Université de technologie de Troyes open archive
Not a member yet
10722 research outputs found
Sort by
Identification of latent biomarkers in brain imaging of Parkinson’s disease using explainable artificial intelligence
No full textInternational audienceParkinson’s disease (PD) is the second most common neurodegenerative disorder worldwide, characterized by the progressive degeneration of dopaminergic neurons. Early diagnosis remains challenging due to the lack of specific clinical tests. Although imaging techniques such as SPECT and MRI are commonly used to support diagnosis, their analysis is most often limited to striatal regions. In this study, we introduce a deep learning-based method for PD detection, while also exploring the role of non-striatal brain regions, which are often overlooked. Using DaTSCAN volumes, we trained a three-dimensional convolutional neural network (3D CNN) to distinguish control subjects from PD patients at different stages (1 to 3). To interpret the predictions, we applied the Grad-CAM technique to localize the regions influencing the model’s decision. Our network achieved remarkable accuracy (>97%) across all stages of the disease, and the Grad-CAM maps revealed a significant involvement of cortical and subcortical regions beyond the striatum. These findings suggest the existence of early or complementary biomarkers, highlighting the value of explainable artificial intelligence in brain imaging to refine PD diagnosis and broaden our understanding of how the disease develops and affects the brai
The grey zone two-echelon vehicle routing problem with customer-to-parcel locations and synchronization for last-mile deliveries
No full textInternational audienc
"Ils nous ont donné tant d'outils". Mouvement dans les Ecologies d'artefact de la sécurité civile francaise.
No full textIn this paper, we investigate the evolutions of digital tools and collaborative practices in a French regional civil security service (SDIS) and its partners (fire-fighters, gendarmes, emergency health services, and local government). Through the lens of Ecologies of Artifacts (EoA), we analyze such evolutions as ”movements” within complex socio-technical collaborative environment. Drawing on empirical material collected through the observation of an inter-service training, three on-site visits, and 40 interviews with 33 different professionals, we characterise a dense, redundant, interconnected, and misunderstood EoA. We identify several movements in its evolutive EoA: personal-to-organizational, outside-introduction, top-down, bottom-up, horizontal, suppression, realignment, replacement, as well as their associated ripple effects, and accelerating factors. From our findings and previous research, we discuss the adaptability of the EoA, the movements’ emerging features at the different levels of EoA, and generalizability to other contexts. We advocate for a movement-sensitive design approach, which means supporting tools integration within existing EoA rather than replacing them outright
L’environnement numérique des soins : déplacer les frontières, renforcer les juridictions
No full textInternational audienc
Diseño conceptual de un reactor heterogéneo fotocatalítico con ozonización usando TiO2
No full textInternational audienceSe hace la propuesta de utilizar TiO2 como catalizador para la degradación de contaminantes emergentes recalcitrantes en agua mediante el diseño conceptual de un reactor heterogéneo fotocatalítico. Una investigación sobre los diferentes tipos de reactores y sobre la fotocatálisis arrojó datos sobre la viabilidad de la propuesta. Se tomaron consideraciones respecto al tipo de luz, la concentración del catalizador, el uso de complementos para la reacción química, la agitación y el control de temperatura. Se culminó con un diseño conceptual de un reactor heterogéneo fotocatalítico
Multifunctional Silicon Surfaces with Enhanced Mechanical Resistance
No full textInternational audienceStructuring silicon (Si) surfaces at the micro‐ and nanoscale is an effective strategy to achieve broadband antireflective behavior, essential for photonic and energy applications. Herein, scalable and low‐cost fabrication techniques including self‐assembly, metal dewetting, and nonchlorine dry etching are used to produce micro‐ and nanostructured Si surfaces with enhanced antireflective properties. Total reflectivity is reduced to below 4% for microstructures and down to 2% for nanostructures, resulting in a pronounced black silicon effect. These optical improvements are validated by finite‐difference time‐domain simulations, which closely match the experimental results. Despite the improved optical performance, surface structuring leads to increased fragility. To address this, a 200 nm conformal Al 2 O 3 coating is applied via atomic layer deposition. Nanoscratch testing reveals significant mechanical reinforcement: critical load values tripled for microstructured surfaces and quadrupled for nanostructures. Notably, embedding the nanostructures beneath the coating preserves the low reflectivity while further enhancing durability. This multifunctional design approach enables the fabrication of surfaces that are both optically efficient and mechanically robust. Moreover, the fabrication techniques are compatible with large‐area processing, offering a promising pathway for industrial‐scale applications in advanced optical and energy systems
