Search

In neuroimaging research, the utilization of multi-centric analyses is crucial for obtaining sufficient sample sizes and representative clinical populations. Data harmonization techniques are typically part of the pipeline in multi-centric studies to address systematic biases and ensure the comparability of the data. However, most multi-centric studies require centralized data, which may result in exposing individual patient information. This poses a significant challenge in data governance, leading to the implementation of regulations such as the GDPR and the CCPA, which attempt to address these concerns but also hinder data access for researchers. Federated learning offers a privacy-preserving alternative approach in machine learning, enabling models to be collaboratively trained on decentralized data without the need for data centralization or sharing.In this paper, we present Fed-ComBat, a federated framework for batch effect harmonization on decentralized data. Fed-ComBat extends existing centralized linear methods, such as ComBat and distributed as d-ComBat, and nonlinear approaches like ComBat-GAM in accounting for potentially nonlinear and multivariate covariate effects. By doing so, Fed-ComBat enables the preservation of nonlinear covariate effects without requiring centralization of data and without prior knowledge of which variables should be considered nonlinear or their interactions, differentiating it from ComBat-GAM. We assessed Fed-ComBat and existing approaches on simulated data and multiple cohorts comprising healthy controls (CN) and subjects with various disorders such as Parkinson’s disease (PD), Alzheimer’s disease (AD), and autism spectrum disorder (ASD).Results indicate that Fed-ComBat outperforms centralized ComBat in the presence of nonlinear effects and is comparable to centralized methods such as ComBat-GAM. Using synthetic data, Fed-ComBat is able to better reconstruct the target unbiased function by 35% (RMSE = 0.5952) with respect to d-ComBat (RMSE = 0.9162) and 12% with respect to our proposal to federate ComBat-GAM, d-ComBat-GAM (RMSE= 0.6751) and exhibits comparable results on MRI-derived phenotypes to centralized methods as ComBat-GAM without the need of prior knowledge on potential nonlinearities.

A semiartificial photosynthesis approach that utilizes enzymes for solar fuel production relies on efficient photosensitizers that should match the enzyme activity and enable long-term stability. Polymer dots (Pdots) are biocompatible photosensitizers that are stable at pH 7 and have a readily modifiable surface morphology. Therefore, Pdots can be considered potential photosensitizers to drive such enzyme-based systems for solar fuel formation. This work introduces and unveils in detail the interaction within the biohybrid assembly composed of binary Pdots and the HydA1 [FeFe]-hydrogenase from Chlamydomonas reinhardtii. The direct attachment of hydrogenase on the surface of toroid-shaped Pdots was confirmed by agarose gel electrophoresis, cryogenic transmission electron microscopy (Cryo-TEM), and cryogenic electron tomography (Cryo-ET). Ultrafast transient spectroscopic techniques were used to characterize photoinduced excitation and dissociation into charges within Pdots. The study reveals that implementation of a donor-acceptor architecture for heterojunction Pdots leads to efficient subpicosecond charge separation and thus enhances hydrogen evolution (88 460 mu mol(H2).g(H2ase)(-1).h(-1)). Adsorption of [FeFe]-hydrogenase onto Pdots resulted in a stable biohybrid assembly, where hydrogen production persisted for days, reaching a TON of 37 500 +/- 1290 in the presence of a redox mediator. This work represents an example of a homogeneous biohybrid system combining polymer nanoparticles and an enzyme. Detailed spectroscopic studies provide a mechanistic understanding of light harvesting, charge separation, and transport studied, which is essential for building semiartificial photosynthetic systems with efficiencies beyond natural and artificial systems.

Biohybrid technologies like semiartificial photosynthesis are attracting increased attention, as they enable the combination of highly efficient synthetic light-harvesters with the self-healing and outstanding performance of biocatalysis. However, such systems are intrinsically complex, with multiple interacting components. Herein, we explore a whole-cell photocatalytic system for hydrogen (H

Small molecules in solution may interfere with mechanistic investigations, as they can affect the stability of catalytic states and produce off-cycle states that can be mistaken for catalytically relevant species. Here we show that the hydride state (H

