Lecture 6: Fundamentals of fMRI - Part 1 Flashcards
1
Q
What does fMRI stand for?
A
- functional Magnetic Resonance Imagining
2
Q
MRI is tuned to detect changes in
A
blood flow that are associated with neural activity
3
Q
“functional” imaging allows us to begin to understand what
A
different parts of the brain do when we perceive, think, remember, understand, act etc…
4
Q
fMRI complements other neuroimagning methods?
A
Yes
5
Q
fMRI is non-invasive and applicable to healthy humans?
A
Yes
6
Q
To resupply oxygen and nturients, blood flow increases to an active region flooding it with fresh
A
oxygenated blood
7
Q
Whats blood-oxygenation dependent (BOLD) contrast?
A
Differences in signal on T2* weighted images as a function of the amount of deoxygenated Hb
8
Q
Fresh oxygenated blood generates a stronger ….. than stale dexoygenated blood
A
T2
9
Q
What does the BOLD contrast depend on? - (2)
A
total amount of deoxygenated hemoglobin present in brain region which in turn depends on balance between oxygen consumption and oxygen supply
the former is dependent
on neural activity and the latter is dependent on blood flow.
10
Q
If the amount of deoxygenated Hb increases locally, the …
A
BOLD signal decreases
11
Q
If the amount of deoxygenated haemoglobin decreases locally, the…
A
BOLD signal increases
12
Q
What does BOLD stand for?
A
Blood Oxygen Level Dependent
13
Q
BOLD ssignal is a key ….
A
DV in almost all fMRI research
14
Q
The BOLD signal is correlated with
A
neural activity
15
Q
Charles Coryell and Linus Pauling disocvered that the
A
haemoglobin molecule has magnetic properities that differ whether it is bounded to oxygen or not
16
Q
Oxygenated Hb is diamagnetic meaning
A
it has 0 magnetic moment
17
Q
Dexoygenated hemoglobin dHb is paramagnetic meaning
A
it has unpaired electrons and a significnat magnetic moment
18
Q
Deoxygenated and fresh oxygenated blood have different magnetic properities such that - (2)
A
- Fresh oxygenated blood has less signal decay due to T2 effects and remains brighter in T2* images as compared to deoxygenated blood and takes longer for signal to die away
- So brighter voxels in part of brain that is supplied with fresh blood and that fresh blood is BOLD signal
19
Q
What does this diagram of Thulborn’s study show? - (2)
A
- MR pulse sequences sensitive to T2* should show more MR signal where blood is highly oxygenated and less MR signal where blood is highly deoxygenated.
- At low field strength (i.e., less than 0.5 T), there was little difference between transverse relaxation values (T2) for oxygenated blood and deoxygenated blood but in higher fields (i.e., 1.5 T or greater) , their values differed significnatly.
- So stronger magnetic fields necessary for MR imagining of T2* Weighted contrast in blood.
20
Q
What does increased neuronal activity lead to? - (4)
A
- increases in blood flow and supply of oxygen that exceed oxygen demand
- As excessive oxygenated blood flows through active regions, it flushes the deoxygenated hemoglobin from the capillaries supporting the active neural tissue and from
the downstream venules. - This process is consistent with the experience of
neurosurgeons, who have long observed regions of the brain “pinking up” (due to the red color of oxygenated hemoglobin) in response to stimulation. - Lead to increased BOLD signal in area
21
Q
Increased BOLD signal following neuronal activity occurs not because ….. but because… - (2)
A
oxygenated Hb increases MR signal
but because oxy Hb displaced the deoxygenated HB that has been suppressing MR signal intensity
22
Q
Summary of BOLD signal generation - (2)
A
- (A) Under normal conditions, oxygenated hemoglobin is
converted to deoxygenated hemoglobin at a constant rate within the capillary bed. - (B) When neurons become active, however, the vascular system supplies more oxygenated hemoglobin than
is needed by the neurons through an overcompensatory increase in blood flow. The result is a decrease in the amount of deoxygenated hemoglobin and a corresponding decrease in the signal loss due to T2* effects, leading to
a brighter MR image.
