A pinboard by
Lisa Jin

Incoming PhD student, University of Rochester


Within a wide variety of domains, graphs are used to express the relationships between entities, such as a social network's friendships between users. In addition to the density of these graphs, time evolution provides further complexity to summarizing the underlying data. Given such a large, time-evolving network, there exist many algorithms to extract their most salient structures, yet visualizing their results remains an open problem. Functional brain networks that are inferred from fMRI data represent one such dataset that benefits from an interpretable visualization of graph summary results. Thus, my research currently focuses on a system that (1) leverages domain expert knowledge to mine and visualize summaries of time-evolving networks and (2) enables comparative analysis of constructed networks. This is done by using communities present in the data, such as a label of DMN (default mode network) that nodes belong to, in order to guide the summarization process. Understanding patterns both within and between these communities are key insights that can be communicated through visualization. Neuroscientists face additional challenges in representing fMRI data as graphs, since the functional brain networks are often thresholded correlations between voxels (volume units of neurons). Quality of graph construction directly affects the accuracy of the graph mining results, so it is a priority to support efficient comparison of graphs themselves. In the preprocessing step of the system is an algorithm that efficiently mines and encodes the summarization ability of substructures. It relies on the expert-provided communities as a first step in segmenting the entire network. The next parts of the system include a back-end and front-end, containing the database and API, and JavaScript-driven web interface, respectively. The former uses a graph database to query for graph substructures, and the latter provides a responsive user interface for visualization.


Hierarchical modularity in human brain functional networks.

Abstract: The idea that complex systems have a hierarchical modular organization originated in the early 1960s and has recently attracted fresh support from quantitative studies of large scale, real-life networks. Here we investigate the hierarchical modular (or "modules-within-modules") decomposition of human brain functional networks, measured using functional magnetic resonance imaging in 18 healthy volunteers under no-task or resting conditions. We used a customized template to extract networks with more than 1800 regional nodes, and we applied a fast algorithm to identify nested modular structure at several hierarchical levels. We used mutual information, 0 < I < 1, to estimate the similarity of community structure of networks in different subjects, and to identify the individual network that is most representative of the group. Results show that human brain functional networks have a hierarchical modular organization with a fair degree of similarity between subjects, I = 0.63. The largest five modules at the highest level of the hierarchy were medial occipital, lateral occipital, central, parieto-frontal and fronto-temporal systems; occipital modules demonstrated less sub-modular organization than modules comprising regions of multimodal association cortex. Connector nodes and hubs, with a key role in inter-modular connectivity, were also concentrated in association cortical areas. We conclude that methods are available for hierarchical modular decomposition of large numbers of high resolution brain functional networks using computationally expedient algorithms. This could enable future investigations of Simon's original hypothesis that hierarchy or near-decomposability of physical symbol systems is a critical design feature for their fast adaptivity to changing environmental conditions.

Pub.: 02 Dec '09, Pinned: 02 Jul '17

Exploring the brain network: a review on resting-state fMRI functional connectivity.

Abstract: Our brain is a network. It consists of spatially distributed, but functionally linked regions that continuously share information with each other. Interestingly, recent advances in the acquisition and analysis of functional neuroimaging data have catalyzed the exploration of functional connectivity in the human brain. Functional connectivity is defined as the temporal dependency of neuronal activation patterns of anatomically separated brain regions and in the past years an increasing body of neuroimaging studies has started to explore functional connectivity by measuring the level of co-activation of resting-state fMRI time-series between brain regions. These studies have revealed interesting new findings about the functional connections of specific brain regions and local networks, as well as important new insights in the overall organization of functional communication in the brain network. Here we present an overview of these new methods and discuss how they have led to new insights in core aspects of the human brain, providing an overview of these novel imaging techniques and their implication to neuroscience. We discuss the use of spontaneous resting-state fMRI in determining functional connectivity, discuss suggested origins of these signals, how functional connections tend to be related to structural connections in the brain network and how functional brain communication may form a key role in cognitive performance. Furthermore, we will discuss the upcoming field of examining functional connectivity patterns using graph theory, focusing on the overall organization of the functional brain network. Specifically, we will discuss the value of these new functional connectivity tools in examining believed connectivity diseases, like Alzheimer's disease, dementia, schizophrenia and multiple sclerosis.

Pub.: 18 May '10, Pinned: 02 Jul '17