Transcript measuring

What are we measuring in fMRI?
Caroline Catmur
Jack Kelly
In BOLD fMRI, we are measuring:
the inhomogeneities introduced into the
magnetic field of the scanner…
as a result of the changing ratio of
oxygenated:deoxygenated blood…
via their effect on the rates of dephasing of
hydrogen nuclei.
Ehhh???
Physics: underlying principles
• Hydrogen nuclei (1H): positively charged
particles which spin around their axes,
producing a (small) magnetic field.
• MDM: magnetic dipole moment: vector of the
magnetic field of the nucleus.
When placed in a uniform
magnetic field, (conventionally
indicated by the z axis), the
particles’ MDMs align with or
against the field. A small percentage
more align with the field than
against, proportional to the strength
of the field, giving the particles a net
magnetization.
• The MDMs also precess around the axis of the field, at a
resonant frequency dependent on the strength of field
and type of nucleus, eg 64MHz for 1H in a 1.5T field.
So what goes on in the scanner?
• Place the nuclei (ie the brain) in a uniform magnetic field
(the scanner).
• The next step: apply an RF pulse, frequency equal to
frequency of precession of the nuclei, normally at 90° to
the magnetic field. This ‘tips’ the MDMs of those nuclei
which have this frequency of precession, ie we only ‘tip’
the 1H nuclei.
• So, the MDMs of the 1H nuclei are now
at 90° to the main field, ie in the x/y plane.
• Terminate the RF pulse and the nuclei relax: their MDMs
return to the original orientation in the z dimension, and
the energy released during relaxation is what is
measured by the receiver coil.
• Three different relaxation times of interest in MRI: T1, T2
and T2*.
Phase
• Before the RF pulse, all the
MDMs precess at the same
frequency but not in phase.
• After they’re tipped, all
precess in phase. Can think
of it as all MDMs moving
together: this produces a strong signal in the x/y plane.
• Once the RF pulse ends, begin to dephase: start to
cancel each other out and the signal decays.
• Two reasons for this dephasing: inhomogeneities in the
magnetic field, and ‘spin-spin’ interactions between
neighbouring nuclei.
• Possible to correct for dephasing due to inhomogeneities
in the field by applying another RF pulse at 180° to the
initial pulse. Known as a spin-echo sequence.
Back to those relaxation times
• T1 relaxation: time course for the
MDMs to return to their original
(z) orientation.
T2 relaxation: time course of the
breakdown of the magnetization
in the x/y plane due to spin-spin
interactions.
• T2* relaxation: time course of the breakdown of the
magnetization in the x/y plane due to variations in the
magnetic field. The T2* processes can be refocused
using a 180° spin-echo sequence, though the T2
processes will still remain.
• Different tissues have different T1 and T2 relaxation
rates.
• T1-weighted scan: measure signal at time when relative
difference (between tissue types) in amplitudes of MDMs
in z dimension is maximum.
• T2-weighted scan: measure at time when relative
difference in amplitudes of MDMs in x/y plane is
maximum.
• To get these different scans, change time between RFpulse and measurement (TE), and between successive
RF pulses (TR).
But why do we need to know all this?
• BOLD (blood oxygenation level dependent) contrast:
measures inhomogeneities in the magnetic field due to
changes in the level of oxygen in the blood. So it’s a T2*
contrast.
• Oxygenated blood contains oxyhaemoglobin: red blood
cells with O2 molecule attached. Not magnetic.
• Deoxygenated blood: deoxyhaemoglobin: red blood cells
without O2. Magnetic.
• So if ratio deoxygenated:oxygenated blood is high,
increases inhomogeneities in the magnetic field  faster
breakdown of magnetism in x/y plane (T2* relaxation) 
decrease in fMRI signal.
• If ratio oxygenated:deoxygenated is high, slower T2*
relaxation  less decrease in signal.
• So we can use the change in fMRI signal to infer the
relative oxygenation of the blood.
So how do we get the actual information?
• Spatial localisation: ‘gradients’. Small magnetic field
gradients (eg 30 mT/m) superimposed onto the main
static magnetic field.
• Remember that the resonant frequency for a nucleus in
a magnetic field depends on the field strength.
• So, differences in the resonance frequencies encode the
positions of the nuclei along the gradient field.
• Switching the small gradients on and off is noisy!
• Receiving the information: the RF coil both transmits and
receives. A volume coil images any part of the brain; a
surface coil gives better images, but only for the nearest
part of the brain, due to distortions. A phased array coil is
a series of surface coils.
fMRI – neurophysiology
fMRI
Outline:
• What is BOLD?
• Correlation of BOLD with electrophys.
• How neurons cause CBF increases
• Localising BOLD
• Summation of BOLD
• Implications for cognitive studies
BOLD and MRI
• BOLD = Blood Oxygenation Level
Dependent
• functional Magnetic Resonance Imaging
• Deoxyhemoglobin is paramagnetic and
produces a reduced signal,
oxyhemoglobin is weakly dimagnetic and
doesn’t reduce the signal.
BOLD and Cerebral Blood Flow
BOLD and electro-physiology:
correlation
Same area in V1 of cat,
Kim DS (2004)
BOLD and electro-physiology:
correlation
Single unit recording, cat V1,
Kim 2004
What causes BOLD?
• The purpose of the increase in blood
oxygenation is to feed neurons…
• …so, what makes a neuron hungry?
• (neurons can’t store much energy)
Vascular density
• Vascular density is proportional to synaptic
density, not soma density
Hungry brains
White matter uses
¼ the energy of
grey matter per
unit volume
62% of
mitochondria are
in dendrites
Attwell & Iadecola 2002
Regulation of blood flow
Regulation of blood flow
• Is it feedback or feed forward?
Activity
Uses energy
Vascular system
must supply more
energy
Activity
Directly commands
more blood flow
Regulation of blood flow
• Feedforward!
Activity
Directly commands
more blood flow
• Energy use does not directly increase blood
flow…
•…so how does tell CBF to increase?
Feed forward pro-active control
Monoamines and blood flow
• DA, NA and 5HT = vasoconstriction
• Cholinergic axons from BF = vasodilation
• This complicates neuropsychiatric studies
• e.g. schizophrenia, PD, ADHD
Localising fMRI
Cat scanner (!);
Kim 2004
Correlation of BOLD and single unit
Kim 2004
Summation of BOLD
BOLD
LFP
Single unit
Comparing different areas
• Different vasculature
• Different neuromodulatory control
• Different circuitry
• BOLD [X] > BOLD [Y] does not mean
NEURAL ACTIVITY [X] > ACTIVITY [Y]
Comparing different subjects
What BOLD does not measure
• The output of an area
• Comparisons of activity between areas
• GABA ??????
What does contribute to BOLD
• Synaptic activity
• Local processing
• Sub-threshold neuromodulatory inputs
What does a blob in area X mean?
• X has changed its local activity
• Change of modulatory inputs arriving at X
• Change of inputs arriving at X
• (beware: the areas giving rise to the inputs
to X may not produce a BOLD signal if
their local synaptic activity levels remain
constant)
References
•
Logothetis NK & Wandell BA (2004) Interpreting the BOLD signal Annu. Rev. Physiol
66:735-69
•
Attwell D & Iadecola C (2002) The neural basis of functional brain imaging signals
Trends in Neurosciences 25:621-625
•
Kim DS et al (2004) Spatial relationship between neuronal activity and BOLD
functional MRI NeuroImage 21:876-885