The metalloenzymes hydrogenases are natures redox catalyst for molecular hydrogen (H2) uptake and production. In [FeFe]-hydrogenases catalysis is facilitated at a unique diiron site. The two iron ions are connected by an azadithiolate bridging ligand (−SCH2NHCH2S−, ADT) and coordinated by a CO and CN− ligand each. Upon reduction of this diiron site the ligand arrangement of a third CO ligand is heavily debated. Two main geometries that differ by the origin of a bridging ligand in between the iron ions and the occupation of a free binding site are discussed. In a first scenario a hydride is bound in the bridging position and the third CO ligand occupies the open coordination site resulting in a geometry not favourable for catalysis. Experimental data indicative of the latter case has been collected at room temperature. In a second scenario the third CO ligand occupies a position in between the two iron ions resulting in a free coordination site favoured for H2 catalysis. Here up to date, experimental data indicative of this geometry, in particular the infrared band of the bridging CO (µCO) ligand, has only been observed at cryogenic temperatures questioning its catalytic relevance. In this study, we characterise the diiron site reduced intermediate that features the µCO geometry and preserves the open coordination site via FTIR spectroscopy at room temperature. The observation of this proposed diiron site reduced reaction intermediate with µCO geometry at room temperature makes its involvement in H2 catalysis more likely.

The progression of neurodegenerative diseases, such as Alzheimer’s Disease, is the result of complex mechanisms interacting across multiple spatial and temporal scales. Understanding and predicting the longitudinal course of the disease requires harnessing the variability across different data modalities and time, which is extremely challenging. In this paper, we propose a model based on recurrent variational autoencoders that is able to capture cross-channel interactions between different modalities and model temporal information. These are achieved thanks to its multi-channel architecture and its shared latent variational space, parametrized with a recurrent neural network. We evaluate our model on both synthetic and real longitudinal datasets, the latter including imaging and non-imaging data, with 𝑁 = 897 subjects. Results show that our multi-channel recurrent variational autoencoder outperforms a set of baselines (KNN, random forest, and group factor analysis) for the task of reconstructing missing modalities, reducing the mean absolute error by 5% (w.r.t. the best baseline) for both subcortical volumes and cortical thickness. Our model is robust to missing features within each modality and is able to generate realistic synthetic imaging biomarkers trajectories from cognitive scores. This work is supported by the European Union’s Horizon 2020 research and innovation programme (grant n◦ 848158). M. Lorenzi is supported by the French government, through the 3IA Côte d’Azur Investments in the Future project managed by the National Research Agency (ANR) (ANR-19-P3IA-0002). G. Piella is supported by ICREA under the ICREA Academia programme. This publication is part of the project PCI2021-122044-2A, funded by the project ERA-NET NEURON Cofund2, by MCIN/AEI/10.13039/501100011033/ and by the European Union “NextGenerationEU”/PRTR. Data collection and sharing for this project was funded by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2- 0012). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and through generous contributions from the following: AbbVie, Alzheimer’s Association; Alzheimer’s Drug Discovery Foundation; Araclon Biotech; BioClinica, Inc.; Biogen; Bristol-Myers Squibb Company; CereSpir, Inc.; Cogstate; Eisai Inc.; Elan Pharmaceuticals, Inc.; Eli Lilly and Company; EuroImmun; F. Hoffmann-La Roche Ltd and its affiliated company Genentech, Inc.; Fujirebio; GE Healthcare; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development, LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Lumosity; Lundbeck; Merck & Co., Inc.; Meso Scale Diagnostics, LLC.; NeuroRx Research; Neurotrack Technologies; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Takeda Pharmaceutical Company; and Transition Therapeutics. The Canadian Institutes of Health Research is providing funds to support ADNI clinical sites in Canada. Private sector contributions are facilitated by the Foundation for the National Institutes of Health ( http://www.fnih.org). The grantee organization is the Northern California Institute for Research and Education, and the study is coordinated by the Alzheimer’s Therapeutic Research Institute at the University of Southern California. ADNI data are disseminated by the Laboratory for Neuro Imaging at the University of Southern California.

