Wrapper to run the AMICO implementation of NODDI on the PMACS cluster
DWI data in the format /path/to/dwi.[nii.gz, bvec, bval] Brain mask (.nii.gz)
The model is explained in detail in the original NODDI paper. Briefly, given
- A_t - normalized DWI signal (the data we're fitting to, normalized by the average b=0 signal)
NODDI fits
A_t = V_{iso} * A_iso + (1 - V_{iso}) * [ V_{icv} * A_{icv} + (1 - V_{icv}) * A_{ecv} ]
where
- A_{iso} - model of the isotropic signal
- A_{icv} - model of the intracellular signal
- A_{ecv} - model of the extracellular signal
- V_{icv} - intracellular volume fraction (output as FITxICVF)
- V_{iso} - isotropic volume fraction (output as FITxISOVF)
The other outputs discussed below, FITxOD and FITxdir, are additional parameters used to model A_{icv} and A_{ecv}.
NODDI metrics computed via AMICO, and a pickle file produced by AMICO:
FITxICVF.nii.gz - Relative intracellular volume fraction. Also called neurite density. It reflects the fraction of the non-isotropic signal that is estimated to be from intracellular diffusion inside neurites.
FITxOD.nii.gz - Neurite orientation dispersion about the estimated principal axis. Normalized to be between 0 (parallel fibers) and 1 (random orientation).
FITxISOVF.nii.gz - Isotropic volume fraction. This is the estimated fraction of the voxel that is occupied by CSF.
FITxdir.nii.gz - Vector image containing the estimated principal neurite axis in each voxel.
Accelerated Microstructure Imaging via Convex Optimization (AMICO) from diffusion MRI data Alessandro Daducci, Erick Canales-Rodriguez, Hui Zhang, Tim Dyrby, Daniel C Alexander, Jean-Philippe Thiran NeuroImage 105, pp. 32-44 (2015)
NODDI: practical in vivo neurite orientation dispersion and density imaging of the human brain Hui Zhang, Torben Schneider, Claudia A Wheeler-Kingshott, Daniel C Alexander NeuroImage. 16;61(4):1000-16 (2012)