Felice Sfravara

Background: Restoration of sensory feedback has become one of the major goals of modern bidirectional neuroprosthetic rehabilitation, enabling the transition from conventional prosthetic replacement toward bidirectional neuroprosthetic systems capable of restoring physiological sensorimotor communication. This narrative review aims to provide a comprehensive and clinically oriented overview of current neural and related sensorimotor interfaces for upper- and lower-limb prostheses, focusing on their mechanisms of sensory restoration, clinical performance, and translational potential. Methods: A targeted literature search was conducted in PubMed, Scopus, and Web of Science to identify clinical, translational, and landmark studies published between January 2010 and June 30, 2026. Sources were selected according to prespecified relevance criteria consistent with the narrative design of the review. Owing to the heterogeneity of the included technologies, study designs, populations, and outcomes, the evidence was synthesized narratively. Earlier seminal publications were included when relevant to the physiological basis of sensory restoration and the historical evolution of neural interfaces. The selected literature was analyzed thematically according to neural interface technology, sensory restoration strategies, functional outcomes, and clinical translation. Results: Evidence from small pilot studies and experimental evaluations suggests that direct peripheral and central neural interfaces, biological or regenerative interfaces, neuromuscular signal interfaces, non-invasive sensory-feedback systems, and integrated prosthetic platforms may contribute to the restoration or substitution of tactile, proprioceptive, thermal, and multimodal sensory information. Emerging technologies, including regenerative peripheral nerve interfaces, fully implantable wireless systems, biomimetic sensory encoding, adaptive closed-loop control, and artificial intelligence-assisted decoding, may support further clinical translation, although most remain investigational. However, the available evidence remains limited by small patient cohorts, heterogeneous methodologies, and a lack of standardized long-term outcome measures. Conclusions: Neural and related sensorimotor interfaces have shown the feasibility of providing sensory feedback in selected experimental and early clinical settings. Although no single technology currently fulfils all clinical requirements, the integration of complementary biological, neural, and computational approaches appears to represent the most promising pathway toward personalized neuroprosthetic systems with the potential to improve sensorimotor function, although long-term clinical and quality-of-life benefits remain uncertain.

Rheumatoid arthritis (RA) is a chronic systemic inflammatory disease that can cause severe joint damage and disability. The management of RA patients has evolved significantly over the past few decades due to improved detection of early disease progression and the initiation of targeted advanced treatments. However, several gaps remain, including disease risk assessment, early diagnosis, phenotypic identification, and risk of treatment failure. These gaps could potentially be addressed by the application of artificial intelligence (AI), defined as the ability of a machine to mimic intelligent human behavior, using machine learning and deep learning models. By analyzing various types of data, including clinical, laboratory, and imaging data, omics, demographics, and data from sensor applications or wearable technologies, it may be possible to improve the management of RA patients, as demonstrated by several studies. However, limitations related to interindividual variability, small datasets, study designs, and intrinsic model bias remain, making it difficult to generalize findings to larger cohorts of RA patients. The aim of this narrative review is to discuss the potential uses and limitations of AI-based algorithms in RA patients in light of the most recent scientific evidence.

Oscillating water column (OWC) devices are a promising technology for wave energy conversion, in which the efficiency of the power take-off (PTO) system plays a critical role. Savonius turbines are attractive candidates for OWC applications due to their mechanical simplicity, robustness, and inherent self-rectifying capability under bidirectional airflow; however, their aerodynamic performance is intrinsically limited. This work aims to enhance the performance of Savonius-based PTO systems through the aerodynamic optimization of external power augmenters. An experimental campaign was conducted on a scaled Savonius turbine operating under bidirectional, oscillatory airflow at two frequencies (0.1 and 1 Hz). A transient computational fluid dynamics (CFD) model based on a sliding mesh approach and dynamic fluid body interaction was developed and validated against experimental data. The numerical framework was then coupled with an optimization algorithm to refine the geometry of the power augmenters, with the objective of maximizing the turbine tip speed ratio. The optimized configuration showed a significant improvement in performance compared to both the bare turbine and the baseline power augmenter design. The maximum tip speed ratio increased by up to 52% at 1 Hz and 58.2% at 0.1 Hz, while the maximum power coefficient increased by up to 12.7%. Flow-field analysis revealed enhanced flow acceleration toward the driving scoop and reduced aerodynamic losses under both maximum and minimum thrust conditions. These results demonstrate the effectiveness of CFD-driven optimization for improving Savonius turbine PTO performance in OWC systems.

