6 - Neuroscience methods

Question 1

Structural MRI: as opposed to functional MRI

Answer 1

• Goals of structural imaging with non-invasive methods:
  a) to study anatomy
  b) to identify abnormalities (as in brain disease)
  c) to follow development (childhood to old age)
  d) to show plasticity
• methods of interest to biological psychology:
  a) computed tomography (CT) scans
  b) MRI imaging
• CT and structural MRI relies on contrast between tissue types (whit vs grey matter vs cerebrospinal fluid)

Question 2

Structural MRI for biological psychology: brain plasticity related to motor learning

Answer 2

• Learning to juggle = changes to brain?
• Clusters of significant expansion of grey matter
• Corresponded to area hMT/V5, a visual motion area
• Extrastriate visual areas: Process input from geniculostraite system
• hMT/V5 location: lateral occito-temporal cortex, part of extrastriate visual cortex (separate from primary visual cortex)
• Learning to juggle requires improving visually guided action (relies heavily on hMT/V5)
• Increased grey matter in these regions supports improved motion processing
• Demonstrates learning-related brain plasticity

Question 3

How to generate structural MR contrast

Answer 3

• Protons rotation axes are randomly orientated before entering the scanner (no overall magnetization is present)
• Inside the scanner: strong external magnetic field is generated by the magnet
• Protons rotation axes become aligned with the magnetic field (net magnetization vector)
• Radiofrequency coil: briefly disturbs proton alignment (RF pulse)
• Magnetization is tipped perpendicular, but returns to alignment with main field after RF pulse (can be repeated many times in milliseconds)
• Why important: different tissues relax at different speeds = create image contrast
• T1 relaxation: recovery of longitudinal magnetization overtime

Question 4

Functional MRI

Answer 4

• Goal: identify brain areas involved in sensory and cognitive processes, build models of brain function
• Measure blood flow changes in activated brain areas
• Require contrast between activated and non-activated tissue
• E.g. foot movement = medial motor cortex activation
• E.g. finger movement = lateral motor cortex activation
• fMRI activation map matches the somatotopic map of motor cortex
• Demonstrates that non-invasive fMRI can reveal functional brain organization

Question 5

How to measure neural activity in functional contrast: BOLD effect

Answer 5

• Problem in fMRI: How to measure neural activity using a functional contrast, Contrast must distinguish activated vs non-activated tissue
• T2 and T2* signals: Unlike T1, T2 and T2* signals decay after the RF pulse
• The rate of decay is tissue-specific
• Similar to T1, but now describing signal decrease, not increase
• Why T2* is used in fMRI: T2* decay changes with the functional state of tissue (depends on the oxygenation state of haemoglobin)
• Haemoglobin properties:
• Oxy haemoglobin: diamagnetic → does not distort the magnetic field
• Deoxy haemoglobin: paramagnetic → makes the magnetic field inhomogeneous
• Effect on MR signal:
• More deoxy-haemoglobin → stronger field inhomogeneity → faster T2* decay → lower signal
• More oxy-haemoglobin → more homogeneous field → slower T2* decay → higher signal
• BOLD contrast: Differences in T2* signal form the Blood Oxygen Level Dependent (BOLD) effect
• Link to brain function:
• Neuronal activation increases local blood flow
• Increased blood flow raises oxy-haemoglobin levels
• Therefore, the BOLD signal indirectly reflects neural activity

Question 6

How to generate functional contrast

Answer 6

• MR signal measured every 2 seconds (typical signal increase is 4%)
• BOLD response to single stimulus: peak signal change = 2% (peak = 6 seconds, return to baseline = 24 seconds)
• BOLD response is slow compared to neural activity
• Signal changes are small
• Experimental design must handle long responses and low signal to noise ratio
  1. Simple design: single stimuli separated by long intervals to allow to return to baseline (time-consuming)
  2. Block design: stimuli grouped into blocks and BOLD responses overlap and add up = produces strong signal (limited flexibility)
  3. Rapid event-related design: different stimulus types presented in a mixed, random order, variable stimulus-onset asynchronies (SOAs)

Question 7

fMRI: block design

Answer 7

• BOLD responses are additive, so block designs increase signal strength
• Higher signal = higher statistical power
• Stimulus timing is predictable
• E.g. data recorded from occipital cortex, measurement taken every 3 seconds
• Results: MR signal lower during rest, increases during each stimulus period, clear separation between rest and stimulus phases
• fMRI measures signal from many brain locations simultaneously
• light grey = higher signal, dark grey = lower signal
• increased signal appears only in occipital cortex
• confirm visual cortex activation in response to visual stimuli

Question 8

fMRI: block designs vs event-related designs

Answer 8

1. Block:
   - Advantages: high statistical power, strong/ continuous activation
   - Disadvantages: inflexible, limited number of experimental conditions, predictable stimulus sequence
2. Event-related:
   - Advantages: reduces habituation, can analyse different response types
   - Disadvantages: lower sensitivity compared to block designs

