FLI
In this section, you will find internships, thesis, post-doctorates, fixed-term and permanent contract vacancies in the network’s partner laboratories.
Liste des annonces
Postdoctoral position – 2024 Microstructural analysis in late-life depression using diffusion MRI multi-compartments models and tractometry
To advance in the understanding of apathy physiopathology in LLD, we conducted a study which
evaluated the relationship between patterns of motor activity measured by actigraphy, and brain modifications of white matter microstructure.
This study found two patterns of motor activity associated with apathy: a reduced diurnal mean activity, and an early chronotype pattern.
These patterns of motor activity were associated with modified intra-network resting-state functional connectivity in key regions associated with the default-mode, the cingulo-opercular and the frontoparietal network. However, our preliminary work on microstructure metrics estimated from diffusion weighted imaging did not find significant associations between microstructural metrics of white matter and patterns of motor activity after adjustment for multiple. To detect more subtle links such as those between patterns of motor activity and microstructure, our approach needs to be improved.
This project will focus on two major subjects:
- Developing a more accurate estimation and projection of microstructure metrics along the
fiber as well as a new statistical method taking into account the shape complexity of the fibers.
- Extracting more accurate markers of patterns of motor activity measured by actigraphy
The developed approach will be tested on a cohort of patients suffering from late-life depression, with the aim of better estimating the microstructure and thus better understanding the neuronal modifications caused by this disease and apathy.
1-year Postdoctoral Position Disentangled and Controllable Latent Representations for Computer Vision and Medical Imaging
Neuroimaging application :
The unsupervised separation of the healthy latent patterns from the pathological ones is not a trivial task in medical imaging. In neuroimaging, pathological brain signatures of psychiatric or neurodevelopmental disorders are not easily visible with the naked eye, even for experienced radiologists. The automatic identification of prognostic brain signatures of clinical courses would pave the way towards personalised medicine in psychiatry. In this project, following our recent works in contrastive analysis (CA), we wish to discover in an unsupervised way the salient imaging patterns that characterize a target dataset of psychiatric patients compared to a control dataset of healthy subjects, as well as what is common between the two domains. Current SOTA methods are based on VAE. However, they all ignore important constraints/assumptions and the generated images have a rather poor quality, typical of VAEs, which decrease their interpretability and usefulness.
Objectives :
- Study and understand the recent advances in disentanglement of latent spaces;
- Review literature on diffusion models with latent spaces;
- Adapt more recent, well-performing models, such as diffusion models, to the CA framework for neuroimaging
RESEARCH ENGINEER IN MRI at CERMN Caen
The recruited research engineer will contribute to the CrIM research project funded by Nor-mandie Valorisation, a project aiming at developing new diagnostic tools for Magnetic Re-sonance Imaging (MRI) using hyperpolarized xenon. This project is developed through a close collaboration between Centre d'Etude et de Recherches sur le Médicament de Normandie (CERMN, boulevard Henri Becquerel, 14000 CAEN, FRANCE; http://cermn.unicaen.fr/), the Laboratory of Catalysis and Spectrochemistry (LCS, boulevard Maréchal Juin, 14000 CAEN; https://www.lcs.ensicaen.fr/) and the Physiopathology and imaging laboratory of neurological disorders (PhIND, UMRs 1237 INSERM, bou-levard Henri Becquerel, 14000 CAEN) and the Blood and Brain Institute @ Caen-Normandie.
Hyperpolarized xenon MRI is an emerging clinical diagnostic technique currently developed for pulmonary functional imaging, with high potential for applications in brain imaging. Since 2019, the CERMN laboratory has been developing a new generation of contrast agents adapted to xenon. The LCS laboratory has expertise in xenon hyperpolarization applied to the study of materials. The objective of CrIM project is to translate these different skills to preclinical MRI as proof of concept.
The recruited research engineer will be trained to use the xenon polarizer, and will have to perform maintenance of this apparatus. He/she will have to continue the implementation of hyperpolarized xenon at the imaging center GIP CYCERON and will actively participate to the in vivo MRI experiments. Data acquisition will be performed on Bruker apparatus. The recruited research engineer will be based in both LCS and PhIND laboratories, under the scientific responsibility of the principal investigator of the project, and will work in close collaboration with CERMN collaborators and Cyceron platform.
