Lookup table of algorithm parameters¶
General algorithm parameters¶
| algorParam.general. | Description |
|---|---|
| isBET | Logical value of performing brain extraction in SEPIA |
| fractional_threshold | Corresponding to ‘-f’ option in FSL’s BET |
| gradient_threshold | Corresponding to ‘-g’ option in FSL’s BET |
| isInvert | Use conjugate of the complex-valued data for processing |
| isRefineBrainMask | Exclude high R2* voxels from the edge of the brain mask |
Total field recovery and phase unwrapping algorithm parameters¶
| algorParam.unwrap. | Description |
|---|---|
| echoCombMethod | Temporal phase unwrapping method |
| unwrapMethod | Spatial phase unwrapping method |
| isEddyCorrect | Logical value of performing phase correction on data acquired with bipolar readout gradient |
| excludeMaskThreshold | Threshold on relative residual map to exclude unreliable voxels |
| excludeMethod | Method of how the unreliable voxels to be excluded |
| isSaveUnwrappedEcho | Logical value of saving unwrapped phase for all echoes |
Background field removal algorithm parameters¶
| algorParam.bfr. | Description |
|---|---|
| refine_method | Method of residual B1 field removal |
| refine_order | Order of polynomial/spherical harmonic fitting |
| erode_radius | Voxel to be removed from mask edges after background field removal |
| erode_before_radius | Voxel to be removed from mask edges before background field removal |
| method | Method for background field removal |
LBV¶
| algorParam.bfr. | Description |
|---|---|
| tol | Iteration stopping criteria on the coarest grid |
| depth | No. of length scales |
| peel | No. of boundary layers to be peeled off |
PDF¶
| algorParam.bfr. | Description |
|---|---|
| tol | Iteration stopping criteria |
| iteration | Maximum no. of iterations allowed |
| padSize | Array size for zero padding |
RESHARP¶
| algorParam.bfr. | Description |
|---|---|
| radius | Radius of the spherical mean value operation |
| alpha | Regularisation parameter used in Tikhonov regularisation |
SHARP¶
| algorParam.bfr. | Description |
|---|---|
| radius | Radius of the spherical mean value operation |
| threshold | Threshold used in truncated SVD |
VSHARP¶
| algorParam.bfr. | Description |
|---|---|
| radius | Vector of radii of the spherical mean value operation in descending order |
VSHARP (STI Suite)¶
| algorParam.bfr. | Description |
|---|---|
| radius | Radius of the spherical mean value operation |
iHARPERELLA¶
| algorParam.bfr. | Description |
|---|---|
| iteration | Maximum no. of iterations allowed |
QSM dipole field invevrsion algorithm parameters¶
| algorParam.qsm. | Description |
|---|---|
| reference_tissue | Magnetic susceptibility map in ppm |
| method | Method for dipole field inversion |
TKD¶
| algorParam.qsm. | Description |
|---|---|
| threshold | Threshold on the dipole kernel |
Closed-form solution¶
| algorParam.qsm. | Description |
|---|---|
| lambda | Regularisation parameter used in L2-norm regularisation |
| optimise | Logical value of self-determination of regularisation parameter based on an L-curve approach |
NDI¶
| algorParam.qsm. | Description |
|---|---|
| tol | Iteration stopping criteria |
| maxiter | Maximum no. of iterations allowed |
| stepSize | Gradient step size of the optimizer for each iteration |
STI suite iLSQR¶
| algorParam.qsm. | Description |
|---|---|
| threshold | Threshold on the dipole kernel |
| maxiter | Maximum no. of iterations allowed |
| tol1 | Iteration stopping criteria at first level |
| tol2 | Iteration stopping criteria at second level |
iLSQR¶
| algorParam.qsm. | Description |
|---|---|
| tol | Iteration stopping criteria |
| maxiter | Maximum no. of iterations allowed |
| lambda | Regularisation parameter used in L2-norm regularisation |
| optimise | Logical value of self-determination of regularisation parameter based on an L-curve approach |
FANSI¶
| algorParam.qsm. | Description |
|---|---|
| tol | Iteration stopping criteria |
| maxiter | Maximum no. of iterations allowed |
| lambda | Gradient L1 penalty, regularisation weight |
| mu1 | Gradient consistency weight |
| mu2 | Fidelity consistency weight |
| solver | Linear or non-linear algorithm for dipole inversion |
| constraint | TV or TGV regularisation |
| gradient_mode | Method for regularisation spatially variable weight |
| isWeakHarmonic | Logical value of using weak harmonic regularisation |
| beta | Harmonic constrain weight |
| muh | Harmonic consistency weight |
Star-QSM¶
| algorParam.qsm. | Description |
|---|---|
| padsize | Array size for zero padding |
MEDI¶
| algorParam.qsm. | Description |
|---|---|
| lambda | Regularisation parameter |
| wData | Method of data weighting |
| zeropad | Array size for zero padding |
| percentage | Percentage of voxels considered to be edges |
| isSMV | Logical value of performing spherical mean value operator |
| radius | Radius of the spherical mean value operation |
| merit | Logical value of performing modal error reduction through iterative tuning |
| isLambdaCSF | Logical value of performing automatic zero reference (MEDI+0) |
| lambdaCSF | Regularisation parameter used on CSF mask |
MRI Suscep. Calc.¶
| algorParam.qsm. | Description |
|---|---|
| solver | Methods to be used for dipole field inversion |
| threshold | Threshold for TKD |
| lambda | Regularisation parameter for Tikhonov algorithms |
| toleance | tiolerance level for CG solver |
xQSM¶
| algorParam.qsm. | Description |
|---|---|
| solver | Pre-trained networks for dipole field inversion |