Stéphanie Allassonnière

Paris Descartes, CPSC | Paris 5 · Centre de Recherche des Cordeliers (UMR_S 872)

Longitudinal medical image data are becoming increasingly important for monitoring patient progression. However, such datasets are often small, incomplete, or have inconsistencies between observations. Thus, we propose a generative model that not only produces continuous trajectories of fully synthetic patient images, but also imputes missing data...

La prédiction de l’agressivité des tumeurs neuroendocrines hypophysaires (PitNETs) est un besoin aujourd’hui fondamental. La première classification multi-omique des PitNETs a démontré une structure basée sur le lignage [1]. Les signatures d’agressivité semblent bien moins déterminantes dans la classification, potentiellement spécifiques de chaque...

This paper introduces a new interpretation of the Variational Autoencoder framework by taking a fully geometric point of view. We argue that vanilla VAE models unveil naturally a Riemannian structure in their latent space and that taking into consideration those geometrical aspects can lead to better interpolations and an improved generation proced...

Stochastic algorithms are at [...]

In this paper, we propose a new method to perform data augmentation in a reliable way in the High Dimensional Low Sample Size (HDLSS) setting using a geometry-based variational autoencoder (VAE). Our approach combines the proposal of 1) a new VAE model, the latent space of which is modeled as a Riemannian manifold and which combines both Riemannian...

In recent years, deep generative models have attracted increasing interest due to their capacity to model complex distributions. Among those models, variational autoencoders have gained popularity as they have proven both to be computationally efficient and yield impressive results in multiple fields. Following this breakthrough, extensive research...

Matrix data sets arise in network analysis for medical applications, where each network belongs to a subject and represents a measurable phenotype. These large dimensional data are often modeled using lower-dimensional latent variables, which explain most of the observed variability and can be used for predictive purposes. In this paper, we provide...

Modelling the mean and variability in a population of images, a task referred to as atlas estimation, remains very challenging, especially in a clinical setting where deformations between images can occur at multiple scales. In this paper, we introduce a coarse-to-fine strategy for atlas estimation in the Large Deformation Diffeomorphic Metric Mapp...

The Expectation Maximisation (EM) algorithm is widely used to optimise non-convex likelihood functions with latent variables. Many authors modified its simple design to fit more specific situations. For instance, the Expectation (E) step has been replaced by Monte Carlo (MC), Markov Chain Monte Carlo or tempered approximations, etc. Most of the wel...

Population monitoring is a challenge in many areas such as public health and ecology. We propose a method to model and monitor population distributions over space and time, in order to build an alert system for spatio-temporal data changes. Assuming that mixture models can correctly model populations, we propose a new version of the Expectation-Max...

Objective: Cushing’s syndrome represents a state of excessive glucocorticoids related to glucocorticoid treatments or to endogenous hypercortisolism. Cushing’s syndrome is associated with high morbidity, with significant inter-individual variability. Likewise, adrenal insufficiency is a life-threatening condition of cortisol deprivation. Currently,...

Conditional correlation networks, within Gaussian Graphical Models (GGM), are widely used to describe the direct interactions between the components of a random vector. In the case of an unlabelled Heterogeneous population, Expectation Maximisation (EM) algorithms for Mixtures of GGM have been proposed to estimate both each sub-population’s graph a...

We propose a new efficient way to sample from a Variational Autoencoder in the challenging low sample size setting (A code is available at https://github.com/clementchadebec/Data_Augmentation_with_VAE-DALI). This method reveals particularly well suited to perform data augmentation in such a low data regime and is validated across various standard a...

The Expectation-Maximization Algorithm (EM) is a widely used method allowing to estimate the maximum likelihood of models involving latent variables. When the Expectation step cannot be computed easily, one can use stochastic versions of the EM such as the Stochastic Approximation EM. This algorithm, however, has the drawback to require the joint l...

Objective: To describe a real-time decision support system (DSS), named SONIO, to assist ultrasound-based prenatal diagnosis and to assess its performance using a clinical database of precisely phenotyped postmortem examinations. Population and Methods: This DSS is knowledge-based and comprises a dedicated thesaurus of 294 syndromes and diseases. I...