Question 9

Preprocessing: motion correction (single subject)

Answer 9

• Spatial preprocessing steps: motion correction, coregistration, normalization, spatial smoothing
• Why motion correction is needed: fMRI data consists of many images collected overtime and head movement caused misalignment = distorted results
• Motion correction: realigns all images to reference image, ensures each brain location is compared across time
• Activation maps: after analysis, activated regions are shown as functional contrast maps
• Functional images are aligned to the pp’s structural MRI
• Improves spatial interpretation of brain activation

Question 10

Preprocessing: normalization (between-subjects)

Answer 10

• Aligning multiple pps’ activation maps
• Purpose of alignment:
  a) Enable group-level analysis – compute average patterns of brain activation
  b) Facilitate anatomical labelling – compare activation with a brain atlas to identify regions
• Labelling brain locations using an atlas: MR images must be in the same coordinate system as the atlas (Talairach-Tournous atlas), using coordinates to identify a region
• Advantages: highly standardized procedure, enables comparison across studies
• Limitations: based on one brain and one hemisphere, brain shape may have changed during fixation, precision is limited

Question 11

fMRI statistics

Answer 11

• fMRI statistical analysis replaces informal inspection with a quantitative approach
• Predicted time course of activation is defined based on stimulus timing
• Each voxel’s MR signal is compared to predicted time course:
  a) Occipital voxel: signal matches prediction– labelled activated
  b) Frontal voxel: signal doesn’t match – labelled not activated
• General linear model is used for voxel-by-voxel statistical testing
• Limitations: low statistical power, type 1 error control, null results are hard to interpret

Question 12

Neuropsychology: effects of brain lesions

Answer 12

• Goal: relate brain anatomy to behaviour
  1. Albert task: pp fixates on central dot and task is to cross out all lines
• Results of typical hemineglect: only lines of the right of fixation are cancelled, left-side lines are ignored, despite intact visual field
  2. Line bisection task: pp fixate on central dot and task is to bisect lines at midpoint
• Typical hemineglect results: marks bisection points far to the right, left half of lines not perceived
  3. Drawing/coping task: copy two flowers on right and left
• Results: both flowers copied but left half of each flower omitted. Shows deficit can be object-centred, not just space-centred
• Lesions usually at right temporo-parietal junction (spatial attention)

Question 13

Goal of neuropsychology: relate brain anatomy behaviour

Answer 13

• Combine two types of knowledge:
  1. Structural & functional anatomy:
• Structural (CT, MRI) – locate brain lesions
• Functional (fMRI, EEG) – identify regions supporting specific function s
  2. Behavioural testing:
• Identify which functions are impaired and which are spared after a lesion
• Linking anatomy and behaviour: look for patterns of co-occurrence between anatomical findings and behavioural deficits
• Sub-goals: 1. Localize impaired behaviour to damaged regions 2. Confirm preserved functions are not localized to damaged regions
  a) Association: damage to one brain region – multiple deficits
  b) Dissociation: damage – impaired task A but normal task B

Question 14

Studying associations

Answer 14

• Damage to a single region – multiple deficits in a typical combination
• Functions A, B, C likely require the same neural circuit
• A, B, C could be separate neighbouring regions affected together
• Region X may be a relay station connecting distinct regions for A, B, C
  1. Balint’s syndrome: example of an association
• Typical cause: lesions in parieto-occipital cortex
• Main symptoms: simultanagnosia (see one object at a time), oculomotor apraxia (inability to planned eye movements), optic ataxia (difficulty reaching seen target)

Question 15

Studying dissociations: visual form agnosia

Answer 15

• Damage – impaired task A, but normal task B
• Reveal functional differences between separate neural networks
• E.g. lesion to ventrolateral occipital cortex
  a) Pp normal results for visuomotor task
  b) Performs poorly for perceptual matching task
• Lesion impairs vision for recognition but vision for action spared
• Clear dissociation between perception and action pathways
• The two tasks rely on separate neural networks

Question 16

Studying double dissociations

Answer 16

• Patient A – visual form agnosia (ventral-occipital lesion)
• Patient B – optic ataxia (bilateral occipitoparietal lesion)
  1. Visual shape discrimination: A poor performance, B good
  2. Visually guided grasping: A good, B poor
• Opposite patterns in two patients – double dissociation
• Strong evidence for separate brain circuits:
  a) Ventral stream: vision for recognition/ perception
  b) Dorsal stream: vison for action/ skilled movement

Question 17

Visual pathways: lesions to the ventral stream

Answer 17

• Separations exists from retina to cortex
  1. Dorsal stream: vision for action
• Input: Magnocellular cells
• Function: guides visually controlled movements
  2. Ventral stream: vision for recognition
• Input: Parvocellular cells
• Function: object recognition and identification