resolution in neuroscience

  • Where did brain activity differ b/w two conditions?

  • Left temporal lobe = low spatial resolution answer

  • Superior and/or middle temporal gyri, fairly posterior, left hemisphere = medium spatial resolution

  • Sulcus b/w superior & middle temporal gyri, in posterior 2 cm, left hemisphere = high spatial resolution

 

  • Same brain area or different brain areas?

Temporal resolution

  • How long does it take to drive to NYC from Binghamton?
  • “More than a minute, but less than a week” = Low temporal resolution answer
  • “About half a day” = Medium temporal resolution answer
  • “2 hours and 45 minutes” = High temporal resolution answer

Temporal resolution in neuroscience

  • 2 brain areas were active.

  • Can our method tell us WHEN they were active?

  • Low temporal resolution: both active while people did the task.

  • High temporal resolution: temporal lobe area was active BEFORE (or AFTER) frontal lobe area.

  • Different interpretations depending on timing of activity

  • Temporal THEN Frontal: Sensory cortex is feeding info to frontal lobe.

  • Frontal THEN Temporal: frontal lobe is guiding sensory cortex about how to process sensory input.

Pros and cons of neuropsychology as a research method

  • Temporal resolution: None really
  • Spatial resolution: Lesion location and size depends on nature of damage. Lesions often as large as a single Brodmann area
  • Safety: Zero risks
  • Sample size: How many similar patients can we find?
  • Cost: Relatively low
  • Causality vs correlation: only secure way of showing a causal link b/w brain area & task performance. A big advantage of the method.
  • Transcranial magnetic stimulation (TMS) as an alternative to neuropsychology

“Temporary lesion” method: Transcranial magnetic stimulation

  • Coil contains a wire carrying an electric current
  • Rapid change in current creates a magnetic field
  • Magnetic field induces a current in neurons near coil
  • electric current adds noise to normal intricate pattern of electrical communication among neurons
  • Disrupts brain activity that was occurring at that point in time
  • “Virtual lesion” idea

Safety of Transcranial Magnetic Stimulation (TMS)

  • Stimulation is capable of causing seizures & short-term complaints

  • never used on people w/ epilepsy

  • Number & rate of pulses is regulated by ethics guidelines

  • Occasional reports of beneficial effects (e.g. in depression) … but used to temporarily impair performance in some task.

Pros and cons of TMS as a research method

  • Temporal resolution is very good

  • Activity of a brain area can be disrupted at a precise time (i.e., turn on pulses 40 milliseconds after stimulus, or 75 ms, or 100 ms, etc)

  • b/c effect has a sudden onset and is temporary, no gradual brain re-organization

  • Spatial resolution: not precisely known, maybe a centimeter, maybe several? But smaller than most natural lesions

  • Safety

  • Sample size: Can apply TMS to same location in numerous subjects – greater standardization than w/ natural lesions. larger sample size easier to get.

  • Cost: Moderate. More expensive than behavioral testing alone, but less than some other methods.