Hydrogenases are metalloenzymes that catalyze the reversible oxidation of molecular hydrogen into protons and electrons. For this purpose, [FeFe]-hydrogenases utilize a hexanuclear iron cofactor, the H-cluster. This biologically unique cofactor provides the enzyme with outstanding catalytic activities, but it is also highly oxygen sensitive. Under in vitro conditions, oxygen stable forms of the H-cluster denoted H

To develop and validate a deep learning method of predicting visual function from spectral domain optical coherence tomography (SD-OCT)-derived retinal nerve fiber layer thickness (RNFLT) measurements and corresponding SD-OCT images.Development and evaluation of diagnostic technology.Two deep learning ensemble models to predict pointwise VF sensitivity from SD-OCT images (model 1: RNFLT profile only; model 2: RNFLT profile plus SD-OCT image) and 2 reference models were developed. All models were tested in an independent test-retest data set comprising 2181 SD-OCT/VF pairs; the median of ∼10 VFs per eye was taken as the best available estimate (BAE) of the true VF. The performance of single VFs predicting the BAE VF was also evaluated. The training data set comprised 954 eyes of 220 healthy and 332 glaucomatous participants, and the test data set, 144 eyes of 72 glaucomatous participants. The main outcome measures included the pointwise prediction mean error (ME), mean absolute error (MAE), and correlation of predictions with the BAE VF sensitivity.The median mean deviation was -4.17 dB (-14.22 to 0.88). Model 2 had excellent accuracy (ME 0.5 dB, SD 0.8) and overall performance (MAE 2.3 dB, SD 3.1), and significantly (paired t test) outperformed the other methods. For single VFs predicting the BAE VF, the pointwise MAE was 1.5 dB (SD 0.7). The association between SD-OCT and single VF predictions of the BAE pointwise VF sensitivities was ROur method outperformed standard statistical and deep learning approaches. Predictions of BAEs from OCT images approached the accuracy of single real VF estimates of the BAE.

Disease progression modeling (DPM) of Alzheimer's disease (AD) aims at revealing long term pathological trajectories from short term clinical data. Along with the ability of providing a data-driven description of the natural evolution of the pathology, DPM has the potential of representing a valuable clinical instrument for automatic diagnosis, by explicitly describing the biomarker transition from normal to pathological stages along the disease time axis. In this work we reformulated DPM within a probabilistic setting to quantify the diagnostic uncertainty of individual disease severity in an hypothetical clinical scenario, with respect to missing measurements, biomarkers, and follow-up information. We show that the staging provided by the model on 582 amyloid positive testing individuals has high face validity with respect to the clinical diagnosis. Using follow-up measurements largely reduces the prediction uncertainties, while the transition from normal to pathological stages is mostly associated with the increase of brain hypo-metabolism, temporal atrophy, and worsening of clinical scores. The proposed formulation of DPM provides a statistical reference for the accurate probabilistic assessment of the pathological stage of de-novo individuals, and represents a valuable instrument for quantifying the variability and the diagnostic value of biomarkers across disease stages.

Despite the increasing effort being devoted to the investigation of the link between imaging endophenotypes (IDPs) and genetic determinants (GDs) in Mild Cognitive Impairment (MCI) and Alzheimer’s disease (AD), many issues remain open and deserve investigation. Among these, the role of functional connectivity (FC) is still blurred. This paper aims at shading some light on the topic relying on the ADNI repository (177 patients, out of which 82 MCI and 95 controls). The within/between-network connectivities were derived from individual FC matrices and used as IDPs. Conversely, the GDs consisted of two Polygenic Risk Scores (PRS) that have recently been proven to play a role in AD. A Partial Least Squares (PLS) model equipped with LASSO regularization was finally fitted to the data for associating IDPs and GDs. In the first component, all FC coefficients had the same sign, and were correlated with PRS2. Connectivities involving the dorsal attention (DAN) and frontoparietal control (CON) networks reached the highest weights, while within/between-network FC for the limbic (LIM) were less represented. Overall, the within-network FC values were less pronounced compared to the between-network ones. In the second component, most of the FC features had zero weights. Visual (VIS) and somatomotory (SMN) showed a correlated trend, while being anti-correlated with LIM, CON and default mode network as well as with PRS1. Our findings suggest that the two PRSs correlated with a possible pattern of aberrant within/between-network FC changes occurring in RSNs devoted to higher cognitive functions and more vulnerable in this pathology.

We propose a novel federated learning paradigm to model data variability among heterogeneous clients in multi-centric studies. Our method is expressed through a hierarchical Bayesian latent variable model, where client-specific parameters are assumed to be realization from a global distribution at the master level, which is in turn estimated to account for data bias and variability across clients. We show that our framework can be effectively optimized through expectation maximization over latent master’s distribution and clients’ parameters. We tested our method on the analysis of multi-modal medical imaging data and clinical scores from distributed clinical datasets of patients affected by Alzheimer’s disease. We demonstrate that our method is robust when data is distributed either in iid and non-iid manners: it allows to quantify the variability of data, views and centers, while guaranteeing high-quality data reconstruction as compared to the state-of-the-art autoencoding models and federated learning schemes.

