|
| Summary document | |
Version
française
The major challenge for the coming years or decades is to set up anatomical,
physical and functional models of the human body that include the biomechanical
properties of tissues and physiological systems.
With parameters adapted to the specific case of a patient, such models
could provide valuable insight for prevention (by simulating the evolution
in time), diagnosis (invisible anomalies) and therapy (simulation, control
and a posteriori assessment).
Project BANG is devising new models of the development of cancereous tumors
which will be used to test and improve therapies.
Projects such as REO, that work on fluid mechanics for the exploration
and visualization of the vascular system, will contribute a great deal
to the modeling of blood flows.
Project MACS specializes in solid mechanics
and is involved in the modeling of such complex structures as the heart
and the liver.
Project SOSSO is currently working on models of cardiac activity under
the control of the autonomic nervous system.
The DREAM team is developing systems that put together diagnosis and decision
for intelligent cardiac prostheses.
Cell cycles, movements and mutations. Can modeling lead to improvements
in cancer therapies by contributing to understanding these phenomena?
 |
© INRIA |
Human
experimentation is extremely limited. It is therefore important for research
scientists to have a model, an actual simulator, that will make it possible
to test and validate new therapeutical approaches. The modeling of biological
systems however constitutes a special challenges for researchers due
to their complexity and diversity.
The BANG team is specialized in numerical analysis and mathematical models.
It came together with the goal of applying a significant part of its
research to medical biology, especially understanding the phenomena linked
to the development of tumors.
For example, does the circadian cycle have an impact on the development
of cancer tumors and on the efficiency of chemotherapies?
Would blocking angiogenesis (the aggregation of cells that precedes the
formation of new capillaries, necessary to feed a tumor) prevent the
development of metastases?
In order to test the hypotheses made by biologists, Bang develops, analyzes
and simulates mathematical models that take into account multiple parameters
as well as their evolution over time (the age of the cells, for example).
Comparison with experimental observations will make it possible to validate
the relevance and robustness of these models that can then be used to
simulate therapeutic choices or optimize their efficiency.
The team started working in collaboration with the cancer chrono-therapeutics
team of the Paul Brousse hospital and with an INSERM team of the Paris
VI School of Medicine. Bang also applies its research to fluid flows
(water quality, pollutant dispersion, etc.) and to the modeling of avalanches
and landslides, in collaboration with the seismology department of the
Globe Physics Institute.
The project in brief
As soon as you try to automate diagnosis a key question is posed: how do you distinguish between a genuine anomaly signaling a problem that requires intervention and a minor, exceptional incident?
This is
one of the questions that the Dream team is working on in the context
of its research. One of the applications concerns cardiac monitoring.
This work could result in striking industrial products such as the development
of "intelligent" cardiac prostheses capable of adapting themselves
to the patient's specific data and lifestyle. Thus, based on continuous
monitoring of the heart, whose rhythm often varies, the prosthesis should
advisedly decide at which point to trigger corrective action-stimulation
or defibrillation for example.
DREAM's approach compares information supplied by the monitoring system
sensors with models whose evolution is insured by automatic acquisition:
good functioning model or "failure" model. Based on a modeling
of cardiac arrhythmias, DREAM leverages artificial intelligence techniques
to develop systems that may be used for online electrocardiogram analysis,
and will in the long run be used to develop intelligent cardiac prostheses.
The work of DREAM is also used for other complex systems-environment
protection and telecommunication or power grid network monitoring.
The project is carried out jointly with CNRS, INSA Rennes and the University
of Rennes I. It is part of the MONET 2 excellence network (Model-based
systems and qualitative reasoning), and of the Bridge workgroup in particular.
The Dream team is manned by Irisa research scientists. The team participates
in the RNRT project CEPICA, a continuation of the PISE ACI, in collaboration
with the LTSI (Signal Processing Laboratory), the teaching hospital of
Rennes and Ela Medical, with the goal of developing intelligent cardiac
prostheses.
The team in brief
 |
© INRIA |
A heart
attack does not leave the heart intact. Damaged tissue may stretch little
by little, causing a hernia and threatening the patient's life. It is therefore
important to precisely diagnose the affected zone of the cardiac muscle
and to assess its functional capacity, in order to identify the repairing
therapy that may possibly be applied. The MACS team works on three-dimensional
modeling of the heart, in the framework of the ICEMA ARC. By using this
model via simulations and integrating measurements made on the patient
(imaging, ECG, etc.), clinicians will be able to better identify the damaged
zone and quantify its ability to contract (contractility index).
