Search

Metallogels built in a bottom-up approach by metal coordination and supramolecular interactions have important potential for the elaboration of smart materials. In this context, we present here the formation of supramolecular coordination polymers driven by the complexation of cobalt(II) or zinc(II) ions with polyoxometalate-based hybrids displaying two terpyridine ligands in a linear arrangement. Thanks to the electrostatic interactions between the polyoxometalate cores and metal nodes, the polymer chains self-assemble into fibers that physically cross-link to form gels above a critical concentration. Using spectroscopy, microscopy, X-ray scattering, and rheometry, complemented by molecular dynamics simulations, we investigated the supramolecular organization of the chains in the fibers and the resulting processes leading to gelation. Compared to previously reported systems, these gels have improved rheological features and appealing properties, such as birefringence, luminescence, and spin crossover, paving the way for their use as building blocks for multifunctional smart materials.

This paper introduces BrainFuseNet, a novel lightweight seizure detection network based on the sensor fusion of electroencephalography (EEG) with photoplethysmography (PPG) and accelerometer (ACC) signals, tailored for low-channel count wearable systems. BrainFuseNet utilizes the Sensitivity-Specificity Weighted Cross-Entropy (SSWCE), an innovative loss function incorporating sensitivity and specificity, to address the challenge of heavily unbalanced datasets. The BrainFuseNet-SSWCE approach successfully detects $93.5\%$ seizure events on the CHB-MIT dataset ($76.34\%$ sample-based sensitivity), for EEG-based classification with only four channels. On the PEDESITE dataset, we demonstrate a sample-based sensitivity and false positive rate of $60.66\%$ and $1.18$ FP/h, respectively, when considering EEG data alone. Additionally, we demonstrate that integrating PPG signals increases the sensitivity to $61.22\%$ (successfully detecting $92\%$ seizure events) while decreasing the number of false positives to $1.0$ FP/h. Finally, when ACC data are also considered, the sensitivity increases to $64.28\%$ (successfully detecting $95\%$ seizure events) and the number of false positives drops to only $0.21$ FP/h for sample-based estimations, with less than one false alarm per day when considering event-based estimations. BrainFuseNet is resource-friendly and well-suited for implementation on low-power embedded platforms, and we evaluate its performance on GAP9, a state-of-the-art parallel ultra-low power (PULP) microcontroller for tiny Machine Learning applications on wearables. The implementation on GAP9 achieves an energy efficiency of $21.43$ GMAC/s/W, with an energy consumption per inference of only $0.11$ mJ at high performance ($412.54$ MMAC/s). The BrainFuseNet-SSWCE method demonstrates effective and accurate seizure detection on heavily imbalanced datasets while achieving state-of-the-art performance in the false positive rate and being well-suited for deployment on energy-constrained edge devices.

The long-term, continuous analysis of electroencephalography (EEG) signals on wearable devices to automatically detect seizures in epileptic patients is a high-potential application field for deep neural networks, and specifically for transformers, which are highly suited for end-to-end time series processing without handcrafted feature extraction. In this work, we propose a small-scale transformer detector, the EEGformer, compatible with unobtrusive acquisition setups that use only the temporal channels. EEGformer is the result of a hardware-oriented design exploration, aiming for efficient execution on tiny low-power micro-controller units (MCUs) and low latency and false alarm rate to increase patient and caregiver acceptance.Tests conducted on the CHB-MIT dataset show a 20% reduction of the onset detection latency with respect to the state-of-the-art model for temporal acquisition, with a competitive 73% seizure detection probability and 0.15 false-positive-per-hour (FP/h). Further investigations on a novel and challenging scalp EEG dataset result in the successful detection of 88% of the annotated seizure events, with 0.45 FP/h.We evaluate the deployment of the EEGformer on three commercial low-power computing platforms: the single-core Apollo4 MCU and the GAP8 and GAP9 parallel MCUs. The most efficient implementation (on GAP9) results in as low as 13.7 ms and 0.31 mJ per inference, demonstrating the feasibility of deploying the EEGformer on wearable seizure detection systems with reduced channel count and multi-day battery duration.

Context New Caledonia is a hotspot of biodiversity, and among threatened species of the territory many are palms. Palms are sensitive to environmental changes and even more so when these species are narrow endemics. Monitoring populations is therefore an essential measure to conserve these emblems of tropical forest. Aims In this study, the population structure for three narrow endemic palms has been evaluated: Burretiokentia koghiensis , Cyphophoenix nucele and Chambeyronia houailouensis . Methods For each species, individuals were counted in several sampling plots. To understand the observed seedling deficit, we also assessed the impact of rat predation on fruit survival on C. nucele . Key results The three endangered species all have different structures. The structure of B. koghiensis shows that the population dynamics are currently undisturbed. The structure of C. nucele shows a very low number of seedlings, which does not affect the regeneration of the species. In contrast, the structure of C. houailouensis does not show a class deficit but neither does it fit the expected pattern. Conclusions All three species are subject to common threats such as introduced rats, yet their population structure is different. Implications The findings allow us to prioritise actions needed to restore regeneration of these species. These results underscore the importance of acquiring species-level data before implementing conservation actions.

