Laboratory of Neuro Imaging Resource

Laboratory of Neuro Imaging Resource

UCLA School of Medicine Department of Neurology 635 Charles East Young Drive South Suite 225 Los Angeles, CA 90095-7334 www.loni.ucla.edu Grant No. P41 RR013642

Principal Investigator and Contact Arthur W. Toga, Ph.D. 310-206-2101; Fax: 310-206-5518 Coinvestigators Richard Leahy, Ph.D. David Shattuck, Ph.D. Paul M. Thompson, Ph.D. Roger P. Woods, M.D

Research Emphasis

The Laboratory of Neuro Imaging Resource (LONIR) develops novel strategies to investigate brain structure and function in their full multidimensional complexity. There is a rapidly growing need for brain models comprehensive enough to represent brain structure and function as they change across time in large populations, in different disease states, across imaging modalities, across age and sex, and even across species. International networks of collaborators are provided with a diverse array of tools to create, analyze, visualize, and interact with models of the brain. A major focus of these collaborations is to develop four-dimensional brain models that track and analyze complex patterns of dynamically changing brain structure in development and disease, expanding investigations of brain structure-function relations to four dimensions.

Current Research

The development of modeling approaches focuses on new strategies for surface and volume parameterization that provide an advanced analysis of surface and volumetric brain models, tracking their change across time. Additional research cores focus on anatomic fundamentals, analyzing anatomic and cytoarchitectural attributes across multiple scales and across time. Another core focuses on visualization and animation, for the dissemination of brain models that visualize complex variations in brain structure and function across time. Specialized approaches are under development for handling cortical data. Ongoing national and international collaborations are analyzing normal and aberrant growth processes, brain development, tumor growth, Alzheimer's disease and related degenerative disease processes, schizophrenia, and brain structure in normal and diseased twins.

BIRN

The center is a partner in the Biomedical Informatics Research Network (BIRN) effort of NCRR.

Resource Capabilities

LONIR develops and disseminates to the community computational tools for neuroimaging and brain mapping. In addition, the resource provides state-of-the-art grid computing resources, data storage, and archival to all of our collaborators.

Methods

LONIR designs algorithms that generate average models of brain anatomy and maps of growth, degeneration, and their population statistics, which are based on parametric surfaces, volumetric morphology, and topology-preserving mapping. The resulting algorithms are implemented, validated, and distributed via the LONIR Pipeline environment and are applicable to a variety of computational neuroscience challenges in normal brain and disease. The approach to integrated computational neuroscience is based on scalable, portable, and distributed infrastructure, which uses object-oriented programming, Extensible Markup Language, encrypted distributed computing and open-source design, implementation, and tool dissemination.

Instruments

The complexities of the aforementioned algorithms and research methodologies require considerable computational and storage resources. Currently, the LONIR uses clustered computers comprising approximately 720 processors, from industry partners SUN Microsystems and Dell, and symmetrical multiprocessing systems from Silicon Graphics, totaling about 132 processors. These disparate systems are virtualized into a computational service using SUN Grid Engine to facilitate user access and automate the distribution of computer jobs to the appropriate resources.

A sophisticated hierarchically managed storage system supports the computational infrastructure mentioned above. The LONIR uses a fault-tolerant storage area network (SAN) comprising clusters and supercomputers, approximately 60 terabytes of fault-tolerant redundant array of independent disk storage, and redundant robotic silos located in different data centers with an aggregate capacity of 4 petabytes and 14 high-bandwidth tape drives, all of which are interconnected using 2 gigabit per second fibrechannel. Migration from disk to tape is contingent on the size, age, and ownership of the data and occurs transparently. The facility has partnered with Cisco Systems to provide network access, voice over Internet Protocol telephony, and security. The LONIR network infrastructure is composed of full gigabit and wireless 802.11-bg connectivity. Redundant edge routers ensure service provision in the event of a single router failure. Security is enforced using a hardware firewall; wireless access is permitted only through a virtual private network.

The Data Immersive Visualization Environmnet (DIVE) is a notable feature in the LONIR facility. This stereocapable immersive theatre is dominated by a 12' 120º curved screen powered by three digital projectors. The DIVE provides researchers with a novel environment to investigate multidimensional data in three-dimensional space.

Software

A sampling of available software includes Pipeline, BrainSuite, Debabler, and Image Data Archive (IDA).

LONIR Pipeline Environment: A graphic environment for constructing complex scientific analyses of data. It provides a visually intuitive interface to data analysis while also allowing for diverse programs to interact seamlessly. The Pipeline allows researchers to share their methods of analysis with each other easily and provides a simple platform for distributing new programs, as well as program updates, to the desired community. LONIR has successfully integrated the Pipeline environment with SUN Grid Engine, using SUN Microsystem's Java DRMAA bindings, to facilitate the submission of jobs to distributed computational resources.

BrainSuite: An interactive brain modeling and visualization software for processing and extracting topologically spherical cortical surfaces from raw magnetic resonance volumes, skull and scalp segmentation, surface analysis, and visualization.

Debabler: A graphical user interface-based data format converter that can analyze and categorize imaging data according to metadata values and convert the data into the appropriate output file format through a programmable set of connected translation modules.

LONIR Image Data Archive (IDA): An integrated environment for safely archiving, querying and visualizing neuroimaging data utilizing a Web-browser interface. The archive protects valuable data from unauthorized access while providing the ability to share with collaborators. To meet HIPAA compliance, client-side de-identification is done by an IDA Java applet at the user's local workstation prior to sending the de-identified data over an encrypted channel to be stored in the IDA. The integrity of stored data is protected by fault-tolerant storage as well as redundant offsite tape backups.

Available Resources

Computational resources: Manage and provide access to our supercomputer systems.

Visualization resources: Provide sophisticated scientific visualization and animation that illustrate the results of the algorithmic and analytic strategies used in the core research areas and by our collaborators.

Data management: Includes the LONI IDA–an efficient relational database and supporting infrastructure.

Training Opportunities and Workshops

Formal training program: A comprehensive training program currently exists and is coupled to the Resource. Funding for trainees has been awarded in a T32 Training Program in Neuroimaging grant from NIMH.

Workshops, courses, and education: LONIR provides regular seminars, workshops, and courses on advanced information technology. Building on a mature and established Training Program in Neuroimaging, the resource has designed our workshops and courses to include training on multidimensional modeling.

Research training program: LONI offers 2- to 10-week-long rotations in various research areas, such as atlasing, genotype, and phenotype studies; functional magnetic resonance imaging; and parametric modeling.

Visiting professorship series: This series is designed to provide more long-term training of individuals who are interested in neuroimaging and multidimensional modeling. By providing the appropriate space and resources necessary to equip visiting investigators, the resource can enhance the visitors' experience with the resource.

Publications

1. Toga, A. W., Thompson, P. M., and Sowell, E. S., Mapping brain maturation. Trends in Neuroscience 29:148–159, 2005.

2. Thompson, P. M., Dutton, R. A., Hayashi, K. M., et al., Thinning of the cerebral cortex visualized in HIV/AIDS reflects CD4+ T-lymphocyte decline. Proceedings of the National Academy of Sciences USA 102:15647–15652, 2006.

3. Luders, E., Narr, K. L., Zaidel, E., Thompson, P. M., Jancke, L., and Toga, A. W., Parasagittal asymmetries of the corpus callosum. Cerebral Cortex 16:346–354, 2005.