BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T000604Z
LOCATION:E Concourse
DTSTART;TZID=America/Denver:20231115T100000
DTEND;TZID=America/Denver:20231115T170000
UID:submissions.supercomputing.org_SC23_sess301_drs105@linklings.com
SUMMARY:High Performance Computing for Optimization of Radiation Therapy T
 reatment Plans
DESCRIPTION:Felix Liu (KTH Royal Institute of Technology, Sweden; Raysearc
 h Laboratories)\n\nModern radiation therapy relies heavily on computationa
 l methods to design optimal treatment plans (control parameters for the tr
 eatment machine) for individual patients. These parameters are determined 
 by constructing and solving a mathematical optimization problem. Ultimatel
 y, the goal is to create treatment plans for each patient such that a high
  dose is delivered to the tumor, while sparing surrounding healthy tissue 
 as much as possible. Solving the optimization problem can be computational
 ly expensive, as it requires both a method to compute the delivered dose i
 n the patient and an algorithm to solve a (in general) constrained and non
 linear optimization problem.\n\nThe goal of this thesis project has been t
 o investigate the use of HPC hardware and methods to accelerate the comput
 ational workflow in radiation therapy treatment planning. First, we propos
 e two methods to bring the optimization to HPC hardware using GPU accelera
 tion and distributed computing for dose summation and objective function c
 alculation respectively. We show that our methods achieve competitive perf
 ormance compared to state-of-the-art libraries and scale well, up to the A
 mdahl’s law limit.\n\nThen, we investigate methods to accelerate interior 
 point methods, a popular algorithm for constrained optimization. We invest
 igate the use of iterative Krylov subspace linear solvers to solve Newton 
 systems from interior point methods and show that we can compute solutions
  in reasonable time for our problems, in spite of extreme ill-conditioning
 . This approach presents one avenue by which constrained optimization solv
 ers for radiation therapy could be ported to GPU accelerators.\n\nTag: Acc
 elerators, Artificial Intelligence/Machine Learning, Applications, Cloud C
 omputing, Distributed Computing, Data Analysis, Visualization, and Storage
 , Data Compression, Heterogeneous Computing, I/O and File Systems, Quantum
  Computing, Reproducibility, Security, Software Engineering\n\nRegistratio
 n Category: Tech Program Reg Pass, Exhibits Reg Pass\n\n
END:VEVENT
END:VCALENDAR