François Rousseau

François Rousseau
Institut Mines-Télécom | telecom-sudparis.eu · Département Image et Traitement de l'Information

PhD

About

187
Publications
22,857
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
4,536
Citations
Introduction
Additional affiliations
September 2006 - January 2015
University of Strasbourg
Position
  • CNRS research scientist

Publications

Publications (187)
Article
Background Real-time sequences allow functional evaluation of various joint structures during a continuous motion and help understand the pathomechanics of underlying musculoskeletal diseases. Purpose To assess and compare the image quality of the two most frequently used real-time sequences for joint dynamic magnetic resonance imaging (MRI), acqu...
Article
Full-text available
The human brain develops from a smooth cortical surface in early stages of fetal life to a convoluted one postnatally, creating an organized ensemble of folds. Abnormal folding patterns are linked to neurodevelopmental disorders. However, the complex multi-scale interactions involved in cortical folding are not fully known yet. Computational models...
Article
Full-text available
Data assimilation is a key component of operational systems and scientific studies for the understanding, modeling, forecasting and reconstruction of earth systems informed by observation data. Here, we investigate how physics‐informed deep learning may provide new means to revisit data assimilation problems. We develop a so‐called end‐to‐end learn...
Article
Full-text available
The estimation of ocean dynamics is a key challenge for applications ranging from climate modeling to ship routing. State-of-the-art methods relying on satellite-derived altimetry data can hardly resolve spatial scales below ∼100 km. In this work we investigate the relevance of AIS data streams as a new mean for the estimation of the surface curren...
Conference Paper
Full-text available
Variational models are among the state-of-the-art formulations for the resolution of ill-posed inverse problems. Following recent advances in learning-based variational settings, we investigate the end-to-end learning of variational models, more precisely of the regularization term given some observation model, jointly to the associated solver, so...
Conference Paper
Full-text available
Space oceanography missions, especially altimeter missions, have considerably improved the observation of sea surface dynamics over the last decades. They can however hardly resolve spatial scales below ∼ 100km. Meanwhile the AIS (Automatic Identification System) monitoring of the maritime traffic implicitly conveys information on the underlying se...
Article
Full-text available
Earth observation satellite missions provide invaluable global observations of geophysical processes in play in the atmosphere and the oceans. Due to sensor technologies (e.g., infrared satellite sensors), atmospheric conditions (e.g., clouds and heavy rains), and satellite orbits (e.g., polar-orbiting satellites), satellite-derived observations of...
Article
Abdominal anatomy segmentation is crucial for numerous applications from computer-assisted diagnosis to image-guided surgery. In this context, we address fully-automated multi-organ segmentation from abdominal CT and MR images using deep learning. The proposed model extends standard conditional generative adversarial networks. Additionally to the d...
Article
Full-text available
Abnormal cortical folding patterns, such as lissencephaly, pachygyria and polymicrogyria malformations, may be related to neurodevelopmental disorders. In this context, computational modeling is a powerful tool to provide a better understanding of the early brain folding process. Recent studies based on biomechanical modeling have shown that mechan...
Article
Full-text available
Objective: Studies of motor outcome after Neonatal Arterial Ischemic Stroke (NAIS) often rely on lesion mapping using MRI. However, clinical measurements indicate that motor deficit can be different than what would solely be anticipated by the lesion extent and location. Because this may be explained by the cortical disconnections between motor are...
Article
The binary partition tree (BPT) allows for the hierarchical representation of images in a multiscale way, by providing a tree of nodes corresponding to image regions. In particular, cuts of a BPT can be interpreted as segmentations of the associated image. Building the BPT of an image then constitutes a relevant preliminary step for optimization-ba...
Preprint
BACKGROUND Duchenne Muscular Dystrophy (DMD), the most common neuromuscular disease in children, is a severe, progressive disease that affects skeletal muscle. Abnormal gait patterns in children with DMD result from compensatory adaptations of their locomotor system to maintain free ambulation in response to the slow, progressive muscle weakness, c...
