GLM Modelling and Statistics Flashcards

1
Q

general linear model

A

y = X * beta + error
error ~ N(0, varianz * In) (independent and identically distributed)
- y and error: 1 col, n rows = number of scans/time points
- X: n col = number of regressors, n rows = scans / time points
- beta: 1 col, n rows = number of regressors

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

OLS

A

beta^ = (X^T * X)^-1 * X^T * y
- Estimated optimal regression coefficients are equal to the covariance of the design matrix with the measured signal y divided by the covariance of the design matrix with itself

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

t-contrast

A
  • [number of regressors] x 1 contrast vector c
  • possible research question: Is there a difference between the 1st and 2nd regressor (e.g. condition)?
  • directional
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

F-contrast

A
  • [number of regressors] x [number of ORs + 1]
  • possible research question: Is there an effect of the 1st or 2nd regressor
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are the challenges for GLM when it comes to fMRI data?

A
  • low-frequency drifts
  • hemodynamic response
  • residual movement effects
  • serial auto-correlations
  • multiple comparisons
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

GLM problem - low-frequency drifts

A
  • there are slow trends in the measured fMRI signal that can make model estimation less accurate

solution:
- build a discrete cosine transform (DCT) set and regress it out of the signal (“temporal filtering”)

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

GLM problem - hemodynamic response

A
  • fMRI responses do not have rectangular shape

solution:
- stimulus functions are convolved with the hemodynamic response function (HRF) to give rise to a more realistic prediction of measured fMRI signals

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

GLM problem - residual movement effects

A
  • despite spatial realignment, there are still movement artifacts in the signal, esp. at tissue boundaries

solution:
- include realignment parameters as regressors in the design matrix of the GLM.

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

GLM problem - serial auto-correlations

A
  • consecutive fMRI scans are not statistically independent, i.e. fMRI time series have serial correlation

solution
- auto regressive model of order one [AR(1)] is fit to account for serial correlations

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

GLM problem - multiple comparisons

A
  • when running many parallel statistical tests, there will be a large number of spurious findings

solution:
- instead of controlling the false positive rate per test, control the family wise errorrate (FWE; probability of at least 1 false positive in n tests)
- for independent tests: FWE = 1 - (1 - alpha)^n
- common corrections are too conservative
- Gaussian field theory: do not correct alpha for the number of voxels but the number of resels

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

GLM - parametric modulators

A
  • sometimes, there is a trial-wise variable possibly influencing the BOLD signal (e.g. stimulus intensity)

solution:
- add a parametric modulator regressor describing possible trial-wise effects.

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

second-level GLM

A
  • usually, we want to provide evidence for activation differences at the group level (i.e. multi subject analyses)

solution:
- fit a second level model over first level contrast maps to obtain group level statistics.

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