This INSPIRE project is jointly funded by the Division of Advanced Cyberinfrastructure in the Directorate for Computer & Information Science, Physics of Living Systems in the Division of Physics in the Directorate for Math and Physical Science, Physical and Dynamic Meteorology in the Division of Atmospheric and Geospace Sciences in the Directorate for Geoscience, and the INSPIRE program in the Office of Integrative Activities.
Advances in scientific instrumentation and computational hardware and software have resulted in an unprecedented ability to acquire, simulate, and visualize time resolved three-dimensional (3D) volumes of data, offering the promise of a greater understanding of complex systems previously beyond our technical grasp. However, as the size and complexity of these data increase, analyzing them becomes increasingly problematic, inhibiting scientific discovery and limiting the utility of the data acquired at great expense and effort. Two particularly cogent examples come from two seemingly disparate scientific fields: neuroscience and meteorology. Magnetic resonance imaging (MRI) scanners can now acquire functional MRI (FMRI) volumes of brain activity in almost real time, while mobile Doppler radar (MDR) systems are capable of acquiring time-dependent volumetric images of thunderstorms during tornado formation. In this project, entitled QUantitative Estimation of Space-Time processes in volumetric data (QUEST), the University of California, San Diego, Center for Scientific Computation in Imaging (CSCI), in partnership with the Center for Severe Weather Research (CSWR) and the Cooperative Institute for Meteorological Satellite Studies (CIMSS) will develop a novel framework for the analysis of time-varying 3D volumes, guided by large scale numerical simulations, to investigate two of the outstanding scientific questions of our age: What is the relationship between brain structure and function?, and How do strong, long-track tornadoes form? The resulting computational platform will be disseminated to the NSF community through the open source analysis and visualization platform (STK) to improve the ability of researchers to quantitatively analyze, visualize, and explore complex time varying volumetric datasets.
This INSPIRE project develops advanced methods for automated quantitative characterization of subtle space-time patterns embedded within spatio-temporal data from 3D voxel-based digital imaging modalities based upon the team's recently formulated entropy field decomposition (EFD) theory, a probabilistic method efficiently that employs the information field theoretic approach with prior information supplied using the team's entropy spectrum pathways theory, in conjunction with numerical simulations designed both to constrain results to physically realizable solutions. The cross-disciplinary approach focuses on two outstanding problems in the respective fields of neuroscience and severe weather meteorology: 1) The identification of structural and functional modes of the human brain from high resolution anatomical MRI, diffusion tensor MRI, functional MRI data from the Human Connectome Project combined with numerical simulations of diffusion and functionally weighted MRI signals, and 2) The identification of signatures of tornado genesis and maintenance from MDR data from the Doppler-On-Wheels network in conjunction with tornado simulations using the CM1 model. Significant social impact would result from the ability to categorize states of brain activity in normal and diseased populations and the ability to reduce the lead time between tornado formation and warning to threatened populations. More generally, this novel methodology has the potential to transform the way analysis is conducted in a wide range of disciplines by enabling automated, quantitative detection of important, though perhaps subtle, variations in large, complex datasets undetectable by current traditional techniques.
