BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Isaac Newton Institute Seminar Series
SUMMARY:Reduced order modeling for uncertainty quantificat
ion in cardiac electrophysiology - Andrea Manzoni
(Politecnico di Milano)
DTSTART;TZID=Europe/London:20190606T103000
DTEND;TZID=Europe/London:20190606T110000
UID:TALK125620AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/125620
DESCRIPTION:We present a new\, computationally efficient fra
mework to perform both forward and inverse uncer
tainty quantification (UQ) in cardiac electrophysi
ology. We consider the monodomain model to descr
ibe the electrical activity in a subject-specifi
c left ventricle geometry\, coupled with the Aliev
-Panfilov model to characterize the ionic activi
ty through the cell membrane. We take into accou
nt relevant inputs related to both models\, such a
s electrical conductivities\, pacing times\, and
coefficients affecting the ionic models. We add
ress a complete UQ pipeline\, including: (i) a var
iance-based sensitivity analysis for the selecti
on of the most relevant input parameters\; (ii)
forward UQ to investigate the impact of intra-subj
ect variability on clinically relevant outputs r
elated to the cardiac action potential\, and (ii
i) inverse UQ for the sake of parameter and state
estimation within a Bayesian framework. All thes
e stages exploit stochastic (Monte Carlo) sampli
ng techniques\, thus implying overwhelming computa
tional costs because of the huge amount of queri
es to the high-fidelity\, full-order coupled PDE
-ODEs model. To mitigate this computational burden
\, we replace the high-fidelity model with compu
tationally inexpensive projection-based reduced-
order models aimed at reducing the state-space dim
ensionality. ROM approximation errors on the out
puts of interest are finally taken into account
by means of statistical error models built through
Gaussian process regression\, enhancing the acc
uracy of the whole UQ pipeline.
LOCATION:Seminar Room 1\, Newton Institute
CONTACT:INI IT
END:VEVENT
END:VCALENDAR