Applied Mathematics (LMAC) - EA 2222

The UTC applied mathematics laboratory (UTC-LMAC) is at the core of engineering sciences at UTC. LMAC is a transdisciplinary research unit specialised in inverse problems, stochastic processes, digital analysis, statistics and reliability, which is a set of problems that depend on societal challenges with industrial or socio-economic partners.

Objectives

The UTC Applied Mathematics Laboratory is currently developing high level research in applied, deterministic and stochastic mathematics.

Its areas of excellence lie in applications and development of efficient tools to be used in scientific computations. These two areas are viewed by LMAC as complementary and coherent with UTC's policy to sign varied partnerships, i.e., between UTC's laboratories as well as relevant external research teams.

Research Teams and Thematics

LMAC's research portfolio addresses 2 main sets of problems:

  • inverse problems
  • stochastic systems

These problems, both from a theoretical and from an applied point of view, arise in questions faced by engineers in applied sciences and engineering in general.

The thematics and applications are as follows:

  • Inverse and badly formulated problems (IP): Identification of parameters, data Infilling, fluid-structure Interactions, validity testing in diffusion problems, shape and topology optimization.
  • Stochastic Systems (S2): Semi-Markov processes, stability factors and stationary conditions, non parametric evaluation, non-parametric and semi-parametric test protocols.
  • Recent (or ongoing) Applications: Electro-encephalograms (EEG), optical tomography (inverse NIRS), crack propagation, dislocation phenomena, saline intrusion, atmospheric pollution, identification of metabolic flow in plants, internal combustion engine modelling, reliability, performance and stability of complex systems, random networks and telecommunications, epidemiology, Bayesian Modelling of biological gene networks, seismology, DNA Modelling.

We are developing partnerships with industrial partners (EDF, ONERA, IFP, SNECMA, ALSTOM, RENAULT,...), and pluridisciplinary collaborations (medicine, biology, mechanical engineering,...) that associates several UTC research units and teams.

The LMAC research team is currently developing numerous other projects with its academic partners in France and abroad.

The PSPC project

VALODIM which covers the optimal value of digestates from methanisation process, is a programme that aims structuring a national sector to valorise digestates by creating and organizing local eco-systems for the production of organic fertilizers.

ITE PIVERT, programme Genesys

MetaLipPro-PL1 constitutes a knowledge acquirement phase that will enable the research teams to improve and complete our knowledge about lipoid metabolism for plants and for yeasts. Another aim is to establish the bases for the purpose of developing a pilot platform for lipid production and extraction.

A European Project

Balaton (telecomm. networks) is designed to produce accurate estimations for the performance of certain real time system performances, with re-tests modelling protocols that are essential in the development of 4th generation networks (4G).

The ANR Project

BANHDITS focuses on non-parametric Bayesian methods as a function of 3strongly implicated axes : complex models, asymptotic study and computational challenges.

Sorbonne Universities cluster project

The general objective of the ROBUST project is to study stability of cell cycles in the presence of a stochastic noise, using observations that relate to growth of "silk sensor whiskers" on mayfly thoraxes.

Imaging mathematically pinpointing the sources for pathological brain activities

Over recent decades, we have seen the arrival of numerous medical imaging techniques, often indeed being complementary with each other. Neurophysiologists, biophysicists, radiologists, image treatment specialists, already collaborate to detect and pinpoint pathological sources in the brain. In a very near future, mathematicians will join forces with these aforementioned experts.

Mathematical treatment allows you to synthesis all the data made available by various techniques currently employed on patients. "Our laboratory" explains Abdellatif El Badia, Director of LMAC, "tackles this issue of pinpointing the sources for pathological brain activities by treating it as an inverse problem. Experimental data allow us to progress 'backwards' up to the sources, using mathematical models." In order to be able to do this, mathematicians use the standard electroencephalogram, the MRI which in essence 'draws' the brain contours and also Near Infra-Red Spectroscopy (NIRS) using luminous sources in the near infra-red spectrum. "When a brain reveals an epileptic activity, there is a concomitant change in the blood flow rates and variations in the concentration levels of oxygenated and deoxygenated haemoglobins and these are revealed by the NIRS - certain areas of the brain will beseen to absorb more light than others and this is where the pathology is occurring", adds Abdellatif El Badia.

It now remains for the LMAC research team to try to combine the experimental findings and data, using a numerical model that would lead on to a specific software package that would allow the doctors to accurately locate and position the exact source of the pathology. Clinical tests at the nearby regional teaching hospital at Amiens (CHU-Amiens) should enable this research to progress fairly rapidly.

Contact

Directeur du laboratoire LMAC
Nikolaos Limnios
Phone : 03 44 23 44 99 | Contact by email

La Recherche à l'UTC

pdf

Download