Publications

Les applications potentielles de l’intelligence artificielle, ces algorithmes visant à améliorer l’efficacité et la sécurité de diverses décisions cliniques, sont nombreuses en odontologie. Alors que les premiers logiciels commerciaux commencent à être proposés, la plupart des algorithmes n’ont pas été solidement validés pour une utilisation clinique. Cet article décrit les enjeux entourant le développement de ces nouveaux outils, afin d’aider les praticiens à garder un regard éclairé et critique sur cette nouvelle approche.

Background: Recent studies have stated the relevance of having new parameters to quantify the position and orientation of the scapula with patients standing upright. Although biplanar radiography can provide 3D reconstructions of the scapula and the spine, it is not yet possible to acquire these images with patients in the same position. Methods: Two pairs of images were acquired, one for the 3D reconstruction of the spine and ribcage and one for the 3D reconstruction of the scapula. Following 3D reconstructions, scapular alignment was performed in two stages, a coarse alignment based on manual annotations of landmarks on the clavicle and pelvis, and an adjusted alignment. Clinical parameters were computed: protraction, internal rotation, tilt and upward rotation. Reproducibility was assessed on an in vivo dataset of upright biplanar radiographs. Accuracy was assessed using supine cadaveric CT-scans and digitally reconstructed radiographs. Findings: The mean error was less than 2° for all clinical parameters, and the 95% confidence interval for reproducibility ranged from 2.5° to 5.3°. Interpretation: The confidence intervals were lower than the variability measured between participants for the clinical parameters assessed, which indicates that this method has the potential to detect different patterns in pathological populations.

Integrating remote Internet of Things (IoT) laboratories into project-based learning (PBL) in higher education institutions (HEIs) while exploiting the approach of technology-enhanced learning (TEL) is a challenging yet pivotal endeavor. Our proposed approach enables students to interact with an IoT-equipped lab locally and remotely, thereby bridging theoretical knowledge with practical application, creating a more immersive, adaptable, and effective learning experience. This study underscores the significance of combining hardware, software, and coding skills in PBL, emphasizing how IoTRemoteLab (the remote lab we developed) supports a customized educational experience that promotes innovation and safety. Moreover, we explore the potential of IoTRemoteLab as a TEL, facilitating and supporting the understanding and definition of the requirements of remote learning. Furthermore, we demonstrate how we incorporate generative artificial intelligence into IoTRemoteLab’s settings, enabling personalized recommendations for students leveraging the lab locally or remotely. Our approach serves as a model for educators and researchers aiming to equip students with essential skills for the digital age while addressing broader issues related to access, engagement, and sustainability in HEIs. The practical findings following an in-class experiment reinforce the value of IoTRemoteLab and its features in preparing students for future technological demands and fostering a more inclusive, safe, and effective educational environment.

Innovation models are key to fostering technology-focused entrepreneurship in higher education institutions (HEIs). These models create dynamic environments that encourage collaboration, creativity, and problem-solving skills among students and faculty. HEIs face several challenges in fostering entrepreneurship, including allocating sufficient financial and human resources, integrating entrepreneurship education across disciplines, and managing intellectual property. Overcoming these challenges requires HEIs to cultivate an entrepreneurial culture and establish strong partnerships with industry stakeholders. To achieve these goals, HEIs must adopt successful innovation models proven to work. This article presents an international case study highlighting such models and the factors contributing to their success. This study explores the implementation and impact of innovation models, specifically IDEATION and DEETECHTIVE, within HEIs to foster technology-focused entrepreneurship. By implementing numerous actions focusing on online education integration and the Quintuple Helix Innovation Model, these models support shifting engineering students’ mindsets toward entrepreneurship. This research highlights the importance of academia–industry collaboration, international partnerships, and the integration of entrepreneurship education in technology-focused disciplines. This study presents two models. The first, IDEATION, focuses on open innovation and sharing economy aspects. This model underwent rigorous testing and refinement, evolving into the second model, DEETECHTIVE, which is more comprehensive and deep tech-focused. These models have been validated as effective frameworks for fostering entrepreneurship and innovation within HEIs. This study’s findings underscore the potential of these models to enhance innovation capacity, foster an entrepreneurial culture, and create ecosystems rich in creativity and advancement. Practical implications include the establishment of open innovation-oriented structures and mechanisms, the development of specialized curriculum components, and the creation of enhanced collaboration platforms.