•The number of mechanical thrombectomies has increased a sevenfold since 2012 in France.•The mean number of interventional neuroradiologist per center performing mechanical thrombectomy increased from 2.9±0.9 to 3.7±1.85.•Majority of interventional neuroradiologist use combined (stent and distal aspiration) approach (54.5%), and the use of balloon guide catheter is not standard, with a third of them never using it.•The majority of centers performing mechanical thrombectomy would require at least one more interventional neuroradiologist to stabilize team and insure continuity of care.

Mineralized biological tissues rich in type I collagen (e.g., bone and dentin) exhibit complex anisotropic suprafibrillar organizations in which the organic and inorganic moieties are intimately coassembled over several length scales. Above a critical size, a defect in such tissue cannot be self-repaired. Biomimetic materials with a composition and microstructure similar to that of bone have been shown to favorably influence bone regeneration. This highlights the value of developing a similar formulation in an injectable form to enable minimally invasive techniques. Here, we report on the fabrication and application potential of an injectable collagen/CHA (carbonated hydroxyapatite) cell-free hydrogel. The organic part consists of spray-dried nondenatured and dense collagen microparticles, while the inorganic part consists of biomimetic apatite mineral. By mixing both powders at desired tissue-like ratios with an aqueous solvent in one step, spontaneous co-self-assembly occurs, leading to the formation of a mineralized matrix with suprafibrillar tissue-like features thanks to the induced liquid crystalline properties of collagen on one hand and apatite on the other hand. When injected into soft tissue, the mineralized collagen hydrogel free of chemical cross-linking agents exhibits suitable cohesion and is biocompatible. Preliminary in vitro tests in a tooth cavity model show its integration onto dentin with a biomimetic interface. Based on the results, this versatile injectable mineralized collagen hydrogel shows promising potential as a biomaterial for bone tissue repair and mineralized tissue-like ink for bioprinting applications.

Recently, alginates (ALG) characterized by high mannuronic content (M blocks) have been shown to undergo a reversible sol/gel transition during cooling in the presence of potassium salts. Cold gelling takes place at low temperatures, just below 0 °C for a KCl concentration of 0.3 mol/kg, but the aggregation process can be easily shifted to higher temperatures by increasing the salt concentration. In the present paper, we take advantage of this peculiar behavior to design a copolymer with schizophrenic gelling properties. For this purpose, side chains of poly(N-isopropylacrylamide) (PNIPAM), characterized by a Lower Critical Solution Temperature (LCST) in water, were grafted on the alginate backbone. Working in semidilute solutions, we show by coupling DSC and viscoelastic measurements that ALG-g-PNIPAM solutions are able to form gels either by cooling or heating depending on the ionic environment. As the aggregation process of ALG and PNIPAM depends mainly and respectively on the nature of the cations and anions, the choice of the salt is then critical to control the self-assembly behavior and the gel properties. Moreover, as the gelation process of alginates driven by the aggregation of mannuronic sequences is characterized by a large hysteresis of 20–30 °C between gelling and melting, both ALG and ALG-g-PNIPAM polymers offer a large versatility not only in terms of salt (nature and concentration) but also in preparation history as different states (sol or gel) can be obtained at room temperature.

In order to systematically investigate the effect of the strength of the supramolecular interactions on the debonding properties of associative polymers, a series of model systems have been characterized by probe-tack tests. These model materials, composed of linear and low dispersity poly(isobutylene) chains (Mn≈ 3 kg/mol) center-functionalized by a single bis-urea sticker, are able to self-assemble by four hydrogen bonds. Three types of stickers are used in the present study: a bis-urea with a methylene diphenyl (MDI) spacer, a bis-urea with a tolyl (TOL) spacer, and a bis-urea with a xylyl (XYL) spacer. In order to investigate the influence of stickers in depth, both the nanostructure of the materials and the linear rheology were investigated by small-angle X-ray scattering (SAXS) and oscillatory shear, respectively. For two types of stickers (TOLand XYL), the association of polymers via hydrogen bonds induces the formation of bundles of rodlike aggregates at room temperature and the behavior of a soft elastic material was observed. For bis-urea MDI, no structure is detected by SAXS and a Newtonian behavior is observed at room temperature. In probe-tack experiments, all these materials show a cohesive mode of failure, a signature of flowing materials as previously observed for tri-urea center-functionalized poly(butylacrylate) (PnBA3U). However, XYLcenter-functionalized polyisobutene shows much higher debonding energies than PnBA3U, revealing the importance of the strength of noncovalent bonds in the scission/recombination dynamics. On the basis of the analysis of the debonding images, this effect is discussed via the mechanical behavior at large deformation.

