qMRLab logoQuantitative MRI mOOC

B0 Mapping · Exercise 5-2

Phase unwrapping

In this exercise, we will cover:

  1. The steps from phase images to a B0 map: echo times, masking and unwrapping
  2. When unwrapping is needed, and why masking matters
  3. Unwrapping each echo’s phase, or the phase difference

Exercise 5-2 · (a)

From phase to B0: single-slice dual-echo dataset

Select the settings that produce a B0 map you are satisfied with.

Input

Magnitude
Phase, echo 1
Phase, echo 2
Echo times
Unwrap
Mask

Output

B0 map

Exercise 5-2 · (b)

From phase to B0: QSM challenge dataset

Select the settings that produce a B0 map you are satisfied with.

Input

Magnitude
Phase, echo 1
Phase, echo 2
Echo times
Unwrap
Mask

Output

B0 map

Data: Chan K-S, Marques JP, Spincemaille P, Wang Y. QSM consensus example data (Siemens). Zenodo, doi:10.5281/zenodo.7795492, CC BY 4.0; echoes 1 and 2, averaged to 2 mm, every 4th slice.

Exercise 5-2 · Summary

Summary

  1. You can unwrap either each echo's phase map, then subtract, or the phase difference Δφ = φ2 − φ1.
  2. Unwrapping might not be necessary: Δφ does not wrap as long as the field offset stays within ±1/(2ΔTE), that is when
    ΔTE [s] < 1 / (2 γ̄ ΔB0 [Hz])
    where ΔB0 is the largest field offset expected.
  3. Near air–tissue interfaces (sinuses, ear canals), the field can be strong enough that phase wraps are unavoidable, even with a short ΔTE.
qMRLab Shimming Toolbox

To compute B0 maps from your own data, you can use either qMRLab or Shimming Toolbox, open-source tools developed at NeuroPoly.

NeuroPoly