Postdoctoral position in organic synthesis / fluorine-18 radiochemistry (12 months)
CONTEXT
Positron Emission Tomography (PET) imaging is a major diagnostic tool in medicine for the detection and the monitoring of a wide range of diseases. This imaging technique relies on the injection of a molecule labeled with a positron emitter, a short-lived radioisotope such as fluorine-18 (T1/2 = 109 min). The radioactivity can then be precisely and quantitatively detected by a PET camera to visualize the distribution of the labeled compound.
We aim to use PET imaging for the in vivo detection of metal ions, and more particularly Zn2+, using a dedicated labeled peptide. Indeed, metal ions are involved in many fundamental biological processes and an altered regulation of the concentration of metal ions can be associated with acute and long-term diseases. Zinc is of particular interest as it is an essential micronutrient and imbalance of zinc concentration has been associated with diabetes, neurodegenerative diseases, cancers and epilepsy.
PROJECT
This ANR-funded project aims to design novel bioinspired MRI contrast agents (CAs) based on zinc-finger peptides for the detection of Zn2+ in vivo in the extracellular media. These CAs will be designed by collaborators in Grenoble and Orléans and PET imaging will be used to quantify their biodistribution and help establish a proof-of-concept with a dedicated diabetes model.
In this context, the postdoctoral candidate will work on fluorine-18 labeling of zinc-finger peptides developed by our collaborators. To reach this goal, the preparation of [18F]-fluoride/aluminum complexes or the formation of [18F]-fluoride/boron bonds and their use for the radiolabeling of model compounds will be implemented on a radiosynthesis automate. The most successful conditions will then be applied to relevant compounds for PET imaging and biodistribution studies.
Starting date: The position is to be filled from January 2025
PhD thesis on AI for the segmentation of biomarkers of cardiovascular Risk on 3D-CT
Ischemic heart disease, heart failure, and atrial fibrillation are conditions responsible for a very high cardiovascular mortality. The cardiac scanner is an imaging method that is increasingly used in these conditions, even if its analysis remains subjective, sometimes long and dependent on the expertise of the operator. Artificial intelligence methods offer new perspectives in this field because they are potentially fast, partially or even totally automated to quantitatively analyze conventional or complex and new biomarkers by using radiomics concepts. However, all these post treatment methods in full development today need to be rigorously validated before being used in daily patient care.
The project that we propose to carry out here within the framework of a PhD thesis in medical imaging (University of Paris) is located at the interface of the development and the validation of these new approaches, hence the collaboration between the company Siemens Healthineers which develops these new approaches of AI and an academic center which has a big expertise of cardiac imaging, particularly in scanner because this center carries out more than 2500 cardiac scans per annum since more than 10 years
To conduct such study, we propose the following steps:
1. Identification and collection of a base of more than 12,000 patients having benefited from a scanner during a hospitalization in our institution during the last 15 years in order to constitute a base of development and deep learning of the various methods of AI (CAVIAR cohort).
2. Semi-manual segmentation of cardiac structures of interest with conventional tools proposed by Siemens and/or internal to our research institution for the labeling of various biomarkers, candidate for the estimation of cardiovascular risk (coronary, valvular, aortic calcification, size of the left atrium and other cardiac cavities, mediastinal fat ...).
3. Analysis, validation of conventional methods of post treatment of theses biomarkers insufficiently validated by the scanner on these deep learning data bases.
4. Participation in the development of AI methods with Siemens and academic partners.
5. Analysis and validation of AI methods in the estimation of conventional and new biomarkers brought by the AI method itself.
Chargé.e de mission développement d’expérimentation (MEG)
Réaliser l’implantation d’un équipement de MEG nouvelle génération en lien avec un fournisseur français déjà identifié et avec un groupe d’utilisateurs de renommée internationale à fédérer autour du cahier des charges à rédiger en fonction de leurs protocoles de recherche qui concernent soit des thématiques très applicatives telles que l’épilepsie, les comas jusqu’à des recherches très fondamentales sur le fonctionnement neuronal.
Il faudra déterminer le lieu d’implantation au CHU Purpan en documentant s’il convient mieux que ce soit dans les locaux du pavillon Baudot ou dans ceux des services hospitaliers.
Il y aura également un financement complémentaire à demander au conseil régional dans le cadre du défi- clé Quantique, ce qui amènera à identifier des sujets de collaboration avec des laboratoires Toulousains de
Physique fondamentale pour rédiger une demande dans le cadre de l’appel d’offre PRIO d’accompagnement des plateformes en émergence.