International audience; Initiatives such as the UK Biobank provide joint cardiac and brain imaging information for thousands of individuals, representing a unique opportunity to study the relationship between heart and brain. Most of research on large multimodal databases has been focusing on studying the associations among the available measurements by means of univariate and multivariate association models. However, these approaches do not provide insights about the underlying mechanisms and are often hampered by the lack of prior knowledge on the physiological relationships between measurements. For instance, important indices of the cardiovascular function, such as cardiac contractility, cannot be measured in-vivo. While these non-observable parameters can be estimated by means of biophysical models, their personalisation is generally an ill-posed problem, often lacking critical data and only applied to small datasets. Therefore, to jointly study brain and heart, we propose an approach in which the parameter personalisation of a lumped cardiovascular model is constrained by the statistical relationships observed between model parameters and brain-volumetric indices extracted from imaging, i.e. ventricles or white matter hyperintensities volumes, and clinical information such as age or body surface area. We explored the plausibility of the learnt relationships by inferring the model parameters conditioned on the absence of part of the target clinical features, applying this framework in a cohort of more than 3 000 subjects and in a pathological subgroup of 59 subjects diagnosed with atrial fibrillation. Our results demonstrate the impact of such external features in the cardiovascular model personalisation by learning more informative parameter-space constraints. Moreover, physiologically plausible mechanisms are captured through these personalised models as well as significant differences associated to specific clinical conditions.

In this study we investigate a novel quantitative instrument for the development of intervention strategies for disease modifying drugs in Alzheimer’s disease. Our framework is based on the modeling of the spatio-temporal dynamics governing the joint evolution of imaging and clinical biomarkers along the history of the disease, and allows the simulation of the effect of intervention time and drug dosage on the biomarkers’ progression. When applied to multi-modal imaging and clinical data from the Alzheimer’s Disease Neuroimaging Initiative our method enables to generate hypothetical scenarios of amyloid lowering interventions. The results quantify the crucial role of intervention time, and provide a theoretical justification for testing amyloid modifying drugs in the pre-clinical stage. Our experimental simulations are compatible with the outcomes observed in past clinical trials, and suggest that anti-amyloid treatments should be administered at least 7 years earlier than what is currently being done in order to obtain statistically powered improvement of clinical endpoints.

International audience; Albeit spectral-domain OCT (SDOCT) is now in clinical use for glaucoma management, published clinical trials relied on time-domain OCT (TDOCT) which is characterized by low signal-to-noise ratio, leading to low statistical power. For this reason, such trials require large numbers of patients observed over long intervals and become more costly. We propose a probabilistic ensemble model and a cycle-consistent perceptual loss for improving the statistical power of trials utilizing TDOCT. TDOCT are converted to synthesized SDOCT and segmented via Bayesian fusion of an ensemble of GANs. The final retinal nerve fibre layer segmentation is obtained automatically on an averaged synthesized image using label fusion. We benchmark different networks using i) GAN, ii) Wasserstein GAN (WGAN) (iii) GAN + perceptual loss and iv) WGAN + perceptual loss. For training and validation, an independent dataset is used, while testing is performed on the UK Glaucoma Treatment Study (UKGTS), i.e. a TDOCT-based trial. We quantify the statistical power of the measurements obtained with our method, as compared with those derived from the original TDOCT. The results provide new insights into the UKGTS, showing a significantly better separation between treatment arms, while improving the statistical power of TDOCT on par with visual field measurements.

Hydrogenases are among the fastest H2 evolving catalysts known to date and have been extensively studied under in vitro conditions. Here, we report the first mechanistic investigation of an [FeFe]-hydrogenase under in vivo conditions. Functional [FeFe]-hydrogenase from the green alga Chlamydomonas reinhardtii is generated in genetically modified Escherichia coli cells, by addition of a synthetic cofactor to the growth medium. The assembly and reactivity of the resulting semi-synthetic enzyme was monitored using whole-cell electron paramagnetic resonance as well as Fourier-transform infrared spectroscopy. Through a combination of gas treatments, pH titrations and isotope editing, we were able to corroborate the physiological relevance of a number of proposed catalytic intermediates, including reactive iron-hydride species. We demonstrate the formation of the so-called hydride state in vivo. Moreover, two previously uncharacterized redox species are reported herein, illustrating the complex metal hydride chemistry of [FeFe]-hydrogenase.

