
Eleftherios Garyfallidis
Eleftherios Garyfallidis, an assistant professor of intelligent systems engineering at the School of Informatics, Computing, and Engineering, has been awarded a grant from the National Institute of Health to support Diffusion Imaging in Python (DIPY), which is open source software designed to implement a range of smart methods for computational neuroanatomy.
The grant, which is a collaboration between SICE and the University of Washington, is part of the Collaborative Research in Computational Neuroscience program. The research and development work proposed in the grant will implement and share both well-established and new algorithms for analysis of dMRI, as well as new datasets that will serve as benchmarks for ongoing methods development.
“The DIPY project will overcome the inherent limitations of current analysis methods, the lack of transparency and cohesion in the implementations of new algorithms, and the difficulties in scaling computationally intensive methods to large datasets,” Garyfallidis said.
DIPY implements a broad range of algorithms for the de-noising, registration, reconstruction, tracking, clustering, visualization, and statistical analysis of MRI data. The software allows users to visualize and measure long-range connections in the brain, allowing medical personnel and researchers to understand the development of brain networks during childhood, study the breakdown in neurological and psychiatric disorders, and track the pathways for nerve bundles using a non-invasive technique.
DIPY also provides the foundation for a community of researchers to improve and refine the software to work more efficiently and at a higher resolution. Since the nerve bundles and pathways control such functions as speech, movement, vision, and more, having a clear picture of those pathways can be critical in treating ailments.
“If you can map in-vivo the pathways of the brain, a surgeon can make better plans when removing a tumor,” Garyfallidis said. “For instance, if a language pathway is damaged, you can think about speaking, but you can’t actually speak because you can’t tell the muscles in your face to speak. We build software to protect from damaging important pathways but also study recovery of those pathways across the lifespan.”
Garyfallidis started developing DIPY while he was a Ph.D. student at the University of Cambridge. DIPY now features more than 80 contributing scientists from more than 20 universities around the world and thousands of users, and the grant will further support the work for three years. Garyfallidis is the director of the Garyfallidis Research Group (GRG) at SICE.
“The Department of Intelligent Systems Engineering is excited about this grant, which endorses our approach to Engineering Health research and education, with DIPY illustrating the importance of open source and an Artificial Intelligence-centric approach,” said Geoffrey Fox, the chair of the intelligent systems engineering program.
For more information on neuroengineering at SICE, visit our website.