Diffusion Imaging Flashcards

1
Q

What is diffusion?

A

Diffusion is the random Brownian motion of molecules due to thermal processes

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

What is diffusion imaging and how does the signal link to diffusion?

A

An MRI technique that measures the mobility of molecules

It compares x, y, z diffusion gradients to a baseline T2 weighted image to look for signal loss

If the image shows mild signal loss compared to baseline, the machine assumes normal diffusion and assigns a grey voxel to the area

If all 3 gradients show no or minimal signal loss compared to baseline, then this is consistent with diffusion restriction and the machine assigns a white voxel to that area

(if the protons do not move throughout the pulse sequence i.e., diffusion restriction, the protons will be back in phase following the second gradient pulse giving off strong signal, whereas, if the protons are moving freely they will not all be in phase and therefore produce no signal)

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

What is the principle behind DWI?

A

In living tissue, molecules of water move freely through to various tissues in the body

However, in certain pathologic conditions such as the tightly packed cells of a tumour or the local swelling and pressure produced by the lack of blood flow during a stroke in the brain, creates an environment which restricts the movement of these molecules

DWI measures this restriction of diffusion

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

What is the commonly applied method for producing diffusion-weighted contrast?

A

Pulsed gradient spin echo

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

Explain the DWI sequence

A

Consists of a 90 degree pulse which flips the magnetic field perpendicular to the main field

Diffusion gradient turned on causing the protons to precess at different frequencies and out of phase

Another 180 degree RF pulse

Equal and opposite diffusion gradient pushing the hydrogen atoms back into the 90 degree perpendicular plane and back in phase giving maximum signal

Depending on whether the atom is stationary or freely moving it will change the output signal

Freely moving atoms cause more signal loss they also cause the protons to be out of phase meaning there is less signal and appear dark. Restricted protons precess in phase giving off the strongest signal so appear white

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What contributes to the signal?

A

DWI uses the same sequence as T2 but with the addition of diffusion gradients

The fundamental idea behind diffusion-weighted imaging is the attenuation of T2* signal based on how easily water molecules are able to diffuse in that region. The more easily water can diffuse (i.e. the further a water molecule can move around during the sequence) the less initial T2* signal will remain

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

How does DW contrast behave?

A

Like inverse T2 weighting
Watery tissues that have very mobile molecules give lower signal intensity whilst more static and solid tissues give a stronger signal

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

Why is DWI a T2 weighted image?

A

Due to the 2 diffusion gradients and the length of time to dephase and rephase the protons, a long TE is needed to allow for this before the signal is recorded

Any image with a long TE is considered a T2 weighted image

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

What are ADC maps?

A

To differentiate between artefact and true pathology, we use the apparent diffusion coefficient, they are independent of T2 effects and show only the effects of diffusion

Used to confirm diffusion restriction from DWI images

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

What do ADC maps represent?

A

In an ADC map, the computer analyses the diffusion in each of the x y z planes and assigns a grey scale value reflecting the apparent diffusion in each voxel of tissue

Areas of free diffusion show as bright white, whereas areas of diffusion restriction show up as darker

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

What does the signal attenuation depend on?

A

In many human tissues, the amount of signal attenuation due to diffusion varies depending on the direction of the applied diffusion gradient (x, y, z)

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

What are quantitative diffusion images?

A

As we also acquire a baseline image S(0) with no diffusion gradients, we can calculate the diffusion coefficient (D) in each gradient

These are quantitative diffusion images where each voxel represents a diffusion value and the different images represent diffusion measurements in different directions

We can average the maps across all of the directions to produce a map of the mean diffusivity

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

What is diffusion restricted by in human tissue?

A

By the complex microstructural environment

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

What is the reason for diffusion tensor imaging?

A

DWI can only tell us whether there is or isn’t any diffusion restriction

It doesn’t give us any information about the direction or magnitude of the molecular movement

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

What is a diffusion tensor?

A

In human tissue, diffusion is restricted by the complex microstructural environment, to characterise this environment, the diffusion process must be measured in many directions and is often described mathematically by a diffusion tensor

The diffusion tensor defines an ellipsoid in each voxel that characterises the diffusion properties

Represents the matrix of the xyz coordinates in any given voxel

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

How are different types of diffusion represented?

