Politecnio die Bari - Catalogo di prodotti della Ricerca
Not a member yet
36616 research outputs found
Sort by
Explainable AI for Brain Age Prediction: Design, Implementation, and Formative Evaluation of an Interactive Tool
No full textBrain age prediction is a valuable tool for distinguishing normal and pathological aging, offering quantitative insights into subtle brain structure changes. However, clinical adoption remains limited due to the complexity and low interpretability of Machine Learning (ML) algorithms. Human-Centered Artificial Intelligence (HCAI) and eXplainable AI (XAI) address these challenges by emphasizing user involvement, explainability, and facilitation of clinical applications. By combining advanced ML models with intuitive interfaces and transparent visualizations, HCAI fosters trust and usability in neurological practice. This paper presents the “Brain Age Predictor”, a web-based tool combining a deep learning model for brain age estimation with SHAP-based interpretability techniques and an interactive simulation panel. We conducted a preliminary study involving five neurology residents to assess its usability, interpretability, and potential value in clinical practice. The results show that the Brain Age Predictor effectively supports clinical exploration of brain aging, with participants praising its ease of use and clarity. Feedback highlighted areas for improvement, including richer visualizations, more detailed explanations, and tools for longitudinal patient monitoring
Multistain Multicompartment Automatic Segmentation in Renal Biopsies with Thrombotic Microangiopathies and other Vasculopathies
No full textAutomatic tissue segmentation is a necessary step for the bulk analysis of whole slide images (WSIs) from paraffin histology sections in kidney biopsies. However, existing models often fail to generalize across the main nephropathological staining methods and to capture the severe morphological distortions in arteries, arterioles, and glomeruli common in thrombotic microangiopathy (TMA) or other vasculopathies. Therefore, we developed an automatic multi-staining segmentation pipeline covering six key compartments: Artery, Arteriole, Glomerulus, Cortex, Medulla, and Capsule/Other. This framework enables downstream tasks such as counting and labeling at instance-, WSI- or biopsy-level. Biopsies (n = 158) from seven centers: Cologne, Turin, Milan, Weill-Cornell, Mainz, Maastricht, Budapest, were classified by expert nephropathologists into TMA (n = 87) or Mimickers (n = 71). Ground truth expert segmentation masks were provided for all compartments, and expert binary TMA classification labels for Glomerulus, Artery, Arteriole. The biopsies were divided into training (n = 79), validation (n = 26), and test (n = 53) subsets. We benchmarked six deep learning models for semantic segmentation (U-Net, FPN, DeepLabV3+, Mask2Former, SegFormer, SegNeXt) and five models for classification (ResNet-34, DenseNet-121, EfficientNet-v2-S, ConvNeXt-Small, Swin-v2-B). We obtained robust segmentation results across all compartments. On the test set, the best models achieved Dice coefficients of 0.903 (Cortex), 0.834 (Medulla), 0.816 (Capsule/Other), 0.922 (Glomerulus), 0.822 (Artery), and 0.553 (Arteriole). The best classification models achieved Accuracy of 0.724 and 0.841 for Glomerulus and Artery plus Arteriole compartments, respectively. Furthermore, we release NePathTK (NephroPathology Toolkit), a powerful open-source end-to-end pipeline integrated with QuPath, enabling accurate segmentation for decision support in nephropathology and large-scale analysis of kidney biopsies
A Heterogeneously Integrated Photonic-Electronic Architecture for FMCW LiDAR
No full textThe development of compact, robust Light Detection and Ranging (LiDAR) systems capable of long range (>200 m) and high accuracy (<10 cm) is essential for applications such as autonomous driving, robotics, and space exploration. Frequency-Modulated Continuous-Wave (FMCW) LiDAR offers inherent advantages for these applications, including simultaneous range and velocity measurement and high interference immunity, facilitating widespread adoption potential. However, realizing the target performance presents significant integration challenges. This paper proposes a heterogeneous photonic-electronic architecture designed to meet these demanding requirements. The architecture integrates an Indium Phosphide (InP) based highly linear laser source and amplifier with Silicon-on-Insulator (SOI) photonics, which includes passive circuitry and Optical Phased Array (OPA) beam steering. This Photonic Integrated Circuit (PIC) is intended for co-packaging with dedicated control and processing electronics, forming a complete, compact, solid-state FMCW LiDAR system projected to achieve 200 m range and 10 cm accuracy
Mapping the Integration of Urban Air Mobility into the Built Environment: A Bibliometric Analysis and a Scoping Review
Get PDFUrban Air Mobility (UAM) has the potential to revolutionize urban transportation, largely with the deployment of Unmanned Aerial Vehicles (UAVs), commonly known as drones. After an initial stage focused on technology requirements, research is now shifting toward investigating operational requirements, which are unavoidably affected by urban characteristics. This study aims to explore the implementation of UAM services within urban environments by mapping the current scientific landscape from a city-focused perspective. Following a systematic search procedure, a bibliometric analysis was conducted on studies published between 2010 and 2024, examining over 350 articles that address UAM and urban-related topics. Trends in publication volume and scientific impact were analysed, along with influential manuscripts, collaborations, and leading countries in the field. Through a keyword co-occurrence analysis, five main research themes were identified: air traffic management, risk assessment, environmental factors (wind and noise), and vertiport location. These themes were further explored through a scoping review to assess current research and emerging directions. The findings highlight that urban characteristics are not just operational constraints but also fundamental elements that shape UAM strategies, influencing UAV path planning, safety, environmental constraints, and infrastructure design. Future research directions include the development of urban digital twins, comprehensive urban spatial databases, and multi-objective optimization frameworks to support the effective implementation of UAM into cities
Risk management of Circular Economy: a framework based on Interpretive structural model
