Lecture 1 - intro Flashcards

1
Q

why fMRI

A
  1. non invasive measurement of brain activity
  2. appealing balance between temporal resolution (seconds), spatial resolution (millimeter), and coverage (whole brain)
  3. MRIs are available in many hospitals
  4. inherently interdisciplinary
  5. imaging analysis skills are highly sought after
  6. highest end technology that is continuously evolving
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

why not fMRI

A
  1. indirect measure (hemodynamic (blood) response) of neural activity since it measures metabolic demands of active neurons
    -> but hemodynamic coupling is well established
  2. limited spatial and temporal resolution (e.g., no single cells, no action potentials)
    -> but constant improvements, and who decides what the relevant scale is for understanding the brain
  3. constrained set of experiments (e.g., head motion not possible)
    -> but many experiments are feasible and data can be combined with other techniques
  4. MRI signal is noisy (low signal to noise ratios)
    -> someones noise is someones signal, constant improvements
  5. analytical challenges (e.g., autocorrelations)
    -> not unique to fmri
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

main components MRI scanner

A
  1. main magnet: creates strong magnetic field
    -> outer circle
  2. radiofrequency (RF) coil: transmits and receives radiofrequency waves
  3. gradient coils: create additional magnetic fields whose strength varies along XYZ dimensions (important for localizing the signal)
    -> inner circle
  4. patient table: moves patient in and out
  5. computer system: controls the scanner from another room
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

no net-magnetization

A

random axis
random proton phase

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

what happens during longitudinal magnetization

A

axis aligned to B0 (= main magnetic field) (vertical)

random proton phase

  1. Initially, the protons in your body are spinning on their own axes and are randomly oriented.
  2. When placed in an MRI machine, a strong magnetic field (labeled B0) causes these protons, which act like tiny magnets, to align with the field.
  3. Although they’re aligned, each proton still spins and wobbles (precession) about the direction of the magnetic field in a random phase, meaning they are not synchronized with each other.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

what happens during transverse magnetization

A

axis flipped orthogonal to B0 (= main magnetic field) (horizontal)

proton now phase aligned

protons resonate if the RF-pulse matches their precession frequency (i.e., they take on energy)

  1. An MRI machine sends a radiofrequency (RF) pulse that matches the wobbling frequency of the protons.
  2. When the RF pulse is applied, it knocks the protons out of their alignment with the magnetic field, causing them to flip into a new plane that is perpendicular to the B0 field (transverse plane).
  3. After the RF pulse, all the protons are temporarily phase-aligned, meaning they wobble in sync, which is not their natural state.
  4. This temporary alignment allows the MRI machine to detect signals from the protons as they emit energy when they attempt to return to their original alignment with the magnetic field. This emitted energy is what gives us the MRI signal that can be translated into an image.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

longitudinal relaxation (T1)

A

axis flips back into B0 direction

After RF-pulse ceases,protons emit energy in the form of RF waves that induce currents in receiver coils (That’s the MRI signal)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

transverse relaxation (T2)

A

phase coherence gets lost

After RF-pulse ceases,protons emit energy in the form of RF waves that induce currents in receiver coils (That’s the MRI signal)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

phase and precession

A

phase refers to the position of the proton spins in their precession around the main magnetic field (B0). Each proton spin precesses at a specific frequency when placed in a magnetic field, and this precession has an angular component, which is the phase. When protons are in a coherent phase, their precessional phases are aligned, leading to a stronger and more coherent signal.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

T1 weighted images

A

structural images (can distinguish different types of tissue)

high spatial resolution (~ 1mm at 3T) - low temporal resolution

high contrast, fewer artifacts

longitudinal magnitude

white matter appears lighter than gray matter

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

T2 weighted images

A

functional images

lower spatial resolution (~ 2mm at 3T) - high temporal resolution

susceptible to blood oxygenation (sensitivity of images to changes in the oxygenation levels of blood within the brain)

transverse magnitude

gray matter (and cavities) appear lighter than white matter

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

field strength

A

strength of the magnetic field used in MRI.