A

Isotropic diffusion = spherical
Anisotropic diffusion = ellipsoid

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Briefly describe the mathematics of diffusion tensor imaging

A

Voxels can be described in the x y and z planes by vectors which each have its own coordinate system in 3D space

This creates a 3x3 matrix of vectors (Txx, Txy, Txz/ Tyx, Tyy, Tyz/ Tzx, Tzy, Tzz) = this represents the diffusion tensor

Now the rows describe the x y and z direction - Txx, Tyy, Tzz represent the axis perpendicular to the plane and therefore, diffusion measurements are made in 6 directions

This matrix can be multiplied to create an overall vector by multiplying the matrix by a vector

The new vector is compared to the first vector to find a scalar value (is the new vector a multiple of the first vector)

If there is a scalar value then the vector is called an eigenvector and the scalar multiple is an eigenvalue

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

What do the eigenvalues and eigenvectors represent?

A

The eigenvalues and eigenvectors of the tensor indicate the principal diffusion magnitudes and orientations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

What are scalar diffusion images?

A

We can calculate different scalar images to provide a visual summary of the diffusion properties

Commonly used scalar diffusion images
1. Mean diffusivity
2. Fractional anisotropy

20
Q

What is fractional anisotropy?

A

Diffusion anisotropy describes how variable the diffusion is in different directions and is most commonly quantified via a measure known as fractional anisotropy (FA)

21
Q

How are scalar diffusion images made?

A

The diffusion coefficient (D) is calculated from the T2 and DWI images and is directionally averaged

If diffusion is calculated in multiple directions, a measure of the directional diffusion variability or ‘anisotropy’ can be calculated

22
Q

What are directional colour encoded images?

A

Usually fractional anisotropy maps with the colour indicating the principal diffusion direction

23
Q

How can DWI be applied to stroke?

A

Diffusion is initially reduced as water moves into the cells - cytotoxic oedema

Diffusion is reduced within hours of stroke onset, before it becomes visible on a T2w image.

Over time, diffusion recovers towards more normal values, eventually becoming elevated above normal due to vasogenic oedema (membranes broken down causing reduced structure)
Therefore, have to be careful when interpreting DWI images with stroke, as you have an initial fall in diffusion and then a pseudonormalisation (diffusion values are normal but the brain tissue is not normal)

24
Q

Why is DWI good for stroke detection?

A

Diffusion is reduced within hours of stroke onset before it becomes visible on a T2 weighted image

25
Q

What is the purpose of motion correction?

A

As with fMRI, motion correction can significantly improve diffusion image quality

Subject motion during fMRI scans reduces statistical significance of the activation maps and increases the prevalence of false activations

Motion correction involves using computational image registration to align all images in the diffusion dataset

26
Q

EPI images are heavily distorted, particularly at tissue boundaries, what does this distortion depend on?

A

The polarity of the phase encode gradient blips

27
Q

How do we correct for distortion?

A

Susceptibility induced distortion correction

== The difference between the positive PE blip and the negative PE blip is exploited by the FSL tool ‘topup’ which corrects for these distortions

28
Q

How can we image white matter connectivity?

A

The principal eigenvector can be plotted at each voxel to give an indication of the major white matter fibre orientation
Finding pathways through this vector field is the basis of tractography

29
Q

What is streamline tractography?

A

Starting from a seed point, voxels are connected together using a tractography algorithm
The tract propagates until termination criteria are met

30
Q

What are the termination criteria in streamline tractography?

A
  1. Angle between principal eigenvectors exceeds a certain threshold
  2. Fractional anisotropy drops below a certain value
31
Q

What is a problem with streamline tractography?

A

A streamline algorithm can be used to identify pathways within the vector field originating from a specified seed point
However, these algorithms only identify a single pathway and do not deal with branching fibres
They also give little indication of how likely the pathway is to be a true anatomical pathway

32
Q

How can we improve the issues with streamline tractography?

A

Can look at the diffusion ellipsoids to decide how likely it is to go in a particular direction

33
Q

Explain diffusion ellipsoids and profiles

A

The diffusion ellipsoid represents the r.m.s (root mean square) diffusion distance at a given snapshot in time

The diffusion profile is calculated from the diffusion tensor using d(r) = r^T Dr and represents the diffusion measurement along the direction r

34
Q

What is ODF?