In this paper, we propose a new method to perform data augmentation in a reliable way in the High Dimensional Low Sample Size (HDLSS) setting using a geometry-based variational autoencoder. Our approach combines a proper latent space modeling of the VAE seen as a Riemannian manifold with a new generation scheme which produces more meaningful sample...

This article describe the Inria, Inserm, Univ. de Paris project team HeKA.

Network analysis provides a rich framework to model complex phenomena, such as human brain connectivity. It has proven efficient to understand their natural properties and design predictive models. In this paper, we study the variability within groups of networks, i.e., the structure of connection similarities and differences across a set of networ...

Alzheimer’s disease (AD) is characterized by the progressive alterations seen in brain images which give rise to the onset of various sets of symptoms. The variability in the dynamics of changes in both brain images and cognitive impairments remains poorly understood. This paper introduces AD Course Map a spatiotemporal atlas of Alzheimer’s disease...

While much efforts have been focused on improving Variational Autoencoders through richer posterior and prior distributions, little interest was shown in amending the way we generate the data. In this paper, we develop two non \emph{prior-dependent} generation procedures based on the geometry of the latent space seen as a Riemannian manifold. The f...

The expectation–maximization (EM) algorithm is a powerful computational technique for maximum likelihood estimation in incomplete data models. When the expectation step cannot be performed in closed form, a stochastic approximation of EM (SAEM) can be used. The convergence of the SAEM toward critical points of the observed likelihood has been prove...

Given repeated observations of several subjects over time, i.e. a longitudinal data set, this paper introduces a new model to learn a classification of the shapes progression in an unsupervised setting: we automatically cluster a longitudinal data set in different classes without labels. Our method learns for each cluster an average shape trajector...

A patient suffering from a rare disease in France has to wait an average of two years before being diagnosed. This medical wandering is highly detrimental both for the health system and for patients whose pathology may worsen. There exists an efficient network of Centres of Reference for Rare Diseases (CRMR), but patients are often referred to thes...

Variational auto-encoders (VAEs) have proven to be a well suited tool for performing dimensionality reduction by extracting latent variables lying in a potentially much smaller dimensional space than the data. Their ability to capture meaningful information from the data can be easily apprehended when considering their capability to generate new re...

Conditional correlation networks, within Gaussian Graphical Models (GGM), are widely used to describe the direct interactions between the components of a random vector. In the case of an unlabelled Heterogeneous population, Expectation Maximisation (EM) algorithms for Mixtures of GGM have been proposed to estimate both each sub-population's graph a...

The Expectation Maximisation (EM) algorithm is widely used to optimise non-convex likelihood functions with hidden variables. Many authors modified its simple design to fit more specific situations. For instance the Expectation (E) step has been replaced by Monte Carlo (MC) approximations, Markov Chain Monte Carlo approximations, tempered approxima...

Gaussian Graphical Models (GGM) are often used to describe the conditional correlations between the components of a random vector. In this article, we compare two families of GGM inference methods: nodewise edge selection and penalised likelihood maximisation. We demonstrate on synthetic data that, when the sample size is small, the two methods pro...

Gaussian Graphical Models (GGM) are often used to describe the conditional correlations between the components of a random vector. In this article, we compare two families of GGM inference methods: nodewise edge selection and penalised likelihood maximisation. We demonstrate on synthetic data that, when the sample size is small, the two methods pro...

In this work, we study the problem of inferring a discrete probability distribution using both expert knowledge and empirical data. This is an important issue for many applications where the scarcity of data prevents a purely empirical approach. In this context, it is common to rely first on an a priori from initial domain knowledge before proceedi...

In this work we study the problem of inferring a discrete probability distribution using both expert knowledge and empirical data. This is an important issue for many applications where the scarcity of data prevents a purely empirical approach. In this context, it is common to rely first on an initial domain knowledge a priori before proceeding to...

Several methods have been proposed recently to learn spatiotemporal models of shape progression from repeated observations of several subjects over time, i.e. a longitudinal data set. These methods summarize the population by a single common trajectory in a supervised manner. In this paper, we propose to extend such approaches to an unsupervised se...

