Dmitry S. Novikov¹, Ileana O. Jelescu¹, and Els Fieremans^1
1Center for Biomedical Imaging, Department of Radiology, NYU School of Medicine, New York, NY, United States

We present exact relations between the measured diffusion signal moments, and the microstructural parameters of neurites (internal and external diffusivities, volume fraction), as well as coefficients of the neurite orientation distribution function in the spherical harmonics basis, up to arbitrary order. Our solution eliminates the need for notoriously unstable nonlinear fitting, and allows one to determine all microstructural parameters using signal moments or cumulants, which are found using linear b-matrix pseudoinversion in the low-b acquisition regime. We apply our framework up to the 6th order to the human connectome data.

Susie Y. Huang¹, Thomas Witzel¹, Qiuyun Fan¹, Jennifer A. McNab², Lawrence L. Wald^1,3, and Aapo Nummenmaa^1
1Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Charlestown, MA, United States,²Radiological Sciences Laboratory, Department of Radiology, Stanford University, Stanford, CA, United States, ³Harvard-MIT Division of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, United States

The characterization of white matter pathways by diffusion tractography could be improved by incorporating quantitative information on axonal size provided by axon diameter mapping methods. We demonstrate axon diameter estimation across white matter tracts of arbitrary orientation in the in vivo human brain using gradient strengths up to 300 mT/m. TractCaliber offers consistent axon diameter estimates across the corpus callosum and corticospinal tracts that agree with histological observations, thereby supporting the use of high gradient strengths for accurate in vivo estimation of axon diameters across white matter tracts.

Darya Morozov¹, Leah Bar¹, Nir Sochen¹, and Yoram Cohen^1
1The Raymond and Beverly Sackler Faculty of Exact Science, Tel-Aviv University, Tel-Aviv Yaffo, Tel-Aviv Yaffo, Israel

Water diffusion in neuronal tissues, at high diffusion weighting and sufficient long diffusion times, becomes non-Gaussian. Under such conditions conventional DWI and DTI data does not accurately describe water diffusion in the CNS. However when complex biological specimens are studied it is of value to test the new methodology on real, complex samples where the ground truth is known a priori. In the present work we modeled the signal from single (s-) and angular double-pulsed-field-gradient (d-PFG) diffusion NMR experiments performed on a series of well-defined phantoms of increasing complexity and on isolated pig optic nerves.

Filip Szczepankiewicz¹, Danielle van Westen^2,3, Jimmy Lätt², Elisabet Englund³, Carl-Fredrik Westin⁴, Freddy Ståhlberg^1,3, Pia C. Sundgren^2,3, and Markus Nilsson^5
1Dept. of Medical Radiation Physics, Lund University, Lund, Sweden, ²Imaging and Function, Skåne University Healthcare, Lund, Sweden, ³Dept. of Clinical Sciences, Lund University, Skåne University Healthcare, Lund, Sweden, ⁴Dept. of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, United States, ⁵Lund University Bioimaging Center, Lund University, Lund, Sweden

In this work we separate diffusional kurtosis into components rendered by isotropic and anisotropic microstructural features, by combining conventional and single-shot isotropic diffusion encoding. We show that glioma and meningioma tumors exhibit two radically different origins of kurtosity in vivo. This indicates that the gliomas and meningiomas contain isotropic domains with varying diffusivity, and randomly oriented anisotropic domains, respectively. Finally, we conclude that disentangling the origins of diffusional kurtosis improves the sensitivity and specificity of kurtosis parameters as well as the ability to interpret such parameters.

Chantal M.W. Tax¹, Dmitry S. Novikov², Eleftherios Garyfallidis³, Max A. Viergever¹, Maxime Descoteaux³, and Alexander Leemans^1
1Image Sciences Institute, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, ²Center for Biomedical Imaging, New York University School of Medicine, New York, New York, United States, ³Sherbrooke Connectivity Imaging Lab, Université de Sherbrooke, Sherbrooke, Quebec, Canada

In this work we localize single fiber population (SFP) voxels in the brain by recursively excluding crossing fiber voxels with constrained spherical deconvolution in multi-shell diffusion MRI data. We investigate the average signal decay over all SFPs parallel and perpendicular to the fiber, and extract (microstructural) diffusion indices by fitting various models. By doing so, we have characterized SFPs throughout the brain and in different tracts, which can potentially be used as reference for healthy brain in group studies or simulation studies.

