POAS4SPM: A Toolbox for SPM to Denoise Diffusion MRI Data
- PDF / 1,426,474 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 29 Downloads / 183 Views
SOFTWARE ORIGINAL ARTICLE
POAS4SPM: A Toolbox for SPM to Denoise Diffusion MRI Data Karsten Tabelow & Siawoosh Mohammadi & Nikolaus Weiskopf & Jörg Polzehl
Published online: 5 July 2014 # The Author(s) 2014. This article is published with open access at Springerlink.com
Abstract We present an implementation of a recently developed noise reduction algorithm for dMRI data, called multishell position orientation adaptive smoothing (msPOAS), as a toolbox for SPM. The method intrinsically adapts to the structures of different size and shape in dMRI and hence avoids blurring typically observed in non-adaptive smoothing. We give examples for the usage of the toolbox and explain the determination of experiment-dependent parameters for an optimal performance of msPOAS. Keywords dMRI . Noise reduction . POAS . msPOAS . SPM (RRID:nif-0000-00343)
Introduction Diffusion-weighted magnetic resonance imaging (dMRI) has developed into an extremely versatile tool for the in-vivo structural analysis of tissue, for example in the human brain (Johansen-Berg and Behrens 2009). One reason is that the diffusion signal obtained with the pulsed gradient spin Karsten Tabelow and Siawoosh Mohammadi contributed equally to this paper. K. Tabelow (*) : J. Polzehl WIAS Berlin, Mohrenstr. 39, 10117 Berlin, Germany e-mail: [email protected] J. Polzehl e-mail: [email protected] S. Mohammadi : N. Weiskopf Wellcome Trust Centre for Neuroimaging, UCL Institute of Neurology, 12 Queen Square, London WC1N 3BG, UK S. Mohammadi e-mail: [email protected] N. Weiskopf e-mail: [email protected]
sequence echo (PGSE, Stejskal and Tanner 1965) directly relates, via three-dimensional Fourier transform, to the diffusion propagator which is the probability density of the underlying Random Walk the spin particles experience (Mitra and Sen 1992). Therefore, if we measured the diffusion signal for all possible diffusion gradient directions, times and strengths, i.e. cover the whole q-space, we would know the full propagator. Its spatial and directional dependence would allow us to infer on boundaries for the diffusing particles and hence the underlying structure. However, in practice, only a limited coverage of the q-space is feasible. Therefore, a number of models have been developed in the past, which reveal at least partial information contained in the diffusion propagator. Most require dMRI data measured on at least one q-shell, that is characterized by a single b-value (Basser et al. 1994b) subsuming diffusion gradient strength and diffusion time. The most prominent example of a diffusion model gives rise to Diffusion Tensor Imaging (DTI, Basser et al. 1994a, b). Surprisingly, although this model actually describes free diffusion in anisotropic media it has proven to relate well to the underlying tissue geometry in the brain in general, and to the main fiber directions in the white matter in particular (Johansen-Berg and Behrens 2009). More sophisticated descriptions of the diffusion signal have been examined to infer on mor
Data Loading...
