Introduction
Materials & Methods
Data Overview
-
T1-weighted data is acquired using an MPRAGE acquisition at 1.1 mm isotropic and resampled to DWI space (Cases1-6, Siemens Trio) or PCASL space (CasesA-B, Philips Achieva),
-
Inversion Recovery data is acquired at up to five inversion times between 500 and 5000 ms at 2.5 mm isotropic resolution (Cases1-6, Siemens Trio only),
-
Multi-echo T2 relaxometry data is acquired at roughly 21 echo times, finely sampled between 19 and 50 ms and coarsely sampled from 50 to 150 ms at 2.5 mm isotropic resolution (Cases1-6, Siemens Trio only),
-
Refocussed T2 relaxometry is acquired with an echo time of 12 ms for 32 echoes with a TR of 9 s. Resolution is 0.42 × 0.42 × 3mm (Case C, Philips Achieva only),
-
Diffusion Weighted MRI is acquired on 3 shells, 8 directions at b = 300 s.mm −2 , 32 directions at b = 700 s.mm −2 and 72 directions at b = 2000 s.mm −2 with 12 b = 0 volumes. Resolution is 2.5 mm isotropic (Cases1-6, Siemens Trio only),
-
PCASL data is acquired for 30 control-label pairs using a 2D EPI read-out with Label Duration of 1650 ms and Post-Labelling delay of 1800 ms. Resolution is 2.5 × 2.5 × 6 mm (CasesA-B, Philips Achieva only),
-
Pulsed ASL data is acquired for 5 averages using a 3D GraSE read-out with QUIPSSII pulse time and inversion time of 800 ms and 2000 ms respectively at 2.5 mm isotropic resolution, resampled into DWI space (Cases1-6, Siemens Trio only),
Package Overview
-
Arterial Spin Labeled MRI (fit_asl),
-
Single and multi-component T1 relaxometry (fit_qt1),
-
Single and multi-component T2 relaxometry (fit_qt2),
-
Diffusion Weighted MRI (fit_dwi),
-
Diffusion Tensor manipulation routines (fit_tools),
-
Basic image maths and manipulation routines (fit_maths).
Underlying Input/Output Framework
-
-source; the input data which expects a 4D nifti file organised with the dependent variable (e.g. time, echo time or diffusion weighting) along the fourth dimension,
-
-mask; a 3D mask file (optional, but recommended),
-
-init; a 4D initialisation parameter file in which the parameters are organised along the fourth dimension in the same order as the NiftyFit parameter output. These provide an initialisation for non-linear least squares. Currently parameter initialisation only applies to non-linear fitting routines,
-
-slice; select a single slice to run the model-fitting on,
-
-voxel; select a single voxel to run the model-fitting on,
-
Input and output help text is displayed when running each command with no inputs.
-
-mcmap; a multi-parameter map with parameters organised along the 4th dimension. This file contains all the information required to build a synthetic version of the data (when using the input experimental variables),
-
-resmap; a 3D volume of the per-voxel model-fit residuals,
-
-error; a 4D volume of the independent, identically distributed (I.I.D.) parameter errors organised with variances followed by covariances,
-
-syn; a 4D volume of data simulated from the fitted parameters and input variables.
Parameter Fitting Routines
Example Applications and Case Studies
Example Applications
Single Inversion Time Arterial Spin Labeled MRI
Single and Multi-component T1 Relaxometry
Single and Multi-component T2 Relaxometry
Diffusion Weighted MRI
Example Application: Modified NODDI Fitting
Example Application: g-Ratio Estimation in Adult Controls
Case Studies
Single Inversion Time Arterial Spin Labeled MRI
Single Component T1 Relaxometry
Single and Multi-component T2 Relaxometry
Diffusion Weighted Imaging
Modified Diffusion Weighted Imaging
g-ratio Estimation
Volume | Independent fitting | Coupled fitting | |||||||
---|---|---|---|---|---|---|---|---|---|
Case | (mm3) | FA | T2 (ms) | Vin
| Vmwf
| g-ratio | Vin
| Vmwf
| g-ratio |
1 | 495,583 | 0.423 | 71.808 | 0.514 | 0.141 | 0.864 | 0.526 | 0.176 | 0.887 |
2 | 422,227 | 0.439 | 70.021 | 0.498 | 0.137 | 0.869 | 0.509 | 0.134 | 0.868 |
3 | 390,378 | 0.431 | 70.712 | 0.518 | 0.147 | 0.862 | 0.530 | 0.143 | 0.855 |
4 | 433,189 | 0.413 | 70.115 | 0.516 | 0.159 | 0.848 | 0.530 | 0.155 | 0.841 |
5 | 464,354 | 0.409 | 66.101 | 0.502 | 0.162 | 0.845 | 0.515 | 0.158 | 0.841 |
6 | 431,296 | 0.351 | 69.848 | 0.470 | 0.163 | 0.835 | 0.482 | 0.159 | 0.831 |