This project will enable the PI to transition from his research on computational and social networks to the study of Brain Networks for biomedicine and bioimaging. This research will apply old and new ideas from network theory, combining techniques from applied mathematics, computer science and physics to analyze large Brain Networks in order to understand and diagnose conditions arising from brain disorders. This project will create a group of biomedicine and bioimaging researchers which is unique in its breadth and strengths for this endeavor. It will build on the PI's years of experience in the study of algorithms methods of analysis for engineered networks and his interdisciplinary work applying these ideas to problems in computer science, physics, applied mathematics and financial economics. It will also build on pathbreaking research by the three mentors (and their research groups), who are faculty in Radiology at UCSF and core members of the joint UCB/UCSF bioengineering group, on the construction and preliminary analysis of Brain Networks. Their expertise includes bioimaging (hi-res MRI, diffusion MRI, fMri, and MEG), tractography (and its statistical analysis) and clinical applications of these tools to a wide variety of brain disorders and injuries. Although we expect this project to be relevant to many brain disorders, we will initially focus on Agenesis of the Corpus Callosum (AgCC), an important connectivity disorder of the brain. We have multi-modal data on a large and growing cohort of AgCC patients and controls which will allow us to both hone our methods and improve our understanding of the correlations between Brain Networks and the many specific conditions induced by AgCC in patients. The PI is a Senior Research Scientist at the International Computer Science Institute (ICSI) in Berkeley and an Associate Professor of Operations Research at Cornell University. He has been a PI on several large interdisciplinary research projects involving networks. However, he is new to biomedicine and bioimaging and this grant would give him the time and freedom to learn about the field (through both coursework and research) and retool his research program. His eventual goal is to become an expert in Brain Networks and their application to biomedicine and bioimaging and to create a research center at ICSI combining students and post-docs in biomedicine and bioimaging with those in computer science, applied mathematics and physics which will focus on the development of Brain Network methodologies. ICSI provides the perfect base for this endeavor as it allows the PI full freedom to focus on research and has strong connections to computer science, applied mathematics and biology at UCB and is extremely supportive of the PI's attempts to create new connections to bioengineering, biomedicine and bioimaging. It's location is central, with easy access to UCB, UCSF and Silicon Valley, with its wealth of networking knowledge. The proposed research is innovative in the application and development of robust localized network measures, based on recent research by the PI and others in network analysis. It will apply more focused network measures and incorporate more information (including the physical and functional location of nodes) to elicit connections that are too subtle to be detected in standard whole network measures. It will lead to new understanding of AgCC and techniques for correlating brain imaging and specific conditions in AgCC patients, providing new and useful diagnostic capabilities which will be useful for treatment of patients with AgCC. These tools will also be made publicly available to researchers in biomedicine and bioimaging as well as clinicians. In addition, since the research will be focussed on the development of robust localized network measures it will include a careful study of the robustness of Brain Network construction. It will compare various methods of constructing these networks from bioimaging data and statistically verify the robustness and repeatability of these techniques. It will consider the scale at which networks are constructed (parcellation), the manner in which nodes are chosen (such as functional regions of the brain) and the choice of cutoffs to define the existence of links between nodes. Lastly, while at least 75% of the PI's time will be spent on this project, the PI is also involved in an NSF-funded project to develop new methods for analyzing networks, which will mesh well with this project and provide impor- tant synergies, allowing ideas to cross freely between the different projects. In particular, Brain Network analysis raises new and interesting questions in the study of networks (such as the scaling behavior of parcellation) which would be in the domain of the NSF-funded project, but the answers to these questions will inform the construction and analysis of Brain Networks. This give and take between theory and application will increase the potential for new and important results in the analysis of Brain Networks and provides an additional boost to the proposed project.
The new insights into the local and global connectivity of the human brain and their alterations in neurological and psychiatric disorders that are gained from these new approaches to brain network analysis may help tailor therapies such as cognitive rehabilitation in TBI and motor rehabilitation in MS and stroke. In particular, when extended further into the clinical realm, this network-based analysis of MEG, fMRI, and DTI/HARDI might offer new insights into the structural and functional neuroplasticity that is the hallmark of successful therapies for these disorders. Also, it and may be helpful for patient selection in clinical trials of experimental interventions for these disorders as well as for treatment monitoring strategies.
| Friedman, Eric J; Landsberg, Adam S; Owen, Julia P et al. (2014) Stochastic geometric network models for groups of functional and structural connectomes. Neuroimage 101:473-84 |
| Friedman, Eric J; Young, Karl; Asif, Danial et al. (2014) Directed progression brain networks in Alzheimer's disease: properties and classification. Brain Connect 4:384-93 |
| Owen, Julia P; Ziv, Etay; Bukshpun, Polina et al. (2013) Test-retest reliability of computational network measurements derived from the structural connectome of the human brain. Brain Connect 3:160-76 |
| Owen, Julia P; Li, Yi-Ou; Ziv, Etay et al. (2013) The structural connectome of the human brain in agenesis of the corpus callosum. Neuroimage 70:340-55 |
| Friedman, Eric J; Landsberg, Adam S (2013) Hierarchical networks, power laws, and neuronal avalanches. Chaos 23:013135 |