Scientific computing is a crucial and indispensable part of this work.
In the absence of computations that can make it "real", a model-a
set of equations-has few applications. Besides, the two approaches
are intrinsically linked: improving models require new computing methods
and simultaneously, advances in mathematics lead to advances in modeling.
MACS specializes in scientific computing for solid mechanics and structural
mechanics. The team applies its research to the three-dimensional modeling
of the heart. Such models are developed by partners (SOSSO in particular).
Computations provide quantitative results on these models that can thus
be validated and/or modified by comparison with observed data. Finally,
by integrating measurements and images of a patient, it is possible to
recover functional indicators for diagnosis aid.
MACS is devoting an increasing part of its work to biomedical applications.
The team is working in particular on the modeling of fluid/solid interaction,
in collaboration with REO (fluid mechanics), in order to represent and
model the flow of blood with deformations of the arterial walls. MACS also
apples its research to the modeling of biological cell behavior.
In addition, the MACS team continues its work in other application fields,
such as tires and MEMS (Micro Electro-Mechanical Systems), tiny mobile
parts activated by electronic devices.
The project in brief
Can blood flow modeling help improve prevention and treatment of
cardiovascular accidents?
 |
© INRIA |
The REO team specializes in biofluid mechanics. It studies blood flows
in the cardiovascular system, and airflow in the respiratory system.
The team's goal is to develop models and simulations for such flows at
the service of medicine. REO work has numerous applications.
REO couples a detailed modeling of part of the circulatory and respiratory
systems with simplified models for other components of these systems
(electrical analogs, reduced base element models, fractal homogenization
and multiphysics). The circulatory and respiratory simulators thus obtained
will give valuable supplementary information for diagnosis and will guide
therapeutics. In the case of aneurisms, for example, numerical simulations
of the blood flow make it possible to identify the wall regions likely
to crack and then break.
The modeling of air transport in the respiratory tract opens up new prospects
for the study of aerosols, especially in the context of less invasive
treatments. For example, diabetes treatment by insulin aerosol would
make it possible to avoid daily shots. Finally, the simulation of the
impact of certain cardiovascular implants can lead to the develop of
better implantable devices.
Project REO has multiple skills: biofluid mechanics, fluid-structure
interaction, solid mechanics, numerical simulation. It is supported by
the National Network for Health Technologies of the Ministry of Research
and by the Research Training Network of the European Commission. The
project also participates in a Franco-Canadian collaboration with the
Mathematics Center of the Laval University in Québec, the Universities
of Ottawa and Montréal and the École Polytechnique of Montréal.
The team in brief
Today, cardiologists have at their disposal images and measurements
(pressure, electrocardiograms) that help them assess the state of
health of a patient and
make a diagnosis. Since the heart is a controlled system that associates a mechanical
part-the muscle-with a control part-electrochemical phenomena
inside the heart or the autonomic nervous system-SOSSO had the idea of
applying automatic control, a modeling of controlled dynamical systems, to the
development of analysis and diagnosis aid tools. The goal is to give doctors
quantitative indicators of the state of the cardiovascular system and its control
based on available information, images and signals. Concerning control, this
will go beyond the descriptive and qualitative elements available today.
The first SOSSO research team had succeeded in developing a multiscale model
of the cardiac muscle compatible with molecular scale descriptions (nano-engines)
and cellular scale descriptions (sarcomers). Today, the team is working on a
model that will reflect the functioning of the system over several heartbeats,
whereas other command systems come into play at each of these scales. This hierarchy
of controlled systems is an example of the multiscale system automatic control
that SOSSO 2 is striving to make emerge.
At the same time, the development of new imaging sources gave access to increasingly
reach and complex information. The current model is used to interpret this information
at the scale of the organ (ICEMA initiative).
Integrating knowledge across the different levels (organ, cell, molecule) is
a scientific challenge for the team, whose goal is to make available to cardiologists
tools based on the available information for each patient and directly useful
for diagnosis and understanding of certain phenomena. In this context, the SOSSO
2 team is working in partnership with the ICEMA ARC teams and the Antoine Béclère
Hospital in Clamart.
The project in brief