Preprint
Full-text available
This paper addresses variational data assimilation from a learning point of view. Data assimilation aims to reconstruct the time evolution of some state given a series of observations, possibly noisy and irregularly-sampled. Using automatic differentiation tools embedded in deep learning frameworks, we introduce end-to-end neural network architectu...
Technical Report
Full-text available
This report presents a synthesis of outcome of the workshop on AI for Ocean, Atmosphere and Climate held in Brest on January 2020 in the framework of LEFE/MANU project IA-OAC, ANR project Melody and AI chair Oceanix with the additional support of Isblue. This report provides a short description of the outcome of each of the 9 working groups which m...
Preprint
Full-text available
Designing appropriate variational regularization schemes is a crucial part of solving inverse problems, making them better-posed and guaranteeing that the solution of the associated optimization problem satisfies desirable properties. Recently, learning-based strategies have appeared to be very efficient for solving inverse problems, by learning di...
Article
Full-text available
Dynamic magnetic resonance imaging (MRI) is a non-invasive method that can be used to increase the understanding of the pathomechanics of joints. Various types of real-time gradient echo sequences used for dynamic MRI acquisition of joints include balanced steady-state free precession sequence, radiofrequency-spoiled sequence, and ultra-fast gradie...
Article
Full-text available
Functional MRI is increasingly being used in the assessment of brain activation and connectivity following stroke. Many of these studies rely on the Blood Oxygenation Level Dependent (BOLD) contrast. However, the stability, as well as the accuracy of the BOLD response to motor tasks in the ipsilesional hemisphere, remains ambiguous. In this work, t...
Preprint
Full-text available
The log Euclidean polyrigid registration framework provides a way to smoothly estimate and interpolate poly-rigid/affine transformations for which the invertibility is guaranteed. This powerful and flexible mathematical framework is currently being used to track the human joint dynamics by first imposing bone rigidity constraints in order to synthe...
Article
Full-text available
This paper addresses the understanding and characterization of residual networks (ResNet), which are among the state-of-the-art deep learning architectures for a variety of supervised learning problems. We focus on the mapping component of ResNets, which map the embedding space toward a new unknown space where the prediction or classification can b...
Article
Background and objective: One of the main issues in the analysis of clinical neonatal brain MRI is the low anisotropic resolution of the data. In most MRI analysis pipelines, data are first re-sampled using interpolation or single image super-resolution techniques and then segmented using (semi-)automated approaches. In other words, image reconstru...
Article
Full-text available
We propose a novel method to quantify brain growth in 3 arbitrary orthogonal directions of the brain or its sub-regions through linear registration. This is achieved by introducing a 9 degrees of freedom (dof) transformation called anisotropic similarity which is an affine transformation with constrained scaling directions along arbitrarily chosen...
Preprint
Full-text available
Objective : Abdominal anatomy segmentation is crucial for numerous applications from computer-assisted diagnosis to image-guided surgery. In this context, we address fully-automated multi-organ segmentation from abdominal CT and MR images using deep learning. Methods: The proposed model extends standard conditional generative adversarial networks....
Article
Objective: To assess the evidence of a relationship between muscle MRI and disease severity in Duchenne muscular dystrophy (DMD). Methods: We conducted a systematic review of studies that analyzed correlations between MRI measurements and motor function in patients with DMD. PubMed, Cochrane, Scopus, and Web of Science were searched using releva...
Article
The stance and swing phases of the gait cycle are defined by foot strike (FS) and foot off (FO). Accurate determination of these events is thus an essential component of 3D motion recordings processing. Several methods have been developed for the automatic detection of these events (based on the heuristics of 3D marker position, velocity and accele...
Preprint
Full-text available
For numerous domains, including for instance earth observation, medical imaging, astrophysics,..., available image and signal datasets often involve irregular space-time sampling patterns and large missing data rates. These sampling properties may be critical to apply state-of-the-art learning-based (e.g., auto-encoders, CNNs,...), fully benefit fr...
