Neuroimaging Analysis Center

Neuroimaging Analysis Center

Research Emphasis

The Neuroimaging Analysis Center (NAC) develops image processing and analysis techniques for basic and clinical neurosciences. The NAC research approach emphasizes both specific core technologies and collaborative application projects.

The core activity of the center is the development of algorithms and techniques for postprocessing of imaging data. New segmentation techniques aid identification of brain structures and disease. Registration methods are used for relating image data to specific patient anatomy or one set of images to another. Visualization tools allow the display of complex anatomical and quantitative information. High-performance computing hardware and associated software techniques further accelerate algorithms and methods. Digital anatomy atlases are developed for the support of both interactive and algorithmic computational tools. Although the emphasis of the NAC is on the dissemination of concepts and techniques, specific elements of the core software technologies have been made available to outside researchers or the community at large.

The NAC's core technologies serve the following major collaborative projects: Alzheimer's disease and the aging brain, morphometric measures in schizophrenia and schizotypal disorder, quantitative analysis of multiple sclerosis, and interactive image-based planning and guidance in neurosurgery. One or more NAC researchers has been designated as responsible for each of the core technologies and the collaborative projects.

Resource Capabilities

Computing Environment

The NAC is a leader in the development and application of advanced image analysis techniques serving a community of local and remote application domain scientists. Local users have access to our integrated computer facility of more than 100 state-of-the-art multi-central processing unit graphics Workstations supported by high-end computer servers in both cluster and multiway symmetric multiprocessor configurations. A unified account management system, gigabit network backbone, and multiterabyte networked storage allow flexible use of this infrastructure from multiple clinical and basic research labs. The NAC welcomes all scientifically qualified users to access these facilities and offers accounts and access to Workstations in drop-in lab facilities. Remote users have the option of Internet access to these facilities, including remote access, Web sites, and file sharing, which support national and international collaborations.

Software

The software infrastructure of the NAC offers both a full range of community-standard research packages and locally developed and supported custom software applications. Software developed by the NAC is focused on the three-dimensional Slicer application software. This open-source cross-platform end user application embodies the latest generation of algorithms from the NAC in an integrated package. Precompiled binaries, mailing lists, user support, training, and example datasets are all provided to enable other research groups to make use of the scientific output of the NAC.

Training Opportunities and Workshops

In collaboration with our research partners, the NAC has offered more than a dozen workshops and training sessions in the past year related to NAC-developed algorithms and software. These events, held at leading research institutions nationwide and internationally, are available to any interested researchers. A similar schedule is planned for upcoming years to address ongoing community interest.

Publications

1. Friman, O., Farneback, G., and Westin, C.-F., A Bayesian approach for stochastic white matter tractography. IEEE Transactions on Medical Imaging 25:965–978, 2006.

2. Talos, I.-F., Zou, K. H., Ohno-Machado, L., Bhagwat, J. G., Kikinis, R., Black, P. M., and Jolesz, F. A., Supratentorial low-grade glioma respectability: Statistical predictive analysis based on anatomic MR features and tumor characteristics. Radiology 239:506–513, 2006.

3. Pohl, K., Fisher, J., Grimson, W. E. L., Kikinis, R., and Wells, W., A Bayesian model for joint segmentation and registration. Neuroimage 31:228–239, 2006.

4. Warfield, S. K., Zou, K. H., and Wells, W. M., Simultaneous truth and performance level estimation (STAPLE): An algorithm for the validation of image segmentation. IEEE Transactions on Medical Imaging 23:903–921, 2004.