GRE Sequences
There are so many different GRE sequences. Can you make sense of these?
|
|
|
Gradient echo (GRE) sequences begin as a regularly spaced series of radiofrequency (RF)-pulses repeated at time interval TR. These RF-pulses are often chosen to have flip angles (α) less than 90°, but can have any value.
|
A train of evenly spaced RF pulses produce "FID-like" and "Echo-like" signals.
|
Each RF-pulse generates a free induction decay (FID) signal immediately after it is applied. The FID decays with time constant T2* that reflects both natural/true/intrinsic T2-relaxation plus accelerated signal loss due to magnetic field inhomogeneities.
Each pair of RF-pulses generates a spin echo (SE), that because of regular spacing of the RF-pulse train, forms and rises to a maximum exactly a the time of the next RF-pulse. Each set of three RF-pulses also generate stimulated echoes (STEs) that coincide with the spin echoes. The "Echo" portion of the GRE sequence is mostly T2-weighted with fewer T2* effects.
Each pair of RF-pulses generates a spin echo (SE), that because of regular spacing of the RF-pulse train, forms and rises to a maximum exactly a the time of the next RF-pulse. Each set of three RF-pulses also generate stimulated echoes (STEs) that coincide with the spin echoes. The "Echo" portion of the GRE sequence is mostly T2-weighted with fewer T2* effects.
After many RF-pulses the two signal components reach a steady-state and begin to lose their original identities as discrete FIDs and SE/STEs. In fact, both result from the summation of echoes and magnetization created from multiple prior cycles For this reason some authors prefer to denote them as S+ and S-, or use more generic terms such as "post-excitation" and "pre-excitation" signals, respectively. For the purposes of this web site I will stay with the older nomenclature, continuing to refer to them as "FIDs" and "Echoes". Perhaps "FID-like" and "Echo-like" would be more appropriate.
|
An interesting phenomenon occurs within an RF-pulse train if TR is made much shorter than T2. Here the "FID" following each RF-pulse does not completely die out before the "Echo" component begins to form. In this situation the transverse signal never completely disappears and the result is called a steady-state free precession (SSFP).
|
A steady-state free precession exists when TR is short, the tails of the "FIDs" merging with the rising edges of the "Echoes".
|
Creation of a GRE by manipulation of spin phases by an external magnetic gradient field. The gradient is typically applied with two polarities, for dephasing and rephasing spins respectively.
|
Either of the signals generated by the RF-pulse train ("FIDs" or "Echoes") can be temporarily suppressed and then made to reappear at a chosen time (TE) by application of an external magnetic gradient field. The gradient is typically applied in two steps: 1) a dephase portion that forces spins out of phase, and 2) a rephase portion that brings them back into phase as the GRE.
|
By changing the timing, strength, and duration of the gradients it is possible to refocus either the "FID", the "Echo", or both at a specified times as the GRE. This is possible because the dephasing gradient field simply "scrambled" the spin phases in a predictable, organized fashion. By reversing this gradient the signal can be regenerated (except for non-gradient-related T2* decay). This gives rise to the alternative translation of the GRE acronym, "gradient recalled echo."
A first-level classification of GRE sequences based on which signal ("FID", "Echo", or both) is sampled can now be made:
- GRE with "FID" refocusing (FISP, GRASS, FFE)
- GRE with "Echo" refocusing (PSIF, SSFP, T2-FFE)
- GRE with combined "FID" and "Echo" refocusing (True FISP, FIESTA, Balanced FFE)
By carefully structuring of the the gradients and phases of the RF-pulses is is possible to either: 1) nurture/preserve these transverse coherences, or 2) disrupt or prevent them from forming. This leads to a second classification scheme for GRE sequences based on how transverse coherences are handled:
- Coherent GRE sequences: preservation of transverse coherences
- Spoiled GRE sequences: complete disruption of transverse coherences
References
Chavhan GB, Babyn PS, Jankharia BG et al. Steady-state MR imaging sequences: physics, classification, and clinical applications. Radiographics 2008;28:1147-1160. [DOI Link]
Elster AD. Gradient echo imaging: techniques and acronyms. Radiology 1993; 186:1-8. (My older review; still accurate, though some vendors have gone out of business. Gives a good history of the development of GRE sequences). [DOI Link]
Hargreaves B. Rapid gradient-echo imaging. J Mag Reson Imaging 2012;36:1300-1313. (A great, but not overly technical, modern review). [DOI Link]
Markl M, Leupold J. Gradient echo imaging. J Magn Reson Imaging 2012; 35:1274-1289. [DOI Link] (good review)
Chavhan GB, Babyn PS, Jankharia BG et al. Steady-state MR imaging sequences: physics, classification, and clinical applications. Radiographics 2008;28:1147-1160. [DOI Link]
Elster AD. Gradient echo imaging: techniques and acronyms. Radiology 1993; 186:1-8. (My older review; still accurate, though some vendors have gone out of business. Gives a good history of the development of GRE sequences). [DOI Link]
Hargreaves B. Rapid gradient-echo imaging. J Mag Reson Imaging 2012;36:1300-1313. (A great, but not overly technical, modern review). [DOI Link]
Markl M, Leupold J. Gradient echo imaging. J Magn Reson Imaging 2012; 35:1274-1289. [DOI Link] (good review)
Related Questions
It seems as if every manufacturer has adopted a different name for their gradient echo sequences. Why is this? Can you sort this out for me?
If a spin echo results from 2 pulses, and a stimulated echo from 3 pulses, what do you get from 4 pulses?
It seems as if every manufacturer has adopted a different name for their gradient echo sequences. Why is this? Can you sort this out for me?
If a spin echo results from 2 pulses, and a stimulated echo from 3 pulses, what do you get from 4 pulses?