Skills

Python Research Training Deep Learning Databases git

Job Specifications

Sujet : Apprentissage de la reprï¿½ntation pour la sï¿½ction de donnï¿½ d'imagerie mï¿½cale

Contribuez au futur de la pratique mï¿½cale en participant ï¿½'avancement des soins de santï¿½umï¿½ques. L'ambition stratï¿½que de Dassault Systï¿½s est de rï¿½lutionner les sciences de la vie grï¿½ au dï¿½loppement du jumeau virtuel, transformant ainsi la mï¿½cine de demain.

Vous rejoindrez le dï¿½rtement Corporate Research and Science, au sein d'une ï¿½ipe dï¿½ï¿½au dï¿½loppement d'outils transversaux basï¿½sur l'IA pour la recherche mï¿½cale. Cet environnement dynamique vous permettra d'ï¿½e exposï¿½ un large ï¿½ntail de sujets de recherche pionniers et vous offrira des opportunitï¿½d'apprentissage et d'ï¿½nouissement.

Vos missions

Dans un contexte oï¿½ lï¿½acquisition de donnï¿½ mï¿½cales est un facteur limitant dans lï¿½entraï¿½ment de modï¿½s de segmentation, il devient essentiel de valoriser les donnï¿½ dï¿½imagerie mï¿½cale dï¿½ collectï¿½. Ces jeux de donnï¿½, souvent complexes et dï¿½urvus de sï¿½ntique explicite, posent un dï¿½ en matiï¿½ de sï¿½ction et dï¿½interprï¿½tion. De plus, la redondance frï¿½ente dans les grands corpus dï¿½imagerie nï¿½ssite des mï¿½odes capables dï¿½extraire des reprï¿½ntations robustes et significatives.

Ce stage a pour objectif lï¿½exploration dï¿½un large corpus dï¿½images mï¿½cales afin dï¿½apprendre des reprï¿½ntations pertinentes, en vue de concevoir des systï¿½s avancï¿½de recherche et d'analyse.

Les thï¿½tiques porteront sur plusieurs axes mï¿½odologiques, incluant :

Apprentissage de reprï¿½ntations (reprï¿½ntation learning)
ï¿½hantillonnage intelligent
Apprentissage actif
Exploration et manipulation de bases de donnï¿½ mï¿½cales

Vos missions incluront :

La rï¿½ction et la prï¿½ntation d'un ï¿½t de l'art sur les sujets abordï¿½/li>
La conception et la mise en place de nouvelles mï¿½odes
Lï¿½implï¿½ntation et lï¿½ï¿½luation des mï¿½odes choisies
Lï¿½exploration des techniques de reprï¿½ntation dans le domaine mï¿½cal

Vous aurez la possibilitï¿½e mener un projet de recherche de maniï¿½ autonome, dans un environnement dynamique, animï¿½ar le partage de connaissances et la collaboration.

ï¿½ English Version ï¿½

Topic: Representation learning for medical imaging data selection

Help shape the future of medical practice by contributing to the advancement of digital healthcare. The strategic ambition of Dassault Systï¿½s is to revolutionize life sciences through the development of the virtual twin, thereby transforming the medicine of tomorrow.

You'll be joining the Corporate Research and Science department, working within a team dedicated to developing cross-functional, AI-driven tools for medical research. This dynamic environment offers exposure to a wide range of pioneering research topics and provides opportunities for learning and growth.

Your Responsibilities

In a context where acquiring medical data is a limiting factor in training segmentation models, it becomes essential to leverage already collected medical imaging data. These datasets, often complex and lacking explicit semantics, present challenges in terms of selection and interpretation. Moreover, the frequent redundancy in large imaging corpora calls for methods capable of extracting robust and meaningful representations.

The goal of this internship is to explore a large corpus of medical images to learn relevant representations, with the aim of designing advanced systems for search and analysis.

The internship will focus on several methodological areas, including:

Representation learning
Sampling strategies
Active learning
Exploration and manipulation of medical databases

Your responsibilities will include:

Writing and presenting a state-of-the-art review on the topics covered
Designing and implementing new methods
Implementing and evaluating the selected approaches
Exploring representation techniques in the medical domain

You will have the opportunity to lead a research project independently, within a dynamic environment driven by knowledge sharing and peer collaboration.

Vos Qualifications

Actuellement en Master 2 / Bac+5 en Ecole dï¿½ingï¿½eur ou Universitï¿½avec une spï¿½alisation en Informatique, deep learning ou mathï¿½tiques appliquï¿½, vous recherchez un stage.

Compï¿½nces techniques souhaitï¿½ :

De solides connaissances en deep learning et en apprentissage des reprï¿½ntations
Une maï¿½ise de Python et des bonnes pratiques de gestion de projet logiciel (environnements, Git, tests, etc.)
Connaissances en imagerie biomï¿½cale seront considï¿½es comme un plus

De nature curieuse, autonome et force de proposition, lï¿½innovation et la recherche sont des moteurs pour vous.

Vous apprï¿½ez la collaboration et savez prï¿½nter les rï¿½ltats de votre travail.

Nous rejoindre c'est aussi

Intï¿½er une entreprise scientifique au cï¿½ur de lï¿½innovation technologique, portï¿½par une forte croissance depuis plus de 40 ans

Principaux avantages et bï¿½fices :

Environnement multiculturel
Cad

About the Company

Dassault Systèmes, the 3DEXPERIENCE Company, is a catalyst for human progress. We provide business and people with collaborative virtual environments to imagine sustainable innovations. By creating virtual twin experiences of the real world with our 3DEXPERIENCE platform and applications, our customers can redefine the creation, production and life-cycle-management processes of their offer and thus have a meaningful impact to make the world more sustainable. The beauty of the Experience Economy is that it is a human-centered... Know more