|
|
|
| e-Pub |
Previous | 
Home
Section: New Software and Platforms
Contributions to the open source dMRI platform DIPY
Participants : Demian Wassermann, Rutger Fick.
DIPY (Diffusion Imaging in Python) is a fast growing open source platform for dMRI image processing. It aims to be a reference implementation platform for most dMRI processing technologies and it has several contributors around the world including Stanford Universtiy, USA; Berkeley University, USA; Sherbrooke University, Canada; and University of Cambridge, UK. This aims to provide a dMRI library easy to use in research-intensive cases where developments of new technologies are simpler than in high performance C++ libraries.
In 2014 D. Wassermann and R. Fick got involved in this open source platform. Their work spans from minor public extensions to private developments within this framework. They developed an improved implementation of the 3D-SHORE [72] basis, which is designed to reconstruct the three-dimensional diffusion propagator from three-dimensional q-space measurements. Moreover, they optimized the computation of the basis coefficients and introduced the analytical Laplacian regularization [19] . They also implemented the MAP-MRI basis [73] , which is an extension of the 3D-SHORE basis to better deal with highly anisotropic data. Finally, they extended this work by again introducing the analytical Laplacian regularization. Also, we implemented a novel generalized basis that fits diffusion MRI data over both three-dimensional q-space and diffusion times (3D+t). The theoretical developments related to these two last contributions have been submitted to ISBI 2015 and IPMI 2015 respectively.