Gradient Non-Linearity Correction for Arbitrary Gradient Waveform
We currently support "Topgaard style" and "NOW style" waveform files.
gnl_b_tensor.py Has the basic math of the problem of GNL for tensor valued diffusion encoding. It contains the gist of the closed-form formula for the distorted B-tensor.
proof_helper.py Contains the symbolic math computation of the earlyclosed-form derivation.
btensor.py contains the functions to compute q-vectors and B-tensor numerically and analytically and distort them with GNL.
gnl_tensor.py contains the GNL tensor computation formula.
viz.py contains quick vizualisation functions for waveform and q-vectors.
io_waveform.py Contains reading function for gradient waveforms.
vector_math.py contains functions to generate and manipulate eigenvectors in the frame of tensor generation.
tensor_math.py contains functions to manipulate tensors and generate dMRI signal.
dtd_cov.py contains a simplistic version of Westin2016 DTD covariance model.
Various (overlapping) example serves as documentation for the various functions.
example1.py loads a Topgaard waveform and numerically computes q-vector and B-tensor.
example2.py loads a NOW waveform, resample it, numerically computes q-vector and b-tensor then distort it with a GNL tensor and recompute numerically waveform, q-vector and B-tensor.
example3.py loads a Topgaard waveform, numerically computes q-vector and b-tensor then distort it with a GNL tensor and recompute numerically waveform, q-vector and B-tensor.
example4.py loads a NOW (2 files style) waveform, resample it, numerically computes q-vector and b-tensor then distort it with a GNL tensor and recompute numerically waveform, q-vector and B-tensor.
example5.py Comparaison of the close-form B-tensor distortion formula to the numerical approximation for random GNL tensor.
exampleGNL1.py Shows b-value and related metric acorss a full brain with Connectom level GNl tensor.