While data in healthcare is produced in quantities never imagined before, the feasibility of clinical studies is often hindered by the problem of data access and transfer, especially regarding privacy concerns. Federated learning allows privacy-preserving data analyses using decentralized optimization approaches keeping data securely decentralized. There are currently initiatives providing federated learning frameworks , which are however tailored to specific hardware and modeling approaches, and do not provide natively a deployable production-ready environment. To tackle this issue, herein we propose an open-source fed-erated learning frontend framework with application in healthcare. Our framework is based on a general architecture accommodating for different models and optimization methods. We present software components for clients and central node, and we illustrate the workflow for deploying learning models. We finally provide a real-world application to the federated analysis of multi-centric brain imaging data.

International audience; Objective: The Multidomain Alzheimer Preventive Trial (MAPT) was designed to assess the effect of omega-3 supplementation and a multidomain intervention (physical activity, cognitive training and nutritional advice) on cognitive decline of people with subjective memory complaint. In term of cognitive testing, no significant effect on cognitive decline was found over the 3-year follow-up. Yet, in the context of dementia-related conditions, brain morphological changes can be used to foretell the cognitive evolution. In this paper, we evaluate the effect of the interventions on the evolution of the brain morphology using the MR images acquired during MAPT. Methods: Subjects in the MAPT cohort with two MRI acquisitions, at baseline and at 36 months, were included , resulting in a subset of 376 subjects distributed in the 4 intervention groups: multidomain intervention plus omega-3, multidomain intervention plus placebo, omega-3 alone, and placebo alone. The morphological changes were assessed from volume measurements of regions of interest and a voxel-wise deformation-based approach. The primary outcome is the longitudinal deformation observed between the baseline image and the 3-year follow-up. Results: The multi-domain intervention is associated with a significant effect on the 3-year morphological evolution. The effect is similar within the two groups undergoing the intervention regardless of the omega-3 or placebo treatment. The voxel-wise deformation-based approach shows that the differences are mainly located in the left peri-ventricular area next to the temporoparietal junction (TPJ). These morphological changes correspond to a slower morphological evolution and are correlated with a better performance in cognitive assessments. These results could not be observed using the volumetric morphometry approach. No effect of omega-3 was observed.Discussion: In this study, we found that the multidomain intervention has a significant effect on morphological changes that are usually associated with the cognitive decline. This result suggests that effects at the level of cognitive decline may be visible in the long term, and that the cognitive scores may not be powerful enough to detect changes after 3 years. We argue that the use of neuroimaging could help define whether early intervention strategies are effective to delay cognitive decline and dementia.

In computational anatomy, the statistics from the object space (images, surfaces, etc.) are often lifted to the group of deformation acting on their embedding space. Statistics on transformation groups have been considered in the previous chapters by providing the Lie group with a left- or right-invariant metric, which may (or may not) be consistent with the group action on our original objects. In this chapter we take the point of view of statistics on abstract transformations, independently of their action. In this case it is reasonable to ask that our statistical methods respect the geometric structure of the transformation group. For instance, we would like to have a mean that is stable by the group operations (left and right compositions, inversion). Such a property is ensured for Frechet means in Lie groups endowed with a biinvariant Riemannian metric, like compact Lie groups (e.g. rotations). Unfortunately, biinvariant Riemannian metrics do not exist for most noncompact and noncommutative Lie groups, including rigid-body transformations in any dimension greater than one. Thus there is a need for the development of a more general non-Riemannian statistical framework for general Lie groups. In this chapter we partially extend the theory of geometric statistics developed in the previous chapters to affine connection spaces. More particularly, we consider connected Lie groups endowed with the canonical Cartan–Schouten connection (a generally nonmetric connection). We show that this connection provides group geodesics that are completely consistent with the composition and inversion. With such a nonmetric structure, the mean cannot be defined by minimizing the variance as in Riemannian manifolds. However, the characterization of the mean as an exponential barycenter gives us an implicit definition of the mean using a general barycentric equation. Thanks to the properties of the canonical Cartan connection, this mean is naturally biinvariant. In finite dimension this provides strong theoretical bases for the use of one-parameter subgroups. The generalization to infinite dimensions is at the basis of the SVF-framework. From the practical point of view, we show that it leads to efficient and plausible models of atrophy of the brain in Alzheimer's disease.