L’évaluation du positionnement des tunnels de ligamentoplastie après reconstruction du ligament croisé antérieur (LCA) est primordiale. Historiquement, cette évaluation était réalisée à l’aide de radiographies standard, et plus récemment en utilisant le scanner avec des reconstructions tridimensionnelles. L’IRM est considérée comme un outil fiable pour évaluer l’intégrité du transplant de LCA, de même que les complications postopératoires. À notre connaissance, aucune étude n’a évalué ses performances dans l’évaluation du positionnement des tunnels. Le but de cette étude était d’évaluer les performances de l’IRM 3D dans l’évaluation du positionnement des tunnels tibiaux et fémoraux. L’hypothèse était que l’IRM 3D avec reconstructions était aussi précise que le scanner 3D pour l’évaluation du positionnement des tunnels de ligamentoplastie du LCA.

A new approach for the elaboration of low molecular weight pressure-sensitive adhesives based on supramolecular chemistry is explored. The synthesis of model systems coupled with probe-tack tests and rheological experiments highlights the influence of the transient network formed by supramolecular bonds on the adhesion energy. The first step of our approach consists of synthesizing poly(butyl acrylate-co-glycidyl methacrylate) copolymers from a difunctional initiator able to self-associate by four hydrogen bonds between urea groups. Linear copolymers with a low dispersity (Mn= 10 kg/mol, Ip < 1.4) have been synthesized via atom transfer radical polymerization. Films of the copolymers were then partially cross-linked through reaction of the epoxy functions with a diamine. The systematic variation of the average ratio of glycidyl methacrylate and diamine per copolymer shed light on the respective role played by the supramolecular interactions (between bis-urea groups and with the side chains) and by the chain extension and branching induced by the diamine/epoxy reaction. In this strategy, the adhesive performance can be optimized by modifying the strength of “stickers” (via the structure of the supramolecular initiator, for instance) and the polymer network (e.g., via the length and level of branching of the copolymer chains) in order to approach commercial PSA-like properties (high debonding energy and clean removal).

Des études récentes ont montré que l’IRM était performante dans l’évaluation des différentes composantes de la pubalgie du sportif (PS). Cependant, et à notre connaissance, aucune étude n’a évalué la corrélation systématique entre une description précise des lésions tendineuses ou pariétales en IRM, les constatations chirurgicales et l’analyse histologique. De fait, nous avons réalisé une étude cas-témoin afin de déterminer si l’IRM pouvait évaluer précisément une tendinopathie du long adducteur (LA), les lésions pariétales, en comparaison à la chirurgie ainsi qu’à l’analyse histologique.

Model hybrid hydrogels reinforced by silica nanoparticles were designed by polymerizing and cross-linking the gels in situ. The polymer–particle interactions were tuned by using either poly(dimethylacrylamide) (PDMA), which adsorbs on silica, or poly(acrylamide) (PAAm), which does not. Besides, the dispersion state of silica nanoparticles was tuned from well-dispersed to aggregated by changing the pH from 9, which ensures repulsive interactions between nanoparticles and good dispersion state, to about 6, which affects the surface chemistry of silica and promotes aggregation. The dispersion states were characterized by small-angle X-ray scattering (SAXS). The mechanical behavior of hybrid gels with aggregated nanoparticles is markedly different from those where silica is well-dispersed within the matrix. PDMA-based hybrid gels display pronounced nonlinear behavior, somehow similar to those observed in filled elastomers. The nonlinearities are even more pronounced in gels with aggregated particles, with strong strain stiffening along with large dissipation. For those samples, reinforcement can be attributed to the combination of both reversible interactions between PDMA and silica nanoparticles, which provide strain stiffening and recovery, and the response of the silica network. Recovery processes observed in hybrid gels with dispersed particles are preserved when silica particles are aggregated, but the characteristic time needed to fully recover the mechanical response is extended from a few seconds to several hours. In PAAm-based hybrid gels with aggregated silica nanoparticles, no recovery processes are observed. This implies that the properties, namely, the very high linear tensile modulus and high dissipated energy, are driven by the rigid network formed by nanoparticle aggregation, which provides high dissipative capabilities, especially when compared to PAAm-based hybrid gels with dispersed silica, that remain soft and fragile. These gels exhibit a quite inhomogeneous structure, with permanent damage under elongation. The nonlinear dynamical behavior of hybrid gels was investigated by large amplitude oscillatory shear (LAOS) experiments. While unfilled gels show no nonlinearity up to very large strain amplitude, marked nonlinear effects combining a drop in modulus (similar to the Payne effect) and strain stiffening for increasing strain amplitude are observed in PDMA-based hybrid gels, certainly due to polymer adsorption onto nanoparticles. PAAm-based hybrid gels also show nonlinearity, with a drop in modulus for increasing strain but no strain stiffening, indicating that the presence of fillers alone can induce nonlinearity in the absence of strong, reversible polymer–particle interactions. PAAm-based hybrid gels with aggregated silica show very high stiffness and high dissipative properties at the expense of stretchability, though. Also, the structure seems to be permanently damaged under stress, revealing the importance of silica–polymer interactions for permanent mechanical reinforcement. Altogether, the analysis of the nonlinear behavior indicates the importance of combining dynamic adsorption of polymer chains on silica nanoparticles with mechanical reinforcement provided by the silica network.