Ethics and Digital Transition
No full textInternational audienceThis book presents a study carried out by the IEEE-SMC French chapter on ethics and digital transformation in industry and society. Based on a survey of researchers in ICT, artificial intelligence (AI), as well as on presentation seminars, this study examines the various aspects that should be considered when assessing ethical principles in approaches to digital transition, particularly with regard to intelligent systems.Considering this, Ethics and Digital Transition presents the main technologies and uses of intelligent systems. Bringing together specialists from various fields, it explores the different dimensions of ethics that should be considered in the development of these systems, from the engineering sciences to law, sociology and philosophy. It also looks at the future challenges of ethics in the digital transition
Reliable Gait Parameters from a Lower-Back IMU for Fall Risk Assessment
No full textInternational audienceAccurate and reliable gait assessment is essential for fall prevention, particularly among older adults. Inertial measurement units (1MUs) represent a portable and affordable alternative to traditional motion capture systems. This study investigates the reproducibility of gait parameters obtained from a single lower-back IMU during walking at three self-selected speeds: slow, normal, and fast. After signal preprocessing, we extracted features related to gait intensity (e.g., speed, cadence), variability (e.g., stride duration variation), and energy expenditure (via estimated VO2). Reproducibility was evaluated using the Intraclass Correlation Coefficient (ICC2). Results showed that spatiotemporal parameters such as speed, cadence, and stride length were highly reliable, especially at normal and fast speeds (ICC > 0.75). However, variability metrics and medio-lateral power were less consistent at slow speeds, likely due to reduced motion amplitude. These findings support the use of IMUs for gait monitoring in both clinical and research settings. Additionally, by aligning the analysis with the Timed Up and Go (TUG) test, this approach enhances a standard geriatric evaluation with richer biomechanical insights to better inform fall-risk assessment
Adaptive Fall Detection Using WiFi CSI for Unseen Environments and New Individuals
No full textInternational audienceQuick detection of falls is crucial for elderly individuals to prevent severe injuries, reduce hospitalization risks, and ensure timely medical intervention. This work introduces a scalable, non-invasive fall detection system that leverages WiFi Channel State Information (CSI), eliminating the need for wearable devices or intrusive cameras, thus providing a privacy-preserving solution. Our lightweight 2D-CNN architecture consists of only 202,705 parameters, making it computationally efficient while achieving up to 100% accuracy in controlled settings, where training and testing data are collected from the same indoor environments with different sets of activities. It also demonstrates strong generalization, detecting new falls in unseen environments with 87% accuracy and falls from entirely new individuals with 79% accuracy. To enhance adaptability to real-world conditions, various data augmentation techniques such as shadowing and time shifting are applied, leading to a 13% improvement in generalization accuracy compared to baseline models up to 97% larger. Designed for real-time operation, our system is highly effective in practical fall scenarios. It also contributes to human activity recognition (HAR), demonstrating broader applicability beyond fall detection. These findings establish our model as a robust, efficient, and practical solution for fall detection and beyond
Grasping With Occlusion-Aware Ally Method in Complex Scenes
No full textInternational audienceRobotic arm target grasping by vision support is a commonly used method in grasping tasks and is usually used for multi-target complex scenes. Where vision support is generally used to identify the targets and to get their positions, categories and sizes. Most robotic arm grasping tasks using target recognition methods as visual inspection ignore the relationship between target objects such as the occlusion problem between objects. This limits the targets to be grasped and makes the crawling task inefficient. We propose Grasping with Occlusion-Aware aLly (GOAL) method based on binocular stereo-vision. Firstly, occlusion relationships in the view are directly inferred and targets are segmented as well as localized. Subsequently, multi-target grasping pose estimation is performed to obtain effective grasping positions. Ultimately, validation is conducted on a high-resolution dataset using the EPSON robotic arm. Note to Practitioners—This research significantly advances the field by addressing occlusion challenges in robotic grasping, offering effective methods, a valuable dataset, and practical insights. The proposed Grasping with Occlusion-Aware aLly (GOAL) method was validated on a high-resolution dataset using the EPSON robotic arm, showcasing its applicability and efficiency in real-world scenarios. This work provides valuable contributions to practitioners in the field of robotic manipulation and grasping tasks