Hydrogenases are among the fastest H2 evolving catalysts known to date and have been extensively studied under in vitro conditions. Here, we report the first mechanistic investigation of an [FeFe]-hydrogenase under in vivo conditions. Functional [FeFe]-hydrogenase from the green alga Chlamydomonas reinhardtii is generated in genetically modified Escherichia coli cells, by addition of a synthetic cofactor to the growth medium. The assembly and reactivity of the resulting semi-synthetic enzyme was monitored using whole-cell electron paramagnetic resonance as well as Fourier-transform infrared spectroscopy. Through a combination of gas treatments, pH titrations and isotope editing, we were able to corroborate the physiological relevance of a number of proposed catalytic intermediates, including reactive iron-hydride species. We demonstrate the formation of the so-called hydride state in vivo. Moreover, two previously uncharacterized redox species are reported herein, illustrating the complex metal hydride chemistry of [FeFe]-hydrogenase.

International audience; We introduce a probabilistic generative model for disentangling spatio-temporal disease trajectories from series of high-dimensional brain images. The model is based on spatio-temporal matrix factorization, where inference on the sources is constrained by anatomically plausible statistical priors. To model realistic trajectories, the temporal sources are defined as monotonic and time-reparametrized Gaussian Processes. To account for the non-stationarity of brain images, we model the spatial sources as sparse codes convolved at multiple scales. The method was tested on synthetic data favourably comparing with standard blind source separation approaches. The application on large-scale imaging data from a clinical study allows to disentangle differential temporal progression patterns mapping brain regions key to neurodegeneration, while revealing a disease-specific time scale associated to the clinical diagnosis.

Apathy and anhedonia are common behavioral and psychological symptoms across brain disorders. Despite mainly common in presence of pathological conditions, apathy and anhedonia can also be present in healthy people. A comprehensive description of lack of interest and anhedonia in the general population is missing. The UK Biobank (UKB) includes 502,655 community-dwelling participants aged 37—73y, recruited between 2006 and 2010 in the UK. Extensive data were collected, including genetics, brain imaging, blood samples, family and medical history. No participant in the UKB received an apathy diagnosis (ICD10). However, self-report questions on mental health may be used as apathy proxies.

The rabbit model has become increasingly popular in neurodevelopmental studies as it is best suited to bridge the gap in translational research between small and large animals. In the context of preclinical studies, high-resolution magnetic resonance imaging (MRI) is often the best modality to investigate structural and functional variability of the brain, both in vivo and ex vivo. In most of the MRI-based studies, an important requirement to analyze the acquisitions is an accurate parcellation of the considered anatomical structures. Manual segmentation is time-consuming and typically poorly reproducible, while state-of-the-art automated segmentation algorithms rely on available atlases. In this work we introduce the first digital neonatal rabbit brain atlas consisting of 12 multi-modal acquisitions, parcellated into 89 areas according to a hierarchical taxonomy. Delineations were performed iteratively, alternating between segmentation propagation, label fusion and manual refinements, with the aim of controlling the quality while minimizing the bias introduced by the chosen sequence. Reliability and accuracy were assessed with cross-validation and intra- and inter-operator test-retests. Multi-atlas, versioned controlled segmentations repository and supplementary materials download links are available from the software repository documentation at https://github.com/gift-surg/SPOT-A-NeonatalRabbit., Highlights • In magnetic resonance image analysis accurate automatic segmentations tools are critically important. • Multi-atlas based methods are valuable tools to obtain an automatic MRI segmentation. • The rabbit model has become increasingly popular in neuro developmental studies. • Currently there is no multi-atlas of the newborn rabbit available in literature. • In this manuscript we provide the first high resolution MRI multi atlas for the newborn rabbit brain.

Recovering the 3D structure of a stack of histological sections (3D histology reconstruction) requires a linearly aligned reference volume in order to minimize z-shift and “banana effect”. Reconstruction can then be achieved by computing 2D registrations between each section and its corresponding resampled slice in the volume. However, these registrations are often inaccurate due to their inter-modality nature and to the strongly nonlinear deformations introduced by histological processing. Here we introduce a probabilistic model of spatial deformations to efficiently refine these registrations, without the need to revisit the imaging data. Our method takes as input a set of nonlinear registrations between pairs of 2D images (within or across modalities), and uses Bayesian inference to estimate the most likely spanning tree of latent transformations that generated the measured deformations. Results on synthetic and real data show that our algorithm can effectively 3D reconstruct the histology while being robust to z-shift and banana effect. An implementation of the approach, which is compatible with a wide array of existing registration methods, is available at JEI’s website: www.jeiglesias.com.