23
Q
Increased BOLD signal following neuronal activity occurs due to oxy Hb displaces deoxygenated Hb that suppress MR signal intensity
As a result, increased neuronal activity increases
A
the signal of T2* images and results in positive BOLD signal
24
Q
What does this diagram show? - (5)
A
- Rat’s somtatosensory cortex is being simtulated indirectly by stimulating a neuron that activates the cortex
- As you stimulate the cortex, you can record an increase in blood velocity (speed at which blood flow increases) going from 30 to 60 arbitary units , doubling, a short period after neural activiy begins
- It declines over time but it remains steadly increased
- What happens at same time is aertial diameter is increasing so blood vessels supplying oxygen to that part of active brain are dilating to let more blood to go to that area - going from 35 to 50 arbitary units
- Rate of blood flow/velocity and aertial diameter increases when area is active to let more oxy Hb in but (Not in diagram) but blood pressure stays the same when area is active
25
How does negative BOLD signal occur? - (2)
* In some circumstances, local deoxygenated Hb accumulates due to increased oxygen consumption without a accompaning increase in blood flow
* Thus leading to negative BOLD signal
26
Summary of the BOLD responses as a seri
* negative BOLD response - inital dip- of 1-2 duration attributed to transient increase in amount of deoxygenated Hb in voxel
* After, increased neuronal activity over baseline levels results in increase inflow of oxy blood
* More oxy is supplied to area than extracted which results in decrease in oxygenated Hb within the voxel
* The fMRI BOLD signal increases above baseline at about 2s following onset of neuronal activity rising to max value at 5s after short-duration stimulus
* Maximum is known as peak of hemodynamic response
* If neuronal activity is extended (stimulus still on), peak may extended to plateau
* After neuronal activity has stopped, BOLD signal decreases in amplitude to below baseline level and remains below baseline for extended interval (undershoot phase
27
What is an example of a negative BOLD signal?
* so-called inital dip which is sometimes observed before positive BOLD signal
28
Summary of BOLD increases in active neural tissue - (2)
increased blood oxygen
higher T2* signal
29
BOLD contrast depends on the paramagnetic properties of deoxygenated hemoglobin, which causes a loss of phase coherence in nearby protons and is measurable using T
2* imaging, yet the effects of hemoglobin on nearby protons are tiny: even large BOLD effects result in signal changes
of only about 1%.
Another method for
increasing image contrast is to
use contrasts agents= paramagnetic substances that can be injected into the bloodsteam but does not cross blood brain barrier
30
What is example of some common contrast agents?
* Gadolinium dlethylenetraime -pentaacetic acid (Gd-DPTA) which is well tolerated by most people with mild headache and nusea as common side effects
31
What are some of the clinical applications of using contrast agents?
Have great importance for clinical imagning especially detection of brain tumours
32
What does this diagram show? - (3)
* Changes in BOLD activation associated with prsentation of single events of stimulus turning on
* Graph shows first example of increased BOLD response to single events
* While event-related methods are extremely common in modern fMRI, their use did not widepsread until late 1990s
33
The changes in MR signal triggered by neuronal activity is known as
* hemodynamic response or HDR
34
What does this diagram show? - (4)
* These graph represent changes in BOLD signal after some neural activity (HRF)
* Graphs of BOLD hemodynamic responses waveforms to:
* In A, single short duration event after 4-6s of stimulus
* In B, block of multiple consuective events
35
The BOLD fMRI defined measures the changes of
the total amount of deoxygenated Hb in a voxel over time
36
What is the undershoot phase due to?
biophysical and metabolic effects
37
Data from single vowel showing its changes in MR signal over time
Experiment in which pps squeezed both hands whenever there was a brief checkerboard present
Long intervals between stimulus so time for hemodynamic response to return to baseline
38
Diagram of limits of resolution in fMRI
39
What is temporal resolution?