A

Orientation Distribution Function

The diffusion profile can be thought of as an Orientation Distribution Function (ODF), where the function defines the relative likelihood of travelling in any given direction

35
Q

Explain orientation distribution function sampling

A

At each iteration, a sample is drawn from the ODF in each voxel
Each iteration produces a different vector field on which tractography can be run

36
Q

What is probabilistic tractography?

A

Streamline tractography is performed many times (>1000) with the ODF used to define the principal diffusion direction in each voxel for each run
The number of times a streamline passes through a voxel is counted and used to give a connection probability

37
Q

What are the limitations of the tensor model?

A

The single tensor/ellipsoid model is not good for describing voxels that contain multiple fibre populations i.e., crossing fibres

It models restricted diffusion in the plane of the crossing fibre with no preference being given to the two fibre orientations

It’s relatively simple to include more complex models into the tracking procedure, but they require more time consuming acquisitions i.e., more directions/ b values

38
Q

How do we go further than the tensor model?

A

Region of interest analysis

  1. Extract major white matter fasciculi using probabilistic tractography
  2. Region of interest masks can be generated and summary diffusion parameters calculated for the tract
  3. Correlate diffusion parameters with clinical or cognitive investigations
39
Q

What is cortical connectivity?

A

An interesting application for tractography is to find connections between cortical regions, for example those activated by fMRI or that exhibit abnormal cortical thickness
Unfortunately this is an area where streamline-based algorithms perform poorly

40
Q

What methods are good at assessing cortical connectivity?

A

Global tractography

41
Q

What is global tractography?

A

Aims to reconstruct the full brain fibre configuration that best explains the measured data based on a generative signal model

Produces a measure of connectivity from a seed point to every other point in the brain

Utilises a front propagation algorithm to identify geodesic pathways through the brain which are scored to give a connectivity value

42
Q

What is an advantage and disadvantage of global tractography?

A

It is able to penetrate much deeper into the cortex than local streamline methods
However, this also means that by capturing as much of the tract as possible, we are also increasing measurement error

43
Q

How has voxel based morphometry been applied to fractional anisotropy and what are the limitations?

A

In the early 2000s it become popular to use VBM to identify voxel-wise differences between groups based on their fractional anisotropy signal

Results from VBM analysis were found to be highly dependent on the registration - no existing non-linear registration algorithm was capable of accurately aligning the complex white matter tract structures between subjects

Therefore, many false positives were identified on VBM analysis due to misregistration

Results were also found to be highly dependent on the degree of spatial smoothing applied to the images

44
Q

What method was developed to overcome the limitations of VBM analysis?

A

Tract based spatial statistics (TBSS)

45
Q

What is the general idea of TBSS?

A

TBSS is a suite of tools for analyzing diffusion data in FSL

This software uses a tensor-fitting method to generate different measures of diffusion, such as fractional anisotropy (FA) and mean diffusivity (MD)

Once these measurements are created, you can then extract them using ROI tools like you would for fMRI data

In TBSS, we attempt to bring together the strengths of each approach. We aim to solve the alignment and smoothing issues, while being fully automated, investigating the “whole” brain — not requiring prespecification of tracts of interest

46
Q

How does TBSS avoid false positives?

A

TBSS only compares the FA in voxels at the centre of the white matter, far from the boundaries - this is referred to as the white matter skeleton - this avoids false positives

47
Q

How does TBSS work?

A

Identify a common registration target and align all subjects’ FA images to this target using nonlinear registration. At this stage, perfect alignment is not expected or required.

Create the mean of all aligned FA images and apply “thinning” (non-maximum-suppression perpendicular to the local tract structure), to create a skeletonised mean FA image. Threshold this to suppress areas of low mean FA and/or high inter-subject variability.

Project each subject’s (aligned) FA image onto the skeleton, by filling the skeleton with FA values from the nearest relevant tract centre. This is achieved, for each skeleton voxel, by searching perpendicular to the local skeleton structure for the maximum value in the subject’s FA image.

Carry out voxelwise statistics across subjects on the skeleton-space FA data.