At this moment, databanks worldwide contain brain images of previously unimaginable numbers. Combined with developments in data science, these massive data provide the potential to better understand the genetic underpinnings of brain diseases. However, different datasets, which are stored at different institutions, cannot always be shared directly due to privacy and legal concerns, thus limiting the full exploitation of big data in the study of brain disorders. Here we propose a federated learning framework for securely accessing and meta-analyzing any biomedical data without sharing individual information. We illustrate our framework by investigating brain structural relationships across diseases and clinical cohorts. The framework is first tested on synthetic data and then applied to multi-centric, multi-database studies including ADNI, PPMI, MIRIAD and UK Biobank, showing the potential of the approach for further applications in distributed analysis of multi-centric cohorts, Comment: Federated learning, distributed databases, PCA, SVD, meta-analysis, brain disease

Aujourd’hui, il y a un besoin croissant d’harmonisation et d’innovation dans les evaluations cognitives et comportementales. Les outils actuels sont parfois trop invasifs, couteux ou demande des temps de realisations trop important dans le cadre d’une simple consultation. De ce faite de nouvelles methodes ecologiquement valides et sensibles pourraient etre utiles pour ameliorer l’accessibilite en tant que depistage de premiere ligne dans la population souffrant de troubles neuropsychiatriques. Les technologies de l’information et de la communication (TIC) sont des solutions non invasives et qui ont montrees une utilite pour identifier les sujets aux premiers stades cliniques des maladies neurodegeneratives [1] , [2] . Les recherches actuelles s’orientent sur l’interet des ICT a un stade pre clinique, comme marqueur d’evolution, au cours des essais therapeutiques et dans les troubles psychiatriques [3] , [4] . Cette session a pour objectif d’illustrer les scenario d’utilisation d’outils numeriques novateurs, qui pourraient etre utilises pour le depistage a grande echelle et pour le suivi des patients dans les essais cliniques. Lea Domain, interne de psychiatrie au centre hospitalier Guillaume-Regnier de Rennes presentera les resultats preliminaires de l’etude DEFLUENCE. Cette etude a pour objectif de determiner si les alterations qualitatives aux tests de fluences verbales mesurees de facon automatisee peuvent constituer un bio marqueur pronostic de l’evolution de la depression. Alexandra Konig, neuropsychologue et chercheuse au laboratoire CoBTeK presentera l’application Δelta sur tablette qui permet aux cliniciens de faire passer et d’analyser automatiquement des tests neuropsychologiques classiques meme a distance a l’aide de l’intelligence artificielle (IA), de l’analyse automatisee de l’expression faciale et de la voix. Un exemple d’une analyse psycholinguistique informatisee d’une entrevue clinique sera egalement presentee. Clement Abi Nader, doctorant dans l’equipe Epione INRIA presentera les travaux portant sur la modelisation de l’evolution de la maladie d’Alzheimer a partir de donnees cliniques longitudinales acquises. Cette approche consiste a developper des algorithmes integrant des donnees heterogenes (imagerie, donnees biologiques, capteurs, donnees cliniques).

Le but de cette presentation est d’illustrer notre contribution a l’elaboration de methodes d’apprentissage statistique robustes et interpretables, afin d’analyser des donnees issues d’essais cliniques pour l’etude de pathologies telles que l’apathie ou la maladie d’Alzheimer. En particulier, nous nous interesserons a des approches detectant automatiquement la relation jointe entre biomarqueurs de grande dimension tels que les images medicales, l’information genetique ou des tests psychologiques. Nous presenterons egalement nos travaux portant sur la modelisation de l’evolution de la maladie d’Alzheimer a partir de donnees cliniques longitudinales acquises sur une courte duree. Nous verrons comment nos methodes permettent d’identifier differents types de progressions de biomarqueurs impliques dans le processus pathologique. Ces methodes peuvent etre appliquees sur des donnees de grande dimension, notamment des images medicales acquises selon differentes modalites. Enfin, nous montrerons comment ces methodes peuvent etre utilisees comme reference statistique afin d’evaluer et de predire le stade de la maladie a partir de donnees cliniques non etudiees par nos algorithmes.