The ability to predict the progression of biomarkers, notably in NDD, is limited by the size of the longitudinal data sets, in terms of number of patients, number of visits per patients and total follow-up time. To this end, we introduce a data augmentation technique that is able to reproduce the variability seen in a longitudinal training data set...

In this work, we present our various contributions to the objective of building a decision support tool for the diagnosis of rare diseases. Our goal is to achieve a state of knowledge where the uncertainty about the patient's disease is below a predetermined threshold. We aim to reach such states while minimizing the average number of medical tests...

Repeated failures in clinical trials for Alzheimer’s disease (AD) have raised a strong interest for the prodromal phase of the disease. A better understanding of the brain alterations during this early phase is crucial to diagnose patients sooner, to estimate an accurate disease stage, and to give a reliable prognosis. According to recent evidence,...

We introduce a hierarchical model which allows to estimate a group-average piecewise-geodesic trajectory in the Riemannian space of measurements and individual variability. This model falls into the well defined mixed-effect models. The subject-specific trajectories are defined through spatial and temporal transformations of the group-average piece...

We propose a generic Bayesian mixed-effects model to estimate the temporal progression of a biological phenomenon from observations obtained at multiple time points for a group of individuals. The progression is modeled by continuous trajectories in the space of measurements. Individual trajectories of progression result from spatiotemporal transfo...

We introduce a mixed-effects model to learn spatiotemporal patterns on a network by considering longitudinal measures distributed on a fixed graph. The data come from repeated observations of subjects at different time points which take the form of measurement maps distributed on a graph such as an image or a mesh. The model learns a typical group-...

We tackle the problem of template estimation when data have been randomly deformed under a group action in the presence of noise. In order to estimate the template, one often minimizes the variance when the influence of the transformations have been removed (computation of the Fréchet mean in the quotient space). The consistency bias is defined as...

We tackle the problem of template estimation when data have been randomly transformed under an isometric group action in the presence of noise. In order to estimate the template, one often minimizes the variance when the influence of the transformations have been removed (computation of the Fr{\'e}chet mean in quotient space). The consistency bias...

In this article we study the consistency of the template estimation with the Fr{\'e}chet mean in the quotient space when the observations are shifted by a group action. We show that in most cases this estimator is actually inconsistent. We exhibit a sufficient condition for this inconsistency, which amounts to the folding of the distribution of the...

When we have a deformation group acting on a vector space of observations, these data are not anymore elements of our space but rather orbits for the group action we consider. If the data are generated from an unknown template with noise, to estimate this template, one may want to minimise the variance in the quotient set. In this article we study...

Examining the dynamics of stroke ischemia is limited by the standard use of 2D-volume or voxel-based analysis techniques. Recently developed spatiotemporal models such as the 4D metamorphosis model showed promise for capturing ischemia dynamics. We used a 4D metamorphosis model to evaluate acute ischemic stroke lesion morphology from the acute diff...

In this paper we introduce a diffeomorphic constraint on the deformations considered in the deformable Bayesian mixed effect template model. Our approach is built on a generic group of diffeomorphisms, which is parameterized by an arbitrary set of control point positions and momentum vectors. This enables us to estimate the optimal positions of con...

Mixed-effects models provide a rich theoretical framework for the analysis of longitudinal data. However, when used to analyze or predict the progression of a neurodegenerative disease such as Alzheimer's disease, these models usually do not take into account the fact that subjects may be at different stages of disease progression and the interpret...

We propose a Bayesian mixed-effects model to learn typical scenarios of changes from longitudinal manifold-valued data , namely repeated measurements of the same objects or individuals at several points in time. The model allows to estimate a group-average trajectory in the space of measurements. Random variations of this trajectory result from spa...

Computerized anatomical atlases play an important role in medical image analysis. While an atlas usually refers to a standard or mean image also called template, which presumably represents well a given population, it is not enough to characterize the observed population in detail. A template image should be learned jointly with the geometric varia...

