PostDoc position at IRIT Lab Toulouse France

IAPA Project
Infrastructure to Access, Share and Analyse Biomedical Data

Context: The IAPA project is funded by the Paul Sabatier University and
includes the ICR (Institut Claudius Regaud, Hospital and Research Centre
Institute against Cancer) as a partner.
This project aims at building up a software and hardware infrastructure to
allow for the cross-correlation of biomedical data coming from very diverse
sources. The objective is to head for an infrastructure that will allow the
synthesis and the cross-analysis of patient data in order to discover new
diagnoses, therapies and treatments against cancer.
Context of the work:
Medical data are from very different nature, coming from heterogeneous
sources distributed among the territories. An integrated medical system,
able to find all the possible data for one patient, does still not exist.
Many obstacles slowed the process: political, ethical, financial and
informatics problems. This latter dimension of the problem is at the heart
of the project. Available data are nowadays numerous, heterogeneous,
multimodal, time stamped, structured and unstructured, ...:
- administrative information about the patients (age, gender, address, ...)
- clinical analysis (blood test, glucose and cholesterol rate, ...)
- regulations and results of therapies
- images from different nature (PET, MRI, X-Ray)
- gene expressions
To sum up the state of medical information, one can see the following: The
medical practice generates more and more diverse information (text, images,
video, tables, ...), structured or not, raw data or data coming from a
analysis process, stored or not in specialized data centres. The combination
of these data proves however very useful in various fields of exercise of
medicine (research, learning). Paradoxically, while the potentially useful
data do not cease growing in quantity as in quality, it becomes increasingly
difficult, even impossible to control and access them as a whole.

PostDoc subject:
The PostDoc to be recruited will have to work on ONE of the following
subjects in the IAPA project: 

1 - modelling of the data available within a common framework, whatever its
modality (clinical images, data, gene expressions...). Then this modelling
will allow to build up tools for indexing and searching of data from their
storage centres.  The objective of this PostDoc is thus to propose solutions
making it possible to better apprehend the heterogeneity on the level of the
contents of information (heterogeneity in terms of structure and formats of
the data) as well as on the level of access methods to these data. For this,
a proposal is to investigate the problem from 3 different axes: First is
concerned with the definition of a common reference pattern (terminology and
ontology) allowing for the description and annotation of the data. Second is
interested in the description of the data themselves (description of the
metadata associated with them).

2 - medical data mining. The candidate in this part will have to make first
a state of the art of data mining technology that could be used to find some
links between all the available data, and then to design (or use existing
tools) such correlation extraction infrastructure. The aim is to make appear
spontaneously some new knowledge, for instance concerning the therapy
obtained using identified drugs over some (population of) patients.
Multidimensional analysis or rule associations sets, among others can give
some answers to this problematic.
3 - medical data management, especially accessing and requesting medical
data on a grid. The work here consists of building on top of an existing
grid middleware an efficient data management system specifically set for
medical data, with special interests in security and request optimization. 4
- grid medical workflow management. In this part, the candidate will work
collaboratively with medical doctors to establish the workflow of their work
and to map this workflow to a grid infrastructure and the services
associated with their tasks. Using already existing grid workflow
middleware, the work will consist of designing the
architecture/infrastructure that will allow the future integration of the
different services needed in the project in a comprehensive way.

When applying, the candidate should indicate in which field he/she wants to
work.
Scientific Background:
1. PhD in Computer Science in the areas of: Data representations,
ontologies, data grid, data indexing, datamining
2. Working experience in the field of Medical Informatics, being familiar
with medical terminologies is a plus. 3. Fluency in English. French is a plus.
Duration and Salary:
1. 18 months
2. 2100 euros / months, gross salary

Conditions to apply:
This position is opened to foreign applicants ONLY. The conditions are thus:
 1. Not being a French citizen
2. Not being more than 35 years old
3. Not having completed a PhD in France
4. Not being employed in a Research centre in France 5. Apply until March
31st, 2008.
- Contact :
Prof. Mohand Boughanem, bougha at irit.fr
Prof. Jean-Marc Pierson, pierson at irit.fr