Automated detection of lathe checks in wood veneers presents significant challenges due to their variability and the natural properties of wood. This study explores the use of two convolutional neural networks (U-Net architecture) to enhance the precision and efficiency of lathe checks detection in poplar veneers. The approach involves sequential application of two U-Nets: the first for detecting lathe checks through semantic segmentation, and the second for refining these predictions by connecting fragmented lathe checks. Post-processing techniques are applied to denoise the mappings and extract precise lathe check characteristics. The first U-Net demonstrated strong performance in predicting lathe check presence, with precision and recall scores of 0.822 and 0.835, respectively. The second U-Net refined predictions by linking disjointed segments, improving the overall lathe checks mapping process. Comparative analysis with manual methods revealed comparable or superior performance of the automated approach, especially for shallow lathe checks. The results highlight the potential of the proposed method for efficient and reliable lathe check detection in wood veneers.

Study design: Multicentric retrospective study of prospectively collected data. Objective: Based on normative data from a cohort of asymptomatic volunteers, this study sought to determine the rate of abnormal values of proximal junctional angles (PJA) in adult spinal deformity (ASD) surgery patients, and compare it with PJK rate. Summary of Background Data: Proximal junctional kyphosis (PJK) definition does not take the vertebral level into account. Methods: This study included 721 healthy volunteers and 824 ASD surgery patients with 2-year postoperative follow-up. Normative values for each disc and vertebral body between T1 and T12 were analyzed, then normative values for PJA at each thoracic level were defined in the volunteer cohort as the mean±2 standard deviations. PJA abnormal values at the upper instrumented vertebra (UIV) were compared with Glattes’ and Lovecchio’s definitions for PJK in the ASD population at two years. Results: Mean age was 37.7±16.3 in the volunteer cohort, with 50.5% of females. Mean thoracic kyphosis (TK) was -50.9±10.8°. Corridors of normality included PJA greater than 20° between T3 and T12. Mean age was 60.5±14.0 years in the ASD cohort, with 77.2% of females. Mean baseline TK was -37.4±19.9°, with a significant increase after surgery (-15.6±15.3°, P<0.001). There was 46.2% of PJK according to Glattes’ versus 8.7% according to Lovecchio’s and 22.9% of kyphotic PJA compared to normative values (P<0.001). Conclusion: This study provides normative values for segmental and regional alignment of thoracic spine, used to describe abnormal values of PJA for each level. Using level-adjusted PJA values allows a more precise assessment of abnormal proximal angles and question the definition for PJK. Level of evidence: II

Automated detection of lathe checks in wood veneers presents significant challenges due to their variability and the natural properties of wood. This study explores the use of two convolutional neural networks (U-Net architecture) to enhance the precision and efficiency of lathe checks detection in poplar veneers. The approach involves sequential application of two U-Nets: the first for detecting lathe checks through semantic segmentation, and the second for refining these predictions by connecting fragmented lathe checks. Post-processing techniques are applied to denoise the mappings and extract precise lathe check characteristics. The first U-Net demonstrated strong performance in predicting lathe check presence, with precision and recall scores of 0.822 and 0.835, respectively. The second U-Net refined predictions by linking disjointed segments, improving the overall lathe checks mapping process. Comparative analysis with manual methods revealed comparable or superior performance of the automated approach, especially for shallow lathe checks. The results highlight the potential of the proposed method for efficient and reliable lathe check detection in wood veneers.