Higher Tesla values mean a stronger magnetic field, which can lead to clearer, higher-resolution images.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

slice

A

An individual 2D image of a specific layer or “slice” of the brain is acquired. This is like a single, flat picture of one level of the brain.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

volume

A

Several of these slices are combined to form a 3D representation of the brain called a volume

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

voxels

A

The 3D volume is divided into tiny cubes called voxels (volume pixels). Each voxel contains data for a small part of the brain, like a tiny 3D pixel.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

proton spin

A

Protons, which are components of atomic nuclei, have a property called spin. In MRI, the spin of hydrogen protons in water molecules in the body is exploited to generate images.

17
Q

precession

A
  • Precession is the motion of spinning protons as they wobble around the axis of the magnetic field. This is due to the torque of the magnetic field on the spinning protons.
  • Protons resonate if the RF-pulse frequency matches their precession frequency (i.e., they take on energy).
18
Q

radiofrequency pulse

A

An RF pulse is used in MRI to perturb the alignment of the spinning protons, causing them to move out of alignment with the B0 field and produce a signal.

After RF-pulse ceases, protons emit energy in the form of RF waves that induce currents in receiver coils (That’s the MRI signal!)

19
Q

preprocessing

A
  1. motion correction
  2. slice time correction
  3. unwarping
  4. coregistration
  5. normalization
  6. spatial smoothing
20
Q

motion correction

A

problem:
head movements shift voxels

solution:
realignment - images are rotated and moved until they align

21
Q

unwarping

A

problem:
recorded image is often distorted

solution:
map the B0 field and correct magnetic field distortions

22
Q

slice time correction

A

problem:
Since the slices are acquired sequentially, each slice is captured at a slightly different time within the TR period.

solution:
shifting signal depending on acquisition time and then resampling at interpolated time points. this becomes more important the longer the RT (repetition time).

23
Q

coregistration

A

problem:
structural and functional images need to be aligned but differ in contrast and artifacts

solution:
maximize mutual information. aligning images with an algorithm handling different contrasts (unlike motion correction)

24
Q

normalization

A

problem:
anatomical differences between subjects - coordinates are not comparable

solution:
convert images into common space (warping all images such that they align with the same template brain)

25
Q

spatial smoothing

A

problem:
weak signals & residual anatomical differences

solution:
smoothing images suppresses noise and increases statistical power. each voxel is replaced by a gaussian-weighted average of itself and its neighbours.

26
Q

BOLD signal
(Blood Oxygenation Level Dependent) signal

A

measure detected by fMRI that indicates changes in blood oxygen levels within the brain

27
Q

Hemodynamic response function (HRF)

A

a model of the blood flow changes that occur in response to neural activity

28
Q

2 types of brain imaging

A
  1. structural brain imaging: brain structure + disease/injury diagnosis
    - PET, MRI, CAT
  2. functional brain imaging: study cognitive and affective processes
    - PET, fMRI, EEG, MEG
29
Q

difference MRI and fMRI

A
  • MRI is structural, fMRI is functional
  • MRI is 1 picture, fMRI is sequence of pictures to study change over time
30
Q

spatial resolution (low to high)

A
  1. BOLD fMRI
  2. MEG & EEG + PET + ASL FMRI
31
Q

temporal resolution (low to high)

A
  1. MEG & EEG
  2. BOLD fMRI + ASL fMRI
  3. PET + ASL fMRI
32
Q

3 main goals of fMRI data analysis

A
  1. localization: determine which regions of the brain are active during a task
  2. connectivity: determine how different brain regions are connected
    –> Correlations across measured values across time
    –> functional. effective, and multivariate connectivity
  3. prediction: use a person’s brain activity to predict their perceptions, behavior, or health status
33
Q

temporal resolution

A
  • determines our ability to separate brain events in time
  • in fMRI this is determined by how quickly each individual image is acquired (TR)
34
Q

spatial resolution

A

our ability to distinguish changes in an image across different spatial locations