FYS-KJM 4740

MR LAB Place:The Intervention Centre, Rikshospitalet.

  1. Get to know the MR scanner. Security aspects.
  1. Scanning ofphantom

A. Aim: estimate T1relaxation times using two different methods

  1. 2D (single slice) acquisition of phantom using a T1-GRE sequence(spoiled GRE) (TR/TE = 20 ms/2 ms). Change flip angle from 5 deg to 90 deg (keeping all other parameters constant).Measure signal intensity (SI) in each compartment of the phantom (see arrows on image below) as function of flip angle. From the resulting SI vs flip angle curves, estimated the T1-relaxation time in each compartment (hint: estimated max SI from the curves).
  2. Find expression for T1 as function of SI, flip angleand TR measured at TWO different flip angles (hint, use ratio of SI; assume TR/T1 < 1). Use this expression to estimate T1in the different compartments (hint: use two flip angles <= ernst-angle).
  3. Repeat 2D (single slice) acquisition of the phantom, but this time use an inversion recovery sequence (IR-TFE). Vary the inversion times (TI) from 50 ms to 3000 ms. Measure SI in every compartmentas function of TI. What is the expression for SI in an IR sequence? What is the relationship between TI and T1 at SI=0.Based on the SI vs TI curves– estimate T1-in each compartment.Compare to the values obtained using method (I).
  4. Comment on sources of error for the two methods (hint slice-excitation profile).

B. Aim: understand and avoid fold-over artifacts.

  1. In exercise A, field of view (FoV) was equal in both directions (220 x 220 mm). Repeat the T1-GRE acquisition (flip = 30 deg). Reduce FoV by 50% in the phase encoding direction. Turn off ’fold’over suppression’. What happens to the resulting image? Explain.
  2. Reset FoV to 220 mm in the phase encoding direction and reduce by 50% in the frequency direction. Why are there no artifacts in the resulting image?
  3. Increase number of signal averages (NSA) to 4 (from 1). What is the effect in the image? (measure change in signal-noise ration SNR). Explain.
  4. Reduce FoV by 50% in the phase encoding direction, but this time turn ON the option: ’fold-over-suppression’ (keep NSA = 4). Explain how fold-over artifacts can be avoided even when the FoV in the phase encoding direction is less than the object size.
  5. In the compendium it says that’fold’over suppression’ leads to increased acquisition time (why?) The option’fold-over-suppression’ does, however, NOT affect the acquisition time (or the NSA) in our scan when we used NSA=4. Why? (hint NSA > 1; time for more NSA vs more phase profiles).

C. Aim: Understand ’Gibbs-ringing’ artifacts and chemical shift parameters

  1. Run same sequence as in B (with full FoV) but increase pixel size to 2 mm x 2 mm (from 1mm x 1 mm). Which artifacts are now visible in areas with large contrast differences (edges). Explain.
  2. The Philips scanner has an option called’water-fat shift’ which is given in pixels and defines how much a pixel made of just fat will be displaced (rel to true position) . If we modify the water-fat shift’ parameter we observe that the bandwidth (in Hz per pixel) also changes . Why? What is the relationship between water-fat shift , pixel size and bandwidth as function of δB (difference in Bo between water and fat).
  3. Repeat the scan in (I) but change water-fat shift so that the bandwidth is doubled What is the effect on the image quality (measure change in SNR). Explain..

Scanning of volunteer

  1. Fast SE (FSE). Axial head scan

a)Use ’echo train length’ (ETL) =13 and (effective) TE = 100 ms. What is the echo spacing (ES) if you assume a linear profile ordering? With a T2-time of 30 ms and a pixel size of 1 mm , what is the FWHH of the point spread function (PSF) with ETL = 13? What happens to the effective TE and PSF if ETL increases to 128 (unchanged ES and profile ordering)? How can you achieve shorter effective TE without changing ES and ETL?

  1. 3D-GRE (T1-FFE). Sagittal head scan.

b)Explain the difference between a 2D and a 3D acquisition.

c)What is the disadvantage of using axial 3D scan in the head? (hint over-folding).

d)Based on the first scan, how can we reduce the scan-time by 30 % without changing the resolution?Explain the different methods. Which parameter affects SNR and which affect the image contrast?

e)In 3D acquisitions, a preparation pulse is often used to increase the contrast or remove unwanted signal. If T1 for white matter is 700 ms and 1000 ms for grey matter, what inversion time (TI) should be used to optimize the contrast (SI1-SI2) between grey- and white matter? (use the signal equation for the IR sequence).