Plusieurs équipements du même type seront installés dans d’autres villes dans le cadre de la même opération nationale et vous devrez vous intégrer dans le réseau des chargé.es de mission de FRANCE LIFE IMAGING ce qui optimisera les efforts d’implantation de tous. Certaines villes telles que Marseille ont une vaste expérience de la MEG première génération et vous bénéficierez du retour d’expérience de ces collègues.
Enfin il faudra établir des règles de fonctionnement collectif permettant la labellisation de la plateforme par GENOTOUL.
PhD thesis Deep learning for assisting clinical decisions in brain imaging: trustworthy validation and benchmarking
This PhD project aims at obtaining a general methodological and experimental framework for trustworthy and reproducible validation and benchmarking of deep learning methods in brain imaging and performing large-scale experiments.
Specific objectives are as follows:
- Enrich the framework with more advanced deep learning models, more tasks and more datasets
- Better account for specificities of brain imaging (multiple acquisitions over time, multiple scanners, multiple hospitals, multiple datasets, multiple disorders)
- Propose an adequate inferential statistics framework for both model validation and model comparison
- Perform benchmarking experiments across deep learning (and also standard machine learning) models, tasks, diseases and datasets to create a new standard for the community
- Demonstrate the importance of accounting for brain imaging specificities when evaluating models
- Implement the approaches in open-source software, in particular ClinicaDL so that they can benefit the entire scientific community
Technicien Imagerie Médicale pour la Recherche
Ce poste se situe au cœur d’un projet européen innovant, axé sur l’imagerie médicale et le développement de l’intelligence artificielle. Il vise à améliorer les capacités diagnostiques des professionnels de santé. Vous serez amené à collaborer avec différents acteurs de la recherche, industriels et universitaires.
Responsabilités :
• Réaliser l'analyse quantitative et qualitative des images
• Effectuer la relecture et l'annotation des images médicales centralisées pour assurer leur qualité et leur exactitude.
• Assurer le suivi des images médicales et vérifier leur conformité aux protocoles d'étude.
• Gérer la sauvegarde et la saisie des données dans les bases de données, en garantissant leur intégrité et leur accessibilité.
• Établir les procédures opératoires spécifiques pour le traitement et l'analyse des images.
• Fournir toute documentation nécessaire concernant le processus d'imagerie.
• Organiser les réunions de travail en préparant les documents nécessaires.
• Participer aux réunions de service et contribuer à l'activité générale de l'équipe de recherche.
Research engineer Statistical analysis of longitudinal medical data
You will work in the context of the project REWIND (pRecision mEdecine WIth longitudinal Data), a multicentric project (Paris, Bordeaux, Lyon, Grenoble, Nice) granted via the “Investissement d’avenir” PEPR Santé Numérique. The project will focus on the development of new mathematical and statistical approaches for the analysis of multimodal multiscale longitudinal data.The project will allow the development of a new generation of decision support systems, which will help clinicians at the bedside to make more informed decisions for the patient. They will contribute to the development of precision medicine in several key areas.
You will be in charge of the:
- development of new features (implementation of new models, algorithms, metrics, visualizations),
-software maintenance,
-user support and animation of the community.
In addition, you will be presenting the software at international scientific conferences and other events, and you will contribute to ambitious medical studies, by using Leaspy on large databases of patients, contributing to the interpretation of results and providing assistance to users.
Data Scientist R&D : Analyse automatique d’images médicales pour l’aide au suivi oncologique
Vous serez responsable de la recherche, du développement, de
l'implémentation et de la maintenance des modèles de machine learning, ainsi que de leur intégration
dans les produits logiciels de Guerbet. Vous travaillerez en étroite collaboration avec les radiologues et les équipes cliniques pour comprendre les besoins spécifiques du domaine médical, afin de développer des solutions pertinentes et efficaces.
Responsabilités :
• Innover, concevoir et optimiser les systèmes de traitement de données et les modèles
d'intelligence artificielle.
• Mener des recherches et des développements en machine learning.
• Implémenter et maintenir les modèles de machine learning.
• Intégrer les modèles de machine learning dans les produits logiciels.
• Collaborer étroitement avec les radiologues et les équipes cliniques.
• Comprendre les besoins spécifiques du domaine médical pour développer des solutions
pertinentes et efficaces.