Diffusion Weighted Magnetic Resonance Imaging (DW-MRI) offers a unique possibility for in vivo insight on neuronal pathways in brain. The potential impact of such a technology is broad and far reaching, ranging from study of normal brain functions and various medical conditions, such as multiple sclerosis, Alzheimer's disease and traumatic brain injuries, to straightforward applications in surgical planning. Despite great promise, DW-MRI and in particular, tractography (reconstruction of neuronal fibers by DW-MRI) has not made significant impact yet. Technical issues plaguing the field come from the fact that current approaches rely on methods that are either: (1) local in nature, reconstructing the neuronal pathways one at the time, in a step-by-step fashion, by using only local information on diffusion -such methods inherently suffer from problems with fiber crossing and splitting;or/and (2) Monte Carlo based methods (with ad hoc parameters) leading to great variability and low reliability of reconstructed fibers. We propose to tackle some of these questions in this project, specifically question of anatomic fiber weight estimation, because existing probabilistic tractography algorithms estimate strength of the connections incorrectly, by favoring the shortest, straightest and simplest paths.
We aim to offer a solution to this issue by incorporating non-diffusion information on the amount of white matter present in voxels and estimating fiber weight globally, by considering all the fibers and all the voxels simultaneously. We would also like to develop consequential improvements in tracking with an aim to decrease the number of ad hoc parameters and variability of results. Set of potential applications for anatomically accurate measure of fiber tract weight (defined as the amount of white matter oriented from one cortical region to another) is substantial. We are primarily interested in the applications to study of neurodegenerative diseases, namely Multiple Sclerosis (MS), because of its pathological characterization (demyelination). Existing DW-MRI metrics correlate weakly with clinical outcomes in MS and need for advanced MRI metrics is well recognized within the MS community. A question of our particular interest is whether grey and white matter loss in MS is directly connected or independent. Previous studies using standard DW-MRI metrics gave inconclusive results. Successful correlation of grey matter volume loss with loss in weight on fibers connected to corresponding grey matter regions would confirm the hypothesis that grey and white matter loss in MS are directly connected. This will have an impact on research, diagnosis, and monitoring. During the course of this fellowship Dr. Ivkovic will receive training in neuroradiology, clinical trials methodology and parallel programming.

Public Health Relevance

The main goal of this project is to develop more accurate and diagnostically more relevant quantitative measures based on magnetic resonance imaging. The development of these measures will enable better understanding on how brain grey matter and white matter are affected by Multiple Sclerosis and other neurodegenerative diseases. During the project, Dr. Ivkovic will receive training in medical imaging, neurology, clinical trial methodology, parallel programming and responsible conduct of research.

National Institute of Health (NIH)
National Institute of Neurological Disorders and Stroke (NINDS)
Postdoctoral Individual National Research Service Award (F32)
Project #
Application #
Study Section
Special Emphasis Panel (ZRG1-F01-L (20))
Program Officer
Utz, Ursula
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Weill Medical College of Cornell University
Schools of Medicine
New York
United States
Zip Code