The ability to distinguish changes in a signal or map across time
40
The limits of the temporal resolution is largely dicated by the
shape of the heodynamic response function (HRF)
41
TR is repetition time and may contribute to the
temporal resolution of an experiment
42
TR of 3 s means
* Acquiring one voleum of the brain every 3 seconds
43
What does this diagram show of limits of temporal resolution?
Consider the simple event-related design in which a subject squeezes her hand whenever she sees a visual stimulus.
To determine whether an area of the brain becomes
active due to hand motion(i.e.,detection of the active region), a relatively slow sampling rate (TR) will suffice.
- (6)
* At a 3-s TR, the hemodynamic response may be
easily identified when compared with the prestimulus baseline, but its exact shape may be difficult to estimate (A)
* Halving the TR to 1500 ms improves our estimates of the shape and timing of the hemodynamic response but does not substantially change the measured amplitude (B)
* Something very interesting becomes evident if we shorten the TR to 750 ms or even 375 ms (Figure 7.20C,D): the measured hemodynamic response, although
sampled much more often, does not change appreciably.
* If we run the scanner faster, less time to measure the BOLD signal so we see less signal compared to noise (SNR) present
* The peak and trough does not change much across as BOLD is quite a slow change
* Due to SNR, a long TR is better
44
Most modern day MRI experiments run the scanner with TR of - (2)
1 to 3 s
unless looking att specialsied region and looking at quick changes than even smaller
45
There might be a preferred temporal resolution for a given
experimental question
46
What does spatial resolution mean?
The ability to distinguish changes in an image or map across different spatial locations
47
The spatial resolution of an fMRI study depends on
many factorss
48
One straight forward factor that influences the spatial resolution of fMRI is
voxel size
49
What are voxels?
Voxels are 3-D rectangular prisims
50
Why do researchers not always use the smallest possible size of voxels to increase spatial resolution of more voxelss within brain regions improve ability to distinguish boundaries between neighbouring functional areas? - (2)
1. Variation in BOLD signal depends on change in total amount of deoxygenated Hb within voxel so if we spilt voxel in half to make smaller, BOLD signal changes will be half as large resulting in smaller signal to noise SNR ratio
2. As voxel size decreases, the time needed to acquire a given volume of brain increases = double slice time acqusition rates which can cause T2* blurring
51
What is T2* blurring?
* Disortions from T2* images that result from having a data acqusition window that is sufficiently long that significant T2* decay occurs over the interval
52
What is the impact of having voxels that are too large?
Reduce detection power and suffer from partial volume effects
53
What does this diagram demonstrating partial volume effects?
The MR signal recorded from that voxel is the sum
of signals recorded from all the different tissue types.
So, if a voxel on a T1 image contains 25% cerebrospinal
fluid (with low signal), 50% gray matter (with medium
signal), and 25% white matter (with high signal), the MR
signal recorded from the voxel will contain contributions
from all three, potentially taking an intermediate value.
54
Explain problem of partial volume effects by using voxel size too large - (8)
* Even the smallest voxel may contain multiple tissue
types, each contributing differently to the total MR signal from th
at voxel.
* Figure 7.17 shows the possible contents of a single 4 × 4 × 5 mm voxel.
* A typi-cal voxel within the brain consists mostly of neurons (and their processes, like axons) and glia, with only about 3% of the volume made up of blood vessels.
* That voxel could include a few million neurons and some tens of billions of synapses, all of which contribute to the combined metabolic demand and thus the total BOLD signal from the voxel.
* In addition to the active neurons of the interest and local capillary bed, there may be other brain tissue that does not contribute to the measured activation
* For example, voxels on the edge of the brain have white matter, gray matter and cerebrospinal fluid
* Only gray matter will contribute to the BOLD singal but protons in other tissue types may contribute to noise
* Thus, a smaller boxel that contains only gray matter activated by a task will thus provide a larger BOLD signal and contribute less to noise
55
What is partial volume effects?
The combination, within a single voxel, of signal con-
tributions from two or more distinct
tissue types or functional regions.
