3 Post-Docs Up to 5 Years -- Multiscale Cutting in Real Time for Surgical Simulation

Three post-doctoral fellow positions
available over 5 years at Cardiff University in the framework of an ERC Starting
Independent Research Grant with Prof. Stéphane P.A. Bordas and Dr.
Pierre Kerfriden in collaboration with Prof. Karol Miller (University
of Western Australia).

(Synopsis at bottom of post)

We are looking for candidates with
experience in either of the following topics:

• model order reduction (e.g. proper orthogonal decomposition - POD)

• advanced discretisation techniques
  (extended finite element methods/meshless methods)

• nonlinear solid mechanics
  simulations

• multiscale methods (especially in
  relation to fracture)

• high performance computing (domain
  decomposition, preconditioning, multigrid algorithms, solvers, etc.)

• error estimation and adaptivity
  (e.g. for fracture, and if possible in the context of multiscale or
  non-linear problems more generally)

• simulation of large deformation
  (if possible fracture/cutting)

• non-rigid image registration

These are prestigious positions and
we expect a high level of competition.

Main criteria:

• a Ph.D. and
  first Degree for a leading institution in a related area;

• recognised
  scientific level with an evidence of potential for publishing in
  leading international journals;

• the names of
  three references who have agreed to support your application.

Please do not apply if you do not
satisfy the above criteria.

The successful
candidate will join the Institute of Mechanics and Advanced Materials
at Cardiff University School of Engineering
http://www.engin.cf.ac.uk/research/resInstitute.asp?InstNo=13

led by Prof.
Stéphane P.A. Bordas.

They will enter a
vibrant research group (10 Ph.D. Students, 2 post-doctoral fellows)
and be integrated within several ongoing projects in related areas:

• goal oriented
  error estimators for fracture using XFEM (EPSRC) 2010-2013

• from CAD to
  Analysis (Marie Curie Initial Training Network) 2011-2014

• multiscale
  fracture for silicon wafer manufacturing (2011-2014)

Collaborators Our
main collaborator on this grant is Prof. Karol Miller who is a world
leader in surgical simulation. His latest results provide quasi-real
time simulations for non-linear materials on GPUs with applications
to brain surgery simulation and non-rigid registration.

We
will also collaborate with leading (neuro)surgeons from Europe and
beyond.

Contact:
stephane dot bordas at alum dot northwestern dot edu

with copy to

pierre dot
kerfriden at gmail dot com

Synopsis

Surgeons
are trained as apprentices. Some conditions are rarely encountered
and surgeons will only be trained in the specific skills associated
with a given situation if they come across it. At the end of their
residency, it is hoped that they will have faced sufficiently many
cases to be competent. Rising healthcare costs, decreasing working
hours (European Working Time Directives) and staff shortage will
aggravate the situation.

If
we were able to reproduce faithfully, in a virtual environment, the
audio, visual and haptic experience of a surgeon as they prod, pull
and incise tissue, then, surgeons would not have to train on
cadavers, phantoms, or on the patients themselves.

Just
as civil pilots train on a flight simulator to rehearse complex
maneuvers, learn to react to unforeseen events, days before a
surgery, surgeons could operate on a faithful computer-representation
of the same patient they would be seeing in the operating room,
greatly reducing risks of complications. They could also use a
simulator to test alternate operating techniques in a safe
environment (surgical planning).

Only
few researchers in the Computational Mechanics community have
attacked the mechanical problems related to surgical simulation, so
that mechanical faithfulness is not on par with audiovisual realism.
This lack of fidelity in the reproduction of the mechanical behaviour
(tissue, cutting) explains why most surgeons who tested existing
simulators report that the "sensation" fed back to them
remains unrealistic. There
are indeed gargantuan difficulties associated with simulating the
mechanical response of human organs accurately. First, human tissue
is highly variable, and their properties are difficult to obtain, in
part because of the challenges associated with in vivo testing.
Realistic models for these materials are therefore very arduous to
construct, especially if accounting for relevant heterogeneity at
lower scales such as the presence of capillaries and anisotropy.
Second, achieving such realism in the material response concurrently
with adequate geometrical detail in the description of the organs
engenders non-linear systems of millions of equations which, to be
solved on today’s supercomputers require the most powerful
algorithms available and computational power beyond that likely to
ever be accessible to hospitals. Third, discretizing the complex
geometry of organs from medical images including material interfaces
and, importantly, simulating the occurrence and propagation of
incisions still requires heavy user intervention. The
above three difficulties may explain why most groups do not address,
at the same time, geometrical accuracy, material realism, evolving
cuts and quality control.

We
propose to develop algorithms and implement them within a
professionally crafted software toolbox which will enable to simulate
real time cutting in surgical simulation through a novel multiscale
approach. Instead of realizing real time results through explicit
methods and vastly simplified geometries, we will use implicit
methods and novel discretisation schemes able to handle very complex
geometries and cuts easily. This can only be done through severe
reductions in the computational complexity (we aim to reduce this
complexity by up to three orders of magnitude without sacrificing
accuracy and have already achieved, for model problems, two orders of
magnitude savings). The savings are expected to be such that we aim
at, for the first time, predicting in real time the initiation of
cuts at the organ level from the effects of surgical instruments on
the meso/microstructure.