|
|
NSF Project: Integrated Modeling and Learning of Multimodality Data across Subjects for Brain Disorder Study
Graphics and Imaging Lab Wayne State University
Summary With ever-improving imaging technologies and ever-increasing high-performance computational power, the complexity and scale of acquired brain imaging data have continued to grow at an explosive pace. Rapid advances in multimodality imaging technologies have significantly accelerated brain disorder studies by providing complementary information on many aspects of the human brain in the normal and diseased states. Capitalizing on the availability of large-scale data, we are now able to computationally integrate, index and model the brain functions across a large population for discovering more detailed understanding and more profound knowledge about complex biological interactions in the human brain. Based on our continuous research effort along this direction, this NSF project is developing a novel, rigorous theoretical framework based on Riemannian geometry, multivariate simplex splines, and statistical learning, which provides a basis for multimodality information integration and understanding across populations. Specifically, our research team will design a fundamental framework for advanced and integrated analysis of brain imaging data. It is expected that the developed, advanced informatics tools will allow the quantitative and integrative analysis of a variety of functional patterns and the relationships between anatomical and functional features in different datasets. The proposed computational framework has the potential to be applied across multiple areas of brain research as well as in clinical diagnosis.
We are pleased to announce the first release of our software, BrainSpace version 1.0 (release date: April 09, 2009). You may browse our brief software manual (PDF Version) for more information regarding the BrainSpace. Detailed instructions regarding how to install and use the software will be provided upon registration. The figure below shows the conformal brain surface model (Figure 1A and Figure 1B) facilitates accurate matching and registration among subjects in the canonical, spherical domain, hence supporting integrated cross-subject analysis of Positron Emission Tomography (PET) (molecular-level brain activity analysis) (Figure 1C), Diffusion Tensor Imaging (DTI) (neural fiber connectivity analysis) (Figure 1D), and Electroencephalography (EEG) (time-varying signal analysis) (Figure 1E) in computer-aided diagnosis of brain disorders. Figure 2 shows the basic processing pipeline in BrainSpace.
The software installation package and manual can be download free after a validated registration. Please follow this link for registration and download. The related publications is as follows. Please make proper references to one or more of the related articles if you use BrainSpace for your research and publications. If you have any question, suggestion, or comment, please contact us via ai2543@wayne.edu.
Selected, Related Publications:
| ||||||||||