Pulse sequence reconsideration (and fMRI techniques)

[UlrikeS91]

Offline discussion with @tgbugs and @lzehl about the general organization of pulse sequences: Start problem is that T1 and T2 are, in the physical sense not pulse sequences (e.g., https://radiopaedia.org/articles/mri-pulse-sequences-1 does not even list it under pulse sequences, and this quote "The MR signal intensity (S) produced using a spin echo sequence is proportional to the proton density (ρH), TR, T1, TE and T2, described by the equation S ∝ ρH [1-eTR/T1]e-TE/T2 4.")

Generic pulse sequences and contrast as separate parameter or have all combinations? - @tgbugs: rather split them

contrast agents (given to specimen or quality of the specimen) vs. weighting/contrast (algorithm that produces the contrasted image, hopefully making sense but theoretically could be applied to any kind of data) vs. pulse sequences (machine settings and which coil(s) were used and such, physically available data)

examples:
contrast agent: oxygenated hemoglobin vs. deoxygenated hemoglobin
pulse sequence: radio echo sequence, etc.
MRI parameter weighting: T1, T2, T2star, proton-density, diffusion-weighted?, more?

Additionally some MRI technique should be reconsidered:
functional MRI:
fMRI as superclass for BOLD fMRI (blood oxygenation level dependent), VASO fMRI (vascular space occupancy) and ASL fMRI (arterial spin labeling)
This is related to #128 (MRI as superclass with structural MRI and functional MRI)

diffusion-weighted imaging:
is a protocol and not a single technique, should probably be sMRI + diffusion-weighted + some pulse sequence; what to do about these complex processes? should the "endpoint" be registered or not?

To Do:

• split weighting and sequences
• fix fMRI techniques

To Discuss:

• solution for the last point about complex processes

also related to #57

[lzehl]

For existing instances this means the following clean up:

Technique library > T1PulseSequence >>>> MRIWeighting library > T1weighting
Technique library > T2PulseSequence >>>> MRIWeighting library > T2weighting
----> new controlledTerms schema MRIWeighting (related PR: openMetadataInitiative/openMINDS_controlledTerms#479)
----> new Terminology instance "MRI weighting"
----> move T1/T2 into new terminology
----> add T2star, proton-density to new terminology

Technique library > spinEchoPulseSequence >>>> MRIPulseSequence library > spinEchoPulseSequence
Technique library > echoPlanarPulseSequence >>>> MRIPulseSequence library > echoPlanarPulseSequence
Technique library > gradientEchoPulseSequence >>>> MRIPulseSequence library > gradientEchoPulseSequence

Additional instance list will follow shortly

[UlrikeS91]

also related: #130

[lzehl]

Additionally we should add / correct these pulse sequences (not finished yet with summary):

spin echo pulse sequence (see comment below)

• definition: An magnetic resonance imaging pulse sequence class which are all composed of a slice selective 90-degree pulse followed by one or more 180-degree refocusing pulses.
• often combined with: T1-weighting, T2-weighting, proton density weighting

inversion recovery pulse sequence (see comment below)

• definition: An magnetic resonance imaging pulse sequence class which are all composed of a 180-degree radiofrequency pulse, followed by a slice selective 90-degree pulse, followed by one or more 180-degree refocusing pulses. The time elapsed between the preparatory 180-degree radiofrequency pulse and the 90-degree excitation (slice selection) radiofrequency pulse is termed time to inversion (TI). By choosing the appropriate TI, suppression of different tissues is possible.
• often combined with: T1-weighting, T2-weighting
• children (based on TI setting):
  • short tau inversion recovery pulse sequence (TI where signal of fat is zero)
  • fluid attenuated inversion recovery pulse sequence (TI where signal of cerebrospinal fluid is zero)
  • double inversion recovery pulse sequence (two inversion pulses with different TIs)
  • phase-sensitive inversion recovery pulse sequence (increases contrast according to tissue magnetization; precise TI less important)

gradient echo sequence (see comment below)

• definition: An magnetic resonance imaging pulse sequence class which are all composed of dephasing or rephasing gradients at the end of the sequence and an additional flip angle of less than 90 degrees.

diffusion weighted sequence

• no pulse sequence should be registered
• should be registered as a fixed technique + sequences can be added

saturation recovery sequence (TBD)

echo-planar pulse sequence

• definition: A magnetic resonance image pulse sequence that is composed of multiple echoes at different phase steps (often collected blocks of 64 or 128 phase steps), which are acquired using rephasing gradients where rephasing is achieved by rapidly reversing the readout or frequency-encoding gradient.

spiral pulse sequence (TBD)

• definition: An magnetic resonance imaging pulse sequence class which are all composed of a k-space acquired in a spiral trajectory, during a single acquisition or interleaved using more than one acquisition.

fast spin echo pulse sequence

• definition: A magnetic resonance image pulse sequence that collects multiple echos instead of a single echo during a spin echo pulse sequence, multiple echos are recorded for each 90-degree pulse by using multiple 180-degree inversion pulses with slightly different phase encoding gradients.
• synonym: turbo spin echo

[UlrikeS91]

During offline meeting with @spieschnik, @Peyman-N and @tgbugs:

• Spin echo: A magnetic resonance imaging pulse sequence composed of a slice selective 90-degree pulse followed by one or more (for fast spin echo sequences) 180-degree refocusing pulses.
  • Possible parts for description: repetition time and echo time
• Inversion recovery pulse sequence: An magnetic resonance imaging pulse sequence composed of a 180-degree radiofrequency pulse, followed by a spin echo pulse sequence.
  • Possible parts for description:
    • The relative timing of the 180-degree pulse makes it possible to suppress the signal from certain tissues.
    • The time elapsed between the preparatory 180-degree radiofrequency pulse and the 90-degree excitation (slice selection) radiofrequency pulse is called the time to inversion (TI). By choosing an appropriate TI, it is possible to suppress signals from certain types of tissue.
• Gradient echo: An magnetic resonance imaging pulse sequence composed of one or more radio frequency pulses that are usually less than 90 degrees interleaved with the application of a spatial magnetic field gradient.

[UlrikeS91]

During offline meeting with @spieschnik, @Peyman-N. @lzehl and @tgbugs:

• Saturation recovery:

  • A magnetic resonance imaging pulse sequence composed of multiple slice selective 90-degree radiofrequency pulses at regular intervals delayed to allow the recovery of all the longitudinal magnetization before another pulse is applied.

• Spiral pulse sequence:

  • Not a pulse sequence but a sampling technique, i.e., seems like (almost?) all pulse sequences can be combined with (almost?) all sampling strategies
  • k-space traversal sampling techniques/strategies (https://pubmed.ncbi.nlm.nih.gov/10415232/):
    • Cartesian
    • radial
    • Zig-Zag
    • Spiral
      • Spiral in
      • Spiral out

Open question:
Add the sampling strategies as techniques or as there own terminology?