Chapter
Online atlasing, i.e. incrementing an atlas with new images as they are acquired, is key when performing studies on databases very large or still being gathered. We propose to this end a new diffeomorphic online atlasing method without having to perform again the atlasing process from scratch. New subjects are integrated following an iterative proc...
Article
Full-text available
The purpose of super-resolution approaches is to overcome the hardware limitations and the clinical requirements of imaging procedures by reconstructing high-resolution images from low-resolution acquisitions using post-processing methods. Super-resolution techniques could have strong impacts on structural magnetic resonance imaging when focusing o...
Conference Paper
Abnormal cortical folding patterns may be related to neurodevelopmental disorders such as lissencephaly and polymicrogyria. In this context, computational modeling is a powerful tool to provide a better understanding of the early brain folding process. Recent studies based on biomechanical modeling have shown that mechanical forces play a crucial r...
Conference Paper
Cortical folding pattern is a main characteristic of the geometry of the human brain which is formed by gyri (ridges) and sulci (grooves). Several biological hypotheses have suggested different mechanisms that attempt to explain the development of cortical folding and its abnormal evolutions. Based on these hypotheses, biomechanical models of corti...
Conference Paper
Magnetic Resonance Imaging (MRI) can provide 3D morphological information on brain structures. Such information is particularly relevant for carrying out morphometric brain analysis, especially in the newborn and in the case of prematurity. However, 3D neonatal MRI acquired in clinical environments are low-resolution, anisotropic images, making seg...
Conference Paper
Spatio-temporal evolution of joint space width (JSW) during motion is of great importance to help with making early treatment plans for degenerative joint diseases like osteoarthritis (OA). These diseases can affect people of all ages leading to an acceleration of joint degeneration and to limitations in the activities of daily living. However, onl...
Conference Paper
Dynamic MRI has made it possible to non-invasively capture the moving human joints in vivo. Real-time Fast Field Echo (FFE) sequences have the potential to reduce the effect of motion artifacts by acquiring the image data within a few milliseconds. However, the short acquisition times affect the temporal resolution of the acquired sequences. In thi...
Preprint
Abnormal cortical folding patterns may be related to neurodevelopmental disorders such as lissencephaly and polymicrogyria. In this context, computational modeling is a powerful tool to provide a better understanding of the early brain folding process. Recent studies based on biomechanical modeling have shown that mechanical forces play a crucial r...
Preprint
Full-text available
Abnormal cortical folding patterns may be related to neurodevelopmental disorders such as lissencephaly and polymicrogyria. In this context, computational modeling is a powerful tool to provide a better understanding of the early brain folding process. Recent studies based on biomechanical modeling have shown that mechanical forces play a crucial r...
Article
Myelin imaging in the central nervous system is essential for monitoring pathologies involving white matter alterations. Various quantitative MRI protocols relying on the modeling of the interactions of water protons with myelinated tissues have shown sensitivities in case of myelin disruption. Some extracted model parameters are more sensitive to...
Article
Full-text available
Although neonatal arterial ischemic Stroke is now well‐studied, its complex consequences on long‐term cortical brain development has not yet been solved. In order to understand the brain development after focal early brain lesion, brain morphometry needs to be evaluated using structural parameters. In this work, our aim was to study and analyse the...
Conference Paper
e present a new method to create a diffeomorphic longitudinal (4D) atlas composed of a set of 3D atlases each representing an average model at a given age. This is achieved by generalizing atlasing methods to produce atlases unbiased with respect to the initial reference up to a rigid transformation and ensuring diffeomorphic deformations thanks to...
Preprint
Full-text available
We propose a novel method to quantify brain growth in 3 arbitrary orthogonal directions of the brain or its sub-regions through linear registration. This is achieved by introducing a 9 degrees of freedom (dof) transformation called anisotropic similarity which is an affine transformation with constrained scaling directions along arbitrarily chosen...