Get PDFOver last years Circular Economy has attracted increasing attention from scholars and businesses, due to its ability to decouple the economic and social growth from the usage of natural resources and the degradation of the environment. Several risks however affect the CE adoption and implementation. In this regard, it is of paramount importance to identify the relationships between circular economy practices, the risks associated with them and their mitigation strategies in order to identify the drivers on which to act for an adequate and efficient transition. Through a review of the literature of main risks and the analysis of case studies, the Interpretive Structural Modeling (ISM) methodology has been used to discover these relationships in the Italian furniture industry. This is the first study of its kind which provides evidence about the adoption and implementation of CE practices and related risks in the furniture sectors as well as mitigation strategies which can be used to manage such risks and overcome such barriers. The outcome of this study will help companies operating in furniture industry as well as policy makers to devise strategies to favor an adequate and efficient CE transition
Enhancing Photovoltaic Thermal System Efficiency Through Geometric Optimization and Intelligent Fuzzy Logic Based Management
Get PDFThe growing demand for sustainable and efficient energy systems has intensified research into improving the performance of photovoltaic thermal (PVT) technologies. This study aims to enhance the electrical and thermal efficiency of PVT systems by integrating material design modifications with intelligent control strategies. A novel methodology is proposed, which involves reducing the thickness of the tedlar layer from 5 mm to 1 mm to improve heat dissipation, alongside the implementation of an adaptive fuzzy logic control (AFLC) system to dynamically regulate the cooling flow rate based on environmental conditions. The results show a notable improvement in system performance, with electrical efficiency increasing from 11.81% to 13.81%, thermal efficiency rising from 27% to 81%, and electrical power output improving from 243.8 W to 271 W under controlled cooling conditions. These findings demonstrate that the proposed integrated approach, named GO-AFLC, effectively maximizes the energy extraction from PVT systems. The combination of passive (material optimization) and active (intelligent control) strategies offers a promising pathway for advancing the performance and sustainability of hybrid solar energy systems
Search for rare decays of the Z and Higgs bosons to a J/ψ or ψ(2S) meson and a photon in proton-proton collisions at s=13TeV
No full textA search is presented for rare decays of the [Figure presented] and Higgs bosons to a photon and a [Figure presented] or a [Figure presented] meson, with the charmonium state subsequentially decaying to a pair of muons. The data set corresponds to an integrated luminosity of 123fb−1 of proton-proton collisions at a center-of-mass energy of 13 TeV collected with the CMS detector at the LHC. No evidence for branching fractions of these rare decay channels larger than predicted in the standard model is observed. Upper limits at 95% confidence level are set: [Figure presented], [Figure presented], [Figure presented], and [Figure presented]. The ratio of the Higgs boson coupling modifiers [Figure presented] is constrained to be in the interval (−157,+199) at 95% confidence level. Assuming [Figure presented], this interval becomes (−166,+208)
Optimal lattice spring models derived from triangular and tetrahedral meshes
No full textLattice Spring Models (LSMs) are fast and stable simulation methods based on networks of discrete deformable elements, well-suited for large deformations or topological changes in real-time applications. Considering that the identification of spring network topologies and elastic constants lacks standardization, defining a robust design workflow for LSMs would promote their wider adoption, enabling efficient and green simulations. In this work, the essential requirements to obtain optimal spring distributions for physically consistent LSM were outlined. Intrinsic rigidity, homogeneity, isotropy, and boundary conformity were identified as fundamental conditions for achieving realistic behaviors of deformable lattices. To generate optimal spring networks, the possibilities offered by high-quality unstructured triangular and tetrahedral meshes were explored, since they satisfy all the outlined design requirements. Several topological metrics were defined and applied to clusters of spring network samples, with the aim of comparing different lattice topologies. By discussing the obtained results, a series of useful considerations were pointed out, and many best practices to generate optimal mesh-derived lattice configurations were clearly identified. The optimal LSMs were validated on several case studies, including comparisons to dual FEM models, analyses on anisotropic spring networks, and fracture simulations, thus giving highly encouraging results. The proposed approach based on high-quality unstructured triangular and tetrahedral meshes resulted in a valuable strategy to design physically consistent LSMs
Biological, Biochemical and Elemental Traits of Clavelina oblonga, an Invasive Tunicate in the Adriatic Sea
No full textClavelina oblonga is an invasive tropical tunicate recently introduced into the Adriatic Sea as a consequence of globalization and climate change. Mussel aquaculture sites provide an ideal environment for this colonial ascidian, where it has recently become the dominant fouling species. This study represents the first investigation of its biological and physical characteristics, as well as its proximal, fatty acid, macroelement, trace element, and toxic metal composition. The entire-tissue chemical composition of C. oblonga resulted in 95.44% moisture. Its composite structure revealed several strong peaks, attributed to O-H, C-H, C-N, and C=O stretching, along with cellulose components overlapping with proteins and carbohydrates. The major fatty acids were palmitic, stearic, and docosahexaenoic acid, followed by docosanoic, elaidic, linoleic, and myristic acid. The saturated fatty acids, polyunsaturated fatty acids, and monounsaturated fatty acids comprised 51.37, 26.96, and 15.41% of the total fatty acids, respectively. Among the analysed trace and macroelements, aluminium and sodium were predominant. C. oblonga exhibited different concentrations of toxic metals, such as arsenic and lead, compared to fouled mussels in the Istria region. It appears that the tunicate has adapted to the environmental conditions of the Adriatic, reaching its maximum spread and biomass in mid-autumn. There is a strong possibility that C. oblonga could colonize and establish itself permanently in the Adriatic. This would have a strong negative impact on shellfish farming, the structure of the ecosystem, plankton biomass, and the distribution of other marine species. However, it also represents a biomass resource with high potential of utilization in different industries
