The clinical application of this project is focused on bone regeneration afterlong time to repair fractures in elderly population or in the case of complex fractures (bone lengthening, regeneration of large bone volume ). Current therapeutic approaches include bone grafts (self- and allograft transplants) or implants composed of various biomaterials, but none of them are currently acceptable. Generally, an autologous bone graft contains all the elements required for bone repair: an osteoconductive structure, osteoinductive and angiogenic growth factors, and cells with osteogenic potential. The use of bone grafts has major disadvantages. If the success rate is high, complications or consolidation defects are observed, especially in the case of large bone regeneration. In addition, replacement of the site damaged by the host bone is often incomplete, and hardening of the autologous bone often results in morbidity of the donor site. Moreover, alternative biomaterials may be given adequate osteoconductivity/inductivity and increased osteogenic potential through cellularization. However, this approach has not shown its fully efficacy in bone regeneration yet. The main causes of this failure may lie in the fact that the initial state of the different tissues is not taken into account.. This also leads to vascular problems. The osteogenic function of the periosteum has been recognized since the 18th century. It occurs in any situations leading to bone surface detachment , whether of traumatic, infectious, tumorous or surgical origin. This fibro-cellular tissue provides dual functions: interwined mechanical and biological. While its outer fibrous layer plays a mechanical role, its deep cellular layer is able to initiate a massive production of skeletal tissue, as diverse as bone, cartilage, tendons and ligaments and even muscle tissue. This "cambial" layer contains mesenchymal progenitor cells which have preserved a potential for multidirectional differentiation. In addition, the periosteum is responsible for the emission of bioactive molecules (growth factors) that allow these various differentiation pathways to be modulated. This project aims at regenerating the periosteum (flexible elastic membrane) and then regenerate bone (viscoelastic hard tissue) based on previousresearch results of some members of the future network The goal of the creation of this network is to develop a consortium which allows s to understand and evaluate the cellular contribution of periosteal cells to bone regeneration, particularly concerning the speed and the spread of the implant vascularization. The aim of the approach will be to make the most effective use of the periosteal function in order to optimize the bone regeneration process. This network will be able to apply for the H2020 SC1-BHC-07-2019 call on the regenerative medicine, and also on the 2020 Biomaterials call. Based on clinical observations of the beneficial role of periosteal tissue, we propose to study the implantation and integration of the periosteal biomimetic composite biomaterials, to accelerate these processes and optimize the quality of regenerated tissue. The approach will include biological, chemical, physical, and tissue and mechanical engineering sciences.

The goal of the Virtual Epileptic Patient Brain (VEP) project is to provide clinicians with novel tools that will increase surgery success and minimize invasiveness for the treatment of focal drug resistant epilepsy (DRE). The most efficient treatment of DRE is resective surgery, which is underused and has barely improved its outcome of seizure-free patients for the past 50 years. The VEP solution towards a better management of DRE is a bioinformatic approach using personalized brain models, derived from each patient’s own anatomy. A personalized virtual brain is constructed from the patient’s non-invasive neuroimaging data. Computer simulations of the virtual brain generate predictions on brain imaging and personalized surgery procedures, which will be tested in France (Timone Hospital in Marseille) and the USA (Cleveland Clinic). Virtual brain technology also proposes novel highly innovative therapeutic solutions, such as spatially distributed multi-focal laser microsurgery, which will be tested in the clinic. The Virtual Brain (TVB) Technology is at the core of the project (http://www.thevirtualbrain.org). TVB is constrained by human subject-specific anatomical information derived from anatomical MRI and diffusion tensor imaging (DTI). A large repertoire of mathematical tools enables mimicking depth EEG recordings, seizure genesis and propagation. In a preliminary study from 15 DRE patients, we computed a score estimating the difference between the Epileptogenic Zone (EZ) identified by TVB, but rejected by clinicians during pre-surgical evaluation. A large difference of the identified EZ was correlated with poor seizure prognosis according to the Engel classification, suggesting that had clinicians performed a pre-surgical patient analysis using TVB technology, they may have changed their surgical strategy.

Many complex autonomous systems (e.g., electrical distribution networks) repeatedly select actions with the aim of achieving a given objective. Reinforcement learning (RL) offers a powerful framework for acquiring adaptive behaviour in this setting, associating a scalar reward with each action and learning from experience which action to select to maximise long-term reward. Although RL has produced impressive results recently (e.g., achieving human-level play in Atari games and beating the human world champion in the board game Go), most existing solutions only work under strong assumptions: the environment model is stationary, the objective is fixed, and trials end once the objective is met. The aim of this project is to advance the state of the art of fundamental research in lifelong RL by developing several novel RL algorithms that relax the above assumptions. The new algorithms should be robust to environmental changes, both in terms of the observations that the system can make and the actions that the system can perform. Moreover, the algorithms should be able to operate over long periods of time while achieving different objectives. The proposed algorithms will address three key problems related to lifelong RL: planning, exploration, and task decomposition. Planning is the problem of computing an action selection strategy given a (possibly partial) model of the task at hand. Exploration is the problem of selecting actions with the aim of mapping out the environment rather than achieving a particular objective. Task decomposition is the problem of defining different objectives and assigning a separate action selection strategy to each. The algorithms will be evaluated in two realistic scenarios: active network management for electrical distribution networks, and microgrid management. A test protocol will be developed to evaluate each individual algorithm, as well as their combinations.

The proposal investigates the differences in physiology, anatomy and organization of the cortex in mouse, non-human primate (NHP) and human. This work requires tight collaborations between physiologists, anatomists and theoreticians. Our capacity to successfully integrate across these approaches is strongly supported by the numerous joint publications linking these disciplines in leading international journals by the PI’s of the consortium. Anatomy: Tract-tracing will be used to build macaque and mouse inter-areal cortical connectomes. This work will generate large data bases on inter-areal connection weights and quantitative measures of laminar distributions as well as atlases of mouse and macaque. The structural basis of hierarchy and local-global integration will be investigated with viral tracers that will be used to map the long distance and local input to the parent neurons of feedforward and feedback connections in visual cortex of mouse and macaque. Physiology: Hierarchical processing in the human, NHP and mouse brains will be compared using electrophysiological and imaging approaches and together with tract tracing, will inform embedded large-scale dynamic models of inter-areal processing in the cortex. Differences in the inter-areal matrix density lead to widely different core structures across the three species, which will be explored by weighted network structural analysis, thereby revealing the core-periphery organization, which we hypothesize could be relevant to the Global Neuronal Workspace theory of consciousness (Figure 4). We will manipulate consciousness with anesthetics and stimulation techniques in macaque and mouse thereby by exploring Global Neuronal Workspace function via auditory signatures of consciousness in a predictive coding paradigm. Modeling: Conditional Granger causality analysis on multi-variate time series recordings will help identify functional subnetwork motifs, in order to explore the links between structural and dynamical features in the networks across the three species. Whole-brain computational modeling will address the functional role of the underlying anatomy by studying in silico theoretical measures of integration and segregation allowing topological hierarchical analyses of effective connectivity as opposed to anatomical or functional connectivity. Altogether, the project aims to provide quantitative metrics of differences in brain organization related to changes in brain size and order, and will demonstrably underpin the relevance of investigations in the mouse and macaque for understanding the human brain.