Week 1

Indirect vs direct brain measures of activity

DIRECT: measures or variables of actual neural activity

INDIRECT: involves a medium to infer neuronal activity from the variable measured

note: spatial vs temporal tradeoff, scale / size depends on question

fMRI (general)

INDIRECT measure → metabolic

• Measures blood oxygen level-dependent signal (BOLD)

  • Occurs due to changes in localised blood flow to the brain in response to neural activity

• Ratio of deoxygenated to oxygenated haemoglobin

  1. At rest → even ratio

  2. Neural activity onset → initial drop in oxygenated Hb

  3. Sustained activity → oxygenated Hb ↑ (overcompensation)

• Reflects neuronal activity via neurovascular coupling

  • Causes a gap / delay for BOLD response output

fMRI strengths vs limitations

Strengths:

• Excellent spatial (allows detailed deep brain structure imaging)

• Non-invasive, safe for humans
  

Limitations:

• Poor temporal (neurovascular coupling, can only image every ~s)

• Artificial imaging environment can modify brain response

• Movement can distort image acuity, produces artefacts

• Cannot image patients with metal implants

• Takes a long time

EEG (general)

DIRECT measure

• Pyramidal neurons on the cortical sheet project their axons superficially

  • Electrical signals flow in parallel with these axonal projections

  • EEG can detect the polarity that occurs with neuronal firing

• Patient’s hair is covered in gel to connect electrodes to the skull

  • Measures superficial voltage changes

• Better for deeper brain structures and gyri (raised ridges)

EEG strengths vs limitations

Strengths:

• Excellent temporal (ms)

• Relatively non-invasive

• Effective for deep brain structures (travels further than magnetic)

• Lower cost relative to MEG
  

Limitations:

• Sensitive, susceptible to interference, requires frequency filtering

• Skull and brain tissue have own electrical current, distorts activity

• Poor localisation (can’t map 2D image into 3D)

• Correlational to behaviour, not causative

MEG (general)

DIRECT measure

• Cortical pyramidal cells generate a small circular electric current

  • Subsequently also produce a magnetic field that propagate radially (perpendicularly)

• Better for superficial detail and sulci (cortical folds)

MEG strengths vs limitations

Strengths:

• Excellent temporal (ms)

• Slightly better spatial resolution than EEG

  • Cleaner signal, magnetic fields unaffected by bone / tissue
    

Limitations:

• Magnetic fields drop quickly, only readable close to a source

• Very expensive (cryogenics, shielding from environment)

• Poor localisation (can’t map 2D image into 3D)

• Correlational to behaviour, not causative

PET (general)

INDIRECT measure → metabolic

• Can measure a protein or antibody of interest using a radioligand probe

  • Tagged using a positron-emitting radioactive isotope

• Individual is injected with said radionucleotide

  • Performs a behavioural task / stimulus exposure

  • Active neurons consume oxygen causing localised ‘pooling’ of the radioligand

• Positrons form the radionucleotide collide with electrons emitting gamma radiation

  • Scanner detects gamma rays to reconstruct the original location(s) of the radioligand

• Not activity over time but average standard uptake value across different regions (i.e., every 30 min for 2h)

PET (strengths vs limitations)

Strengths:

• Shows changes on a molecular level, not just morphology
  → cellular function and metabolism

• Used in neuropharmacology

• In-vitro visualisation
  → can detect early-stage neurodegenerative disease
  

Limitations:

• Poor temporal (mins)

  • Doesn’t reflect changes in activity over time

• Poor spatial relative to fMRI

• Expensive, requires a cyclotron, takes up a lot of space

• Radiotracer injection

TMS (general)

INDIRECT measures

• Capacity to shut down or disrupt activity

• Powerful electromagnetic current pulsed through a wire coil on a person’s scalp

  • Generates large, alternating magnetic fields in a short period → can penetrate skull

  • Collides with conductive brain tissue and creates a localised secondary electrical current

  • Disrupts activity of receiving neurons

• Frequency dictates effect on target brain region

  • High frequency pulses cause depolarisation, excitatory effect

  • Low frequency pulses suppress / reduce activity, inhibitory

TMS strengths vs limitations

Strengths:

• High temporal

• Non-invasive

• Relatively inexpensive

• Causative relationships

  • Temporary neuronal depoliarisaiton → behavioural output)

  • Allows within-subjects analysis (= higher statistical power)
    

Limitations:

• Poor spatial resolution (TMS wand is very large)

• Measures the absence of activity, not activity itself

• Sensitive to magnetic / electric interference

• Mostly superficial readings

• Electric shocking people

• High seizure risk for individuals with epilepsy

Single-unit recordings (general)

DIRECT measure

• Places a microscopic microelectrode near a cell membrane to measure the spiking patterns of a single cell

  • Measures action potentials

  • Requires penetration of brain tissue

Single-unit recordings strengths vs limitations

Strengths:

• Excellent temporal (ns, mean firing or spikes/s)

• Very good spatial, can measure specific neurons and axons)

• Most direct measure of brain activity
  

Limitations:

• Highly invasive, not performed on humans

• Very expensive, only used for research (requires surgery)

• Correlational, not causative

Brain lesions (general)

INDIRECT measure

• Observation of cognitive, physical, behavioural changes in response to localised damage

  • Inference of region function, causative

  • More the absence of brain activity, not activity itself

Brain lesions strengths vs limitations

Strengths:

• Demonstrates causation (structure-function relationship)

• Can study disease modifications, pathology symptoms

• Can be naturally occurring

Limitations:

• Unethical, uncommon even in animal studies

• Non-selective, requires inference

• Poor spatial and temporal

• Irreversible, permanent anatomical modifcaiton

Neural Doctrine for Perceptual Psychology (Barlow, 1972)

The best method is studying the “single neuron / unit”

1. Sparse representations
   - Neurons use a lot of energy, must fire sparingly
   - Sparse representations, few neurons can encode important things
   

2. Environment influences neuronal activity
   - Spiking triggered by patterns of stimulation
   - Neurons adapt to be optimised to recurring environmental stimuli
   

3. We see the world through macroscopic representations
   - Perception is ‘high level’
   - We don’t see lines but ‘stuff’
   

4. A neuron’s activity increases in the presence of its specialised trigger / stimulus
   - ‘High impulse frequency’, fires a lot when it detects its ‘thing’
   - Linked to energy conservation

MRI (general)

INDIRECT measure → structural

• Detects the behaviour of hydrogen atoms to differentiate tissue types on high-resolution images

  • Protons orient themselves parallel to a strong magnetic field

  • When hit with a radio frequency pulse, protons are knocked off their orientation turning 90 deg away

  • Then, they emit a signal as they relax and realign

• MRI measures the time taken to align back to the magnetic field

  • T1 (timepoint 1) = parallel relaxation time
    → protons returning to resting state

  • T2 (timepoint 2) = perpendicular relaxation time
    → protons falling out of phase

• Time taken indicates tissue type

  • Water < grey matter < white matter

DTI (general)

INDIRECT measure → structural

• Specialised type of MRI to measure and map white matter axon tracts using the diffusion of water

  • Instead of a static picture, tracks the direction and speed of water diffusion

    1. Isotropic diffusion: free water molecule will move randomly in all directions in open space (i.e., spinal cord CSF)

    2. Anisotropic diffusion: restricted movement, rate and direction depend on environment like structural paths (i.e., water of axon fibres is wrapped in sheaths preventing sideways movement)

• Radiofrequency pulses are projected in many different directions

  • Measures how much signal is lost along different orientations

  • Compiles to map space in 3D