We extend the image-to-image metamorphosis into constrained longitudi- nal metamorphosis. We apply it to estimate an evolution scenario, in patients with acute ischemic stroke, of both scattered and solitary ischemic lesions vis- ible on serial MR perfusion weighted imaging from acute to subacute stages. We then estimate a patient-specific residual...

Automatic anatomical brain image segmentation is still a challenge. In particular, algorithms have to address the partial volume effect (PVE) as well as the variability of the gray level of internal brain structures which may appear closer to gray matter (GM) than white matter (WM). Atlas based segmentation is one solution as it brings prior inform...

[This corrects the article DOI: 10.1016/j.nicl.2012.10.003.].

Traditional analyses of Functional Magnetic Resonance Imaging (fMRI) use little anatomical information. The registration of the images to a template is based on the individual anatomy and ignores functional information; subsequently detected activations are not confined to gray matter (GM). In this paper, we propose a statistical model to estimate...

Coupling time series of MR Images with reaction–diffusion-based models has provided interesting ways to better understand the proliferative-invasive aspect of glial cells in tumors. In this paper, we address a different formulation of the inverse problem: from a single time point image of a non-swollen brain tumor, estimate the tumor source locatio...

There is considerable interest in how well acute DWI can predict the infarct on final T2-w imaging at >1month after stroke. We estimated the 4D dynamic change, from the acute DWI to the final T2-w lesion, to assess DWI predictive ability. Our model dynamically simulates the evolution of uni- or multi-component lesions. We applied a 4D metamorphosis...

The spatiotemporal evolution of stroke lesions, from acute injury to final tissue damage, is complex. Diffusion-weighted (DWI) and perfusion-weighted (PWI) imaging is commonly used to detect early ischemic changes and attempts to distinguish between permanently damaged and salvageable tissues. To date, 2D and 3D measures of diffusion/perfusion regi...

High dimensional data are more and more frequent in many application fields. It becomes particularly important to be able to extract meaningful features from these data sets. Deformable template model is a popular way to achieve this. This paper is a review on the statistical aspects of this model as well as its generalizations. We describe the dif...

This paper introduces a new parameterization of diffeomorphic deformations for the characterization of the variability in image ensembles. Dense diffeomorphic deformations are built by interpolating the motion of a finite set of control points that forms a Hamiltonian flow of self-interacting particles. The proposed approach estimates a template im...

Over the last 15 years, basic thresholding techniques in combination with standard statistical correlation-based data analysis tools have been widely used to investigate different aspects of evolution of acute or subacute to late stage ischemic stroke in both human and animal data. Yet, a wave of biology-dependent and imaging-dependent issues is st...

Estimation in the deformable template model is a big challenge in image analysis. The issue is to estimate an atlas of a population. This atlas contains a template and the corresponding geometrical variability of the observed shapes. The goal is to propose an accurate algorithm with low computational cost and with theoretical guaranties of relevanc...

The decomposition of a sample of images on a relevant subspace is a recurrent problem in many different fields from Computer Vision to medical image analysis. We propose in this paper a new learning principle and implementation of the generative decomposition model generally known as noisy ICA (for independent component analysis) based on the SAEM...

Existing computer simulations of aircraft infrared signature (IRS) do not account for dispersion induced by uncertainty on input parameters, such as aircraft aspect angles and meteorological conditions. As a result, they are of little use to quantify the detection performance of IR optronic systems: in this case, the scenario encompasses a lot of p...

Existing computer simulations of aircraft infrared signature do not account for the dispersion induced by uncertainty on input data, such as aircraft aspect angles and meteorological conditions. As a result, they are of little use to estimate the detection performance of IR optronic systems: in that case, the scenario encompasses a lot of possible...

The problem of the definition and the estimation of generative models based on deformable templates from raw data is of particular importance for modelling non aligned data affected by various types of geometrical variability. This is especially true in shape modelling in the computer vision community or in probabilistic atlas building for Computat...

The estimation of probabilistic deformable template models in computer vision or of probabilistic atlases in Computational Anatomy are core issues in both fields. A first coherent statistical framework where the geometrical variability is modelled as a hidden random variable has been given by Allassonniere, Amit and Trouve in (1). They introduce a...