Emmanuel Vallée¹, Gwenaëlle Douaud¹, Andreas U Monsch², Achim Gass³, Wenchuan Wu¹, Stephen M Smith¹, and Saad Jbabdi^1
1FMRIB, University of Oxford, Oxford, Oxfordshire, United Kingdom, ²Memory Clinic, University Center for Medicine of Aging Basel, Basel, Switzerland,³Department of Neurology, University Hospital Mannheim, Heidelberg, Germany

The nature and resolution of diffusion MRI make it very sensitive to partial volume effects, thus affecting the indices derived from the diffusion tensor. We propose here a novel way of fitting a two compartments model that does not require spatial regularization, and aims to retrieve the unbiased diffusion tensor. We evaluate our method with simulations and in-vivo Inversion Recovery data where Free Water signal was nulled. Finally, we assess how the new model impacts the results of a clinical study on Alzheimer’s disease.

Sjoerd B Vos¹, Andrew Melbourne¹, Hui Zhang², John S Duncan³, and Sebastien Ourselin^1
1Translational Imaging Group, University College London, London, United Kingdom, ²Centre for Medical Image Computing, University College London, London, United Kingdom, ³Department of Clinical and Experimental Epilepsy, Institute of Neurology, University College London, London, United Kingdom

Perfusion in white matter contributes to diffusion-weighted signal attenuation that is not modelled in most diffusion MRI based microstructural tissue models. This effect is present at low b-values (<300 s/mm2) and microstructural tissue parameters are influenced by this. Fitting these models, in this case NODDI, using only data acquired at b-values > 300 s/mm2 removes this effect. Comparisons of NODDI fitting using multi-shell acquisitions with (b=0,300,700,2500) or without (b=300,700,2500) data in the low-b regime show difference in both the isotropic and intracellular volume fractions. Correlations of NODDI parameters with IVIM-derived perfusion data confirm the effects originate, in part, from perfusion.

Enrico Kaden¹, Nathaniel D Kelm², Robert P Carson³, Mark D Does², and Daniel C Alexander^1
1Centre for Medical Image Computing, University College London, London, United Kingdom, ²Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States, ³Departments of Neurology and Pediatrics, Vanderbilt University, Nashville, TN, United States

This work proposes a new method, which we call spherical mean technique (SMT), for estimating the microscopic diffusion anisotropy unconfounded by neurite orientation dispersion and crossing, which are ubiquitous in the brain. A distinguishing feature of the method is that it does not rely on complex diffusion sequences with multiple gradient pulses or magic-angle spinning. We will demonstrate SMT in an ex-vivo mouse brain study and its capability for detecting hypomyelination conditions.

Maira Tariq¹, Michiel Kleinnijenhuis², Anne-Marie van Cappellen van Walsum^3,4, and Hui Zhang^1
1Department of Computer Science & Centre for Medical Image Computing, University College London, London, England, United Kingdom, ²FMRIB Centre, University of Oxford, Oxford, United Kingdom, ³Department of Anatomy, Radbound University, Nijmegen Medical Centre, Nijmegen, Netherlands, ⁴MIRA Institute for Biomedical Technology and Technical Medicine, Enschede, Netherlands

We present validation of a recently proposed technique based on neurite orientation dispersion and density imaging (NODDI), which allows estimation of the dispersion anisotropy of neurites. Dispersion anisotropy is an important feature to measure in vivo as it can enable more accurate characterisation of complex fibre configurations like fanning and bending. We use high-resolution diffusion MRI data from ex-vivo samples of human V1, to evaluate the metrics of the model. We find that the measure of dispersion anisotropy obtained from the NODDI model reflects the expected cytoarchitecture of V1.

Fernando Calamante¹, Ben Jeurissen², Robert Elton Smith¹, Jacques-Donald Tournier^3,4, and Alan Connelly^1
1The Florey Institute of Neuroscience and Mental Health, University of Melbourne, Melbourne, Victoria, Australia, ²iMinds-Vision Lab, Dept. of Physics, University of Antwerp, Belgium, ³Centre for the Developing Brain, King's College London, London, United Kingdom, ⁴Department of Biomedical Engineering, King's College London, London, United Kingdom

Non-invasive mapping of cortical myeloarchitecture has received increasing interest, with MRI methods based on T1/T2/T2*-weighting producing detailed cortical maps based on myelin content. Recent improvements in hardware, acquisition, and analysis methods make diffusion MRI a promising tool for studying myeloarchitecture. We combine high-quality data and recent advances in fibre-orientation modelling to investigate myeloarchitecture in the living human whole-brain. We show cortical patterns similar to those found with other methods, including well-defined areas in the sensory-motor strip, visual cortex, and auditory areas, among others. In vivo human diffusion MRI data should therefore provide a useful complementary approach to study whole-brain myeloarchitecture.