56
What are the other limits of spatial resolution beside partial effects? - (5)
* many brain structures including visual cortex include horiztonal and vertical cortical columns such as the ocular dominance columns in visual cortex
* In primary visual cortex neurons organised in small columns (~ 1mm) are sensitive coming to one eye
* Same pp participated in 2 fMRI sessions showed in A and C that maped relative sensitivty of visual cortex stimulation from each eye
* Note outline of areas of ocular dominance from first session correspond to results in second session
* These neurons organised in columns exist in smaller scale are difficult but not impossible to elucidate functionally using fMRI
57
Limits of spatial resolution summary - (5)
*Spatal speciticty of of blood flow changes is not precise
The blood vessels that supply brain with fresh blood spread out to large area relative to sixe of neuron
* The aertials that expand during blood flow changes are large vessels but hard to see blood flow changes less than 1 mm
* Experiments have been able to see them on the right in visual cortex
* Blood supply affects large scale compared to individual neurons
* ~1mm features, ocular dominance columns
58
Diagram of limits of spatial and temporal resolution of fMRI - (5)
* At bottom end of temporal scale limited of duration/linearity of hrf - how it evoled
* Upper end of temporal scale - limited by ability of participant to stay in scanner , can't be very still - up to 1 hour
* End of spatial scale is the spatial speciificity of the blood supply and doesn't go down below 1 mm
* At top of spatial scale, can't see changes occur bigger than whole brain, take in whole brain every 3s (TR) which is ke advantage
59
Neural activity can be recorded using different
methods which bias of what aspect of neural activity they represent
60
What is single unit recording?
A small microelectrode placed near a single neuron can record the individual action potentials from that neuron
61
What is multiple unit activity?
Larger electrode can record action potentials from a larger group of neurons
62
What are local field action potentials (LFPs)? - (2)
Coarse measure of overall activiy of brain cells recorded from electrodes some distant from cell bodies and filtered in way of seeing changes in electrical activity not jsut from AP but also incoproate postsynpatic changes in electrical signal that is correlated with integration of info
includes info of all the synaptic activity occur at brain cells even when APs do not occur
63
What is MUA and SUA concerning?
AP of neurons
64
How is local-field potentials (LFPs) recorded?
Microelectrodes record LFPs from a small population of neurons
65
Logothetis and colleagues examined the relationship between neural activity and BOLD signal by - (3)
* simultaneously recording fMRI and electrophysiological data from primary visual cortex
* Anesthetised monkeys viewed rotating checkerboard pattern while being in 4.7 T scanner
* The researchers simultaneosuly recorded single unit activity (SUA), multi unit activity (MUA) and local field potentials (LFP)
66
What are the results of Logothetis and colleagues study examining the relation between BOLD and neuronal activity? - (5)
* The relationship between BOLD activation and neuronal activity.
* Simultane-ous electrophysiological and fMRI data were recorded in monkeys during the presentation of visual stimuli that varied from 24s, 12s to 4s
* Note that the BOLD activation and the LFP activity are extended in time throughout stimulus presentation,
whereas the SUAs and MUAs spike when stimulus appears but rapidly return to baseline.
- The BOLD signal follows similar pattern of LFP activity
* These results suggest that post- synaptic activity that generates LFPs rather than APs may be a primary contributor to the BOLD response.(After Logothetis et al., 2001.)
67
A common preprocessing step of spatia smoothing of fMRI data using 3-D Gaussian filters of many voxels in width reduces
spatial resolution of fMRi
68
What is spatial smoothing?
The blurring of fMRI data across adjacent voxels to
improve the validity of statistical
testing and maximize the functional
signal-to-noise ratio at a cost of spatial resolution.
69
Although spatial smoothing reduces spatial resolution it can improve the
validity of statistical tests
70
Normalisation of transforming individual's MRI data into standardised brain reduces
spatial resolution due to difficulty in matching perosn's individual anatomy to standard template
71
If long-interval blocked design used, changes in hemodynamic response would be slower and even longer TR compared to event-related would
be needed
72
For most experimental questions, TRs of about .. is sufficient
1 to 2s
73
What does linear system of hemodynamic response mean?
