Biometaphorical Computing Research - overview

Blue Brain Project rendering of neocortical column neurons and the locations of potential touches. The Biometaphorical Computing Research Group is dedicated to neuroimaging and to analyzing, modeling, and understanding the brain for scientific, medical, and technology applications.

The initial research focus of the group was identifying and applying the principles of global brain function to the design of technologies with uniquely biological capabilities. The term, "Biometaphorical Computing," was coined to emphasize the use of biological computation as a metaphor for how computation might be done and to distinguish the distinctive, multi-disciplinary, holistic research perspective of the group from the perspectives implied by other terms in the literature.

Due to the intensive use of neuroimaging and neuroscientific data analysis, the group joined the Computational Biology Center in 2005 and significantly expanded its research scope to include the needs of the scientific, medical, and imaging communities. The group exploits high performance computing technology for technology innovation and scientific discovery throughout its research. It is a member of IBM's Blue Gene team, which received the 2009 National Medal of Technology.

The group's current scope of research spans five broad areas:

  • Neuroimaging: to translate imaging system measurements into data about the system under study that are of higher quality and are tailored to the specific scientific questions pursued.
  • Neuroscientific Analysis: to translate data about the system under study into insights and usable, interpretable information.
  • Neural System Modeling: to translate insights and information into understanding through predictive models validated by biological observations and capabilities.
  • Analysis Tools and Applications: to develop machine learning and statistical analysis tools and techniques tailored to the magnitude and complexity of neuroscientific data.
  • Software Tools and Infrastructure: to develop high performance, distributed, parallel software tools supporting the other research areas.

Please follow the links to learn more about each research area, for descriptions of representative projects, and to access relevant publications.