Wood density measurement plays a crucial role in assessing wood quality and predicting its mechanical performance. This dataset was collected to compare the accuracy and reliability of two non-destructive techniques, X-rays and terahertz waves, for measuring wood density. While X-rays have been commonly used in the industry due to their effectiveness, they pose health risks due to ionizing radiation. Terahertz waves, on the other hand, are non-ionizing and offer high spatial resolution. This article presents a database of wood samples measurements obtained using both techniques, on the same 110 samples with a fine location of the measuring points, on a wide range of wood species (tropical and temperate ones) and densities, from 111 kg·m−3 to 1086 kg·m−3. The database includes X-ray and terahertz scans, sample dimensions, moisture content, and color photographs.

Background: Maintaining balance while moving isvital forday-to-day activities. Akeychallenge inthe comprehension ofhuman movement istodetermine howmuscles contribute tobalance-movement coordination. Motor transitions, defined asmovements executed between twosteady balance states, areparticularly interesting phases tostudy balance-movement coordination because alarge, discrete change inwhole-body momentum may disturb balance. During voluntarily-initiated motor transitions, anticipatory muscle patterns provide the biomechanical conditions thatarefavourable toboth maintaining balance andexecuting themovement. Research question: What arethemechanical consequences ofanticipatory muscle activations forbalancemovement coordination during voluntarily-initiated motor transitions? Methods: Wereview thebiomechanical contributions oftheanticipatory muscle activations identified inthe literature during four types ofvoluntarily-initiated motor transitions, through theprism ofthree balance mechanisms (‘moving thecentre ofpressure (CoP)’,‘counter-rotating segments’,and‘applying new external force(s)’).Inparticular, weinvestigate howanticipatory muscle activations modulate whole-body centre ofmass acceleration. Results: Weshow thatthemechanical consequences ofanticipatory muscle activations have been extensively described, butmainly using the‘moving theCoP ’ mechanism. Unlike their roleduring steady balance states, both ‘moving theCoP ’ and‘applying new external force(s)’ mechanisms create arequired mechanical instability during theanticipatory phase ofmotor transitions. The‘counter-rotating ’ mechanism may actasastabiliser during motor transitions, butadditional research isneeded toclarify thisassumption. Significance: Thisreview establishes thatmuscle activation processes have different mechanical consequences for balance-movement coordination during theanticipatory phases ofmotor transitions, compared tosteady balance states. Because themechanical instability thatiscreated canleadtofalls, abetter understanding ofthemechanisms underlying motor transitions isneeded toenable thedesign ofmore effective fallprevention programs and/or devices forpopulation with balance deficits.

This study outlines the performances of finite-length journal bearing with textured liner, shear-thinning Rabinowitsch lubricants, and considers the elastic deformation of the bearing liner. The yielded nonlinear Rabinowitsch-Reynolds PDE system has been solved using the finite difference method combined with Elrod’s algorithm for the case of cylindrical textures. The static performance of hydrodynamic lubrication, in this case, involves different parameters such as depth of texture, eccentricity ratio, elastic deformation factor, rheological index, etc. Results showed that, at a fixed eccentricity ratio, texturing of the bearing’s converging area enhances significantly the load-carrying capacity and reduces the friction coefficient compared to the smooth bearing surface (up to 22% in load capacity and 17% in friction coefficient). On the other hand, using a lubricant with more pronounced shear-thinning behavior and/or elastic liner reduces the journal bearing performances (up to 25% in load capacity and 30% in pressure). By applying the optimization technique (particle swarm optimization), an optimal arrangement of textures that can compensate for these losses in performance even at high eccentricity has been found. This optimal texture with dimensionless dimple’s depth set to 1 (ry=40 µm) improves the load capacity with 10% for a lubricant with a rheological index α= 0.6 and elastic bearing liner with deformation parameter C0=0.05.