A system that obeys the principles of scaling and superposition
74
In the linear system, for a given impulse the hemodynamic response is assumed to always respond in the
same manner
75
Linearity of hemodynamic response considers when
multiple stimuli presented in succession in experiments
76
For the assumption to follow that the hemodynamic response will respond in same manner in linear system it depends on 2 principles - (2)
* scaling
* superposition
77
Linearity of the BOLD signal means - (3)
* The signal has properities of a linear system i.e., scaling, superpostiion
* This is important in working out the relationship between the BOLD signal and the neural activity that caused the BOLD signal changes
* If we don't have properities of linear system then it will be harder to work out from BOLD signal changes what is happening in terms of the neurons
78
The principle of scaling states that the output of a linear system
is porportional to the magntide of its input
79
It seems in some circumstances that the
BOLD signal is approximately linear
80
The principle of scaling means in simple words that - (2)
If the input is doubled, the output is likewise doubled;
if the input is halved, so is the output.
81
The principle of scaling in terms of fMRI data will be that - (2)
this principle would predict that changes in the relative amplitude of neuronal activity should lead to similar changes in the amplitude of thehemodynamic response.
if intensity of stimulus is doubled (e.g., dark grey dot on light grey background to white dot) so will the BOLD signal
82
Whereas scaling refers to amplitude of activation, principle of superposition refers to
the timing of activation
83
Superposition simply means that the - (2)
total response to two or more events is the summation of the individual responses
If a single event generates a hemodynamic response, two events presented in succession will generate a combined response equal to two individual responses added (sum) togethe --> which can be decoded
84
Dale and Bucker investgated linearity of the hemodynamic response to individual stimulus events by .... + HYPOTHESIS - (3)
presenting clusters of one, two or three stimuli at interstimulus intervals of 2s or 5 s
The hemodynamic response to a second stimulus in a cluster was determined by subtracting the single-stimulus response from the response to a pair of stimuli. Likewise, the response to the third stimulus was determined by subtracting the two-stimulus response from the three-stimulus response.
If principle of superposition holds for fMRI BOLD data, all responses have same amplitude
85
What did Dale and Buckner found - evidence of rough linearity - (2)
* All responses had same ampltiduue and shape supporting suerposition = seen from contrast
* Suerposition producing temporal summation of signals from different events
* This suggests BOLD responses scale in roughyl inear fashion
86
Note
The (approximate) linearity of the BOLD response is a key assumption underlying
fMRI analysis using the General Linear Model (GLM) as in FSL, SPM etc
87
What does non linearity mean?
The property whereby the combined response to two or more events is not equivalent to the summation of the responses to the individual events in isolation.
88
Research investigating non linearities in fMRI hemodynamic response investigates whether there is a
refractory period following stimulus presentation during which subsequent stimuli evoke smaller hemodynamic response
89
Blocked design studies shown substantial refractory effects present after
short stimulus durations
90
Event-related designs show substantial refractory effects with short
interstimulus interval
91
What does Huttel and McCarthy event-related study presented visual checkboard stimuli in singly or pairs separated by 1 to 6 interstimulus interval - event-related
(supporting idea of non linearity of refractory periods) - (4)
* * Single stimuli evoked a robust hemodynamic re-
sponse (black line)
* Within the primary visual cortex,
the response to a second stimulus in a pair separated by 1 s was reduced by more than 40% compared with the single-stimulus response and was delayed in time by nearly 1 s.
* At short interstimulus intervals (e.g., 1 to 2 s), the hemodynam-
ic response was reduced in amplitude and increased in latency. These changesreflect the presence of refractory effects in the hemodynamic response --> don't show same response to same event
* By 6 s, both the amplitude and latency of the hemodynamic response to the second stimulus in a pair had returned to near normal values --> response similar to same event and a bit in 4 s
