BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Comparing sparse brain activity networks
DTSTART:20161020T150000
DTEND:20161020T160000
DTSTAMP:20260407T143838Z
UID:095e6f26ac25412d19e8f69798408fe66b11848b844533c4a1034559
CATEGORIES:Conferences - Seminars
DESCRIPTION:Ben Cassidy (Department of Biostatistics\, Columbia Universit
 y) Dr. Ben Cassidy's research interests include statistical signal process
 ing\, ill-conditioned inverse problems\, high dimensional statistics\, and
  mathematical topology\, with particular application to neuroimaging and n
 etwork problems.\nFunctional magnetic resonance imaging of resting state b
 rain activity has become a mainstay of modern neuroscience research. Howev
 er\, there are problems with existing methods for identifying\, characteri
 zing and comparing networks obtained from fMRI data\, leading to many conf
 licting results in neuroimaging research.    In this talk we introduce t
 wo new methods\, for network identification and network comparison.  The 
 first method estimates sparse time series conditional independence network
 s\, jointly dealing with crucial issues of spurious spatial correlations\,
  temporal correlations and data dimensionality.  The second method compar
 es networks\, when the networks may be sparsely connected and of different
  sizes\, by leveraging functional data analysis from statistics with persi
 stent homology from mathematical topology. We use the network comparison t
 ools to investigate the performance of the new network identification meth
 od on real fMRI data\, showing marked improvements over existing methods. 
   This is joint work with Victor Solo\, DuBois Bowman and Caroline Rae.
LOCATION:B1.02 Videoconference Room\, Campus Biotech
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR