Tutorial-advanced-fMRI

Brad Macintosh

Informations

• Publié par: Neseng
• Nombre de lectures: 14
• Langue: English

Extrait

Tutorial: Advanced Functional MRI

Brad MacIntosh (

[e-mail]

) and Stuart Clare (

[e-mail]

)

The purpose of this tutorial is to provide an understanding of the issues associated with conducting

an fMRI study at different magnetic field strengths.

You will not need to analyze images during this tutorial as has happened in the past, but instead

rely on fMRI papers in the literature, the concepts introduced during the lecture and your

imagination on how to design a knock-out fMRI study.

Scenario 1:

You are interested in comparing activation patterns in

young

and

old

individuals. With access to a

1.5 Tesla

MRI scanner that has the latest capabilities for fMRI using BOLD contrast you would like

to design a visual-motor reaction time fMRI study to compare these two cohorts.

1.1. Give two reasons why you would expect to detect different BOLD hemodynamic response

functions between these two groups.

1.2. In light of 1.1, why have you decided to make the total length of the fMRI experiment for the

young and old cohort different?

1.3. Approximately how many trials would be required in the older cohort, where the fMRI time

series noise is higher, to ensure that you detect an equivalent number of supra-threshold voxels?

Note:

A fellow student has emailed you the following paper and asks you what you think of the

study. Use this paper to help with Scenario 1.

Mark DʼEsposito, et al., Neuroimage: “The Effect of Normal Aging on the Coupling of Neural

Activity to the Bold Hemodynamic Response”, 1999.

Scenario 2:

You are a graduate student in a laboratory that has worked on pioneering Arterial Spin Labeling

(letʼs say UPENN or perhaps FMRIB Physics group). You have been approached by a colleague to

come up with a paradigm that is capable of measuring dynamic processes of motor learning. At

your disposal is a

3 Tesla

MRI system with continuous Arterial Spin Labeling (CASL) capabilities.

The experiment involves having young, healthy control participants performing a series of numeric

sequences on an fMRI-compatible mobile phone. Your colleague wants to see how long it takes

participants to memorize mobile phone numbers. They tell you that it is very important that

participants do the task continuously – so as to say that there is no rest condition.

2.1. Explain why you have elected to use the CASL sequence in this dynamic learning study

compared to a conventional BOLD fMRI sequence.

2.2. List the critical imaging parameters that will be used in the study (i.e. repetition time, TR, echo

time, TE, post-labeling delay time, p).

Note:

Your colleague hands you a recent paper that they think is pretty interesting:

Ingrid R Olson, et al., Brain and Cognition: “Using perfusion fMRI to measure continuous changes

in neural activity with learning”, 2006.

Scenario 3:

High field fMRI is increasing in availability and provides new opportunities to study the brain and

you are fortunate to work at a research institute that has MRI scanners at three different field

strengths (e.g. 1.5, 4.0 and 7.0 Tesla or 1.5, 3.0 and 7.0 Tesla). You want to do some calculations

to help you prepare your experiments.

3.1. Search the literature for the R

2

*

and

Δ

R

2

*

values for gray matter at the three field strengths. If

you cannot find exact numbers in the literature, then make an educated guess.

3.2. Assuming all issues associated the signal-to-noise ratio (SNR) BOLD acquisition remain the

same across the three field strengths (i.e. the same RF profiles, bandwidth, etc.):

a) What is the temporal gain in SNR when you perform your 3 T fMRI experiment with in plane

resolution of 3 x 3 mm

2

at 7 T?

b) What is the image gain in SNR when you perform your 3 T fMRI experiment with in plane

resolution of 3 x 3 mm

2

at 7 T?

c) You performed a high resolution 7 T experiment with in place resolution of 1 x 1 mm

2

. How much

longer would you need to scan to achieve the same temporal SNR if you then did the experiment at

1.5 T?

Note:

You know from experience that the folks at the University of Minnesota have pioneered a lot

of the high field fMRI. A quick pubmed search reveals a relevant paper by Yacoub et al., “Imaging

brain function in humans at 7 Tesla” Magnetic Resonance in Medicine. 2001 Apr;45(4):588-94.

And of course there is a recent impressive paper by Triantafyllou et al in Neuroimage 2005;

26:243-250 “Comparison of physiological noise at 1.5 T, 3 T and 7 T and optimization of fMRI

acquisition parameters”.