Point cloud registration is a central theme in computer vision, with alignment algorithms continuously improving for greater robustness. Commonly used methods evaluate Euclidean distances between point clouds and minimize an objective function, such as Root Mean Square Error (RMSE). However, these approaches are most effective when the point clouds are well-prealigned and issues such as differences in density, noise, holes, and limited overlap can compromise the results. Traditional methods, such as Iterative Closest Point (ICP), require choosing one point cloud as fixed, since Euclidean distances lack commutativity. When only one point cloud has issues, adjustments can be made, but in real scenarios, both point clouds may be affected, often necessitating preprocessing. The authors introduce a novel differential entropy-based metric, designed to serve as the objective function within an optimization framework for fine rigid pairwise 3D point cloud registration, denoted as Iterative Differential Entropy Minimization (IDEM). This metric does not depend on the choice of a fixed point cloud and, during transformations, reveals a clear minimum corresponding to the best alignment. Multiple case studies are conducted, and the results are compared with those obtained using RMSE, Chamfer distance, and Hausdorff distance. The proposed metric proves effective even with density differences, noise, holes, and partial overlap, where RMSE does not always yield optimal alignment.

The desire to improve the safety of navigation, especially in restricted and very busy areas like the straits, leads researchers to evaluate possible uses of Artificial Intelligence as an alternative to the traditional probabilistic methods. This is possible thanks to the large amount of available AIS data generated by ships in transit. In this work, a Machine Learning algorithm (Classification Decision Tree) was trained with eight features coming from AIS data of the Strait of Messina (Italy), with the aim of carrying out a two-class classification of the single AIS data to find anomalies in ship transits that could compromise navigation safety. Since anomalous events are relatively rare, compared to the large amount of information related to the normal navigation situations, the challenge of this work was to obtain an artificial dataset with the aim of simulating the possible anomalous navigation conditions for the Strait investigated, known the active risk mitigation means one. For this reason, the dataset containing abnormal events was obtained simulating different risk scenarios. A hyperparameters tuning with a Bayesian optimization approach and a 5-fold cross validation have been performed to improve the quality of the model and a large dataset has been tested. The accuracy of both validation and test phases is <99.5% and <95.9%, respectively. This can make it possible to identify anomalous navigation conditions in real time, in order to quickly classify possible conditions of risk. The method can be used as a Decision Support Tool by the authority in order to improve the capacity of the single operator to identify the possible risk situation inside the Strait of Messina.

Maritime collisions represent one of the main risks for operational safety at sea and can involve various types of contact, such as ship-to-ship impacts, collisions with fixed obstacles (e.g., bridge piers) or groundings. Such events can cause severe structural damage to vessels and ships, compromising their integrity and leading to significant environmental consequences in the event of spills or transported cargo loss. This study is focused on the evaluation of the consequences of a collision event on a reference structure, assessing the energy absorbed, the deformation resulting from an impact and the damage extents and feature. An explicit dynamic finite element model is employed to simulate the temporal evolution of the impact event, with a focus on quantifying energy absorption, deformation patterns, and damage extent. The results offer valuable insights into the mechanics of hull breach formation and energy dissipation, supporting the development of crashworthy structural solutions and effective containment strategies. This is crucial for preventing oil dispersion in the event of a collision and for ensuring safe navigation to the nearest port. This study is conducted within the framework of the EMPATHY project (Enhanced Marine Pollution Emergency Actions to Preserve Ocean Health and the Safety of Ship and Navigation), funded by the National Recovery and Resilience Plan (NRRP), which aims to enhance preparedness and resilience in maritime safety by advancing the modeling of collision scenarios and guiding emergency response planning.