Article
In this paper, we propose a method for non-invasively measuring three-dimensional in vivo kinematics of the ankle joint from a dynamic MRI acquisition of a single range-of-motion cycle. The proposed approach relies on an intensity-based registration method to estimate motion from multi-plane dynamic MRI data. Our approach recovers not only the move...
Preprint
Full-text available
Cortical folding pattern is a main characteristic of the geometry of the human brain which is formed by gyri (ridges) and sulci (grooves). Several biological hypotheses have suggested different mechanisms that attempt to explain the development of cortical folding and its abnormal evolutions. Based on these hypotheses, biomechanical models of corti...
Conference Paper
Online atlasing, i.e. incrementing an atlas with new images as they are acquired, is key when performing studies on databases very large or still being gathered. We propose to this end a new diffeomorphic online atlasing method without having to perform again the atlasing process from scratch. New subjects are integrated following an iterative proc...
Poster
Abstract: https://app.oxfordabstracts.com/stages/123/programme-builder/submission/21064?backHref=/events/123/programme-builder/view/sort/author&view=published
Conference Paper
Full-text available
PURPOSE To evaluate the feasibility of a real-time sequence using balanced Fast Field Echo (bFFE) sequence to study tendon and bone motion during dynamic MRI of the ?inger. METHOD AND MATERIALS A real-time bFFE sequence was used to acquire dynamic data in sagittal plane on 10 index finger, without history of injury or inflammatory rheumatism, on 3T...
Article
Brain structure analysis in the newborn is a major health issue. This is especially the case for preterm neonates, in order to obtain predictive information related to the child development. In particular, the cortex is a structure of interest, that can be observed in magnetic resonance imaging (MRI). However, neonatal MRI data present specific pro...
Poster
Finger motion and function is essential in daily living activities and results from a complex relationship among the anatomical structures of forehand. Injuries and disorders of musculotendinous structures limit joint motion and affect quality of life. To evaluate the state of disorder, it is important to first understand the normal functional fing...
Preprint
Full-text available
This paper addresses the understanding and characterization of residual networks (ResNet), which are among the state-of-the-art deep learning architectures for a variety of supervised learning problems. We focus on the mapping component of ResNets, which map the embedding space towards a new unknown space where the prediction or classification can...
Conference Paper
Cerebral palsy is the leading cause of motor disabilities affecting children. The ankle is the most common equine musculoskeletal strain in children with cerebral palsy. Despite multiple medical and surgical therapies, postoperative recurrence rate is still very high (48%). A major reason for therapy failure is the lack of knowledge of the ankle jo...
Conference Paper
Exploration of myelin content in the brain and spinal cord is essential for monitoring pathologies such as multiple sclerosis. Quantitative MRI methods such as diffusion tensor imaging (DTI) and quantitative magnetization transfer imaging (qMT) already demonstrated efficiency in characterizing demyelination. Ultrashort echo time sequences were pr...
Conference Paper
Conventional slice selection in 2D-UTE sequences is challenging as eddy currents and gradient non-idealities make difficult to achieve an appropriate slice selection and a minimum TE. The sat-UTE sequence proposes a simplification that separates slice selection from excitation, ensuring an easily implementable 2D-UTE sequence. The selection part...
Conference Paper
Short-T2 structures such as myelin and cortical bone often requires the use of inversion-recovery modules in UTE sequences to provide a selective contrast in the components of interest. The sat-UTE sequence allows for an effective slice selection, and avoid issues found in commonly used 2D IR-UTE sequences concerning the use of reshaped half-radio...
Conference Paper
Inversion-Recovery UTE sequences have been used to highlight and quantify short-T2 structures. Covering a 3D volume remains time-consuming, and a trade-off often needs to be found between k-space undersampling and acceptable appearing streaking artifacts. As such, acquiring several radial spokes within a single repetition time represents a supple...
Article
Purpose: To introduce a novel method for long-T2 signal physical suppression in steady-state based on configuration states combination and modulation using diffusion weighting. Its efficiency in yielding a high contrast in short-T2 structures using an ultrashort echo time acquisition module (Diff-UTE) is compared to the adiabatically prepared Inve...