Consorzio Pisa Ricerche

Efforts to map disparities in environmental exposure burdens across populations in varying urban settings are extremely limited. Here, we analyzed disparities in exposure to nitrogen dioxide air pollution and urban greenspaces in Dublin, Copenhagen, and Amsterdam, using hyperlocal environmental metrics derived from large-scale digital datasets. In Dublin, racial/ethnic minority groups were exposed to higher levels of nitrogen dioxide and lower levels of greenspace. Furthermore, immigrant populations were exposed to more pollution and less greenspace than natives in Dublin and Copenhagen, whereas the opposite trend was observed in Amsterdam. However, when considering population wealth, inhabitants of low-income areas were exposed to greater levels of greenspace and lower levels of nitrogen dioxide in all cities. These findings demonstrate that population environmental exposures are inconsistent across cities. Our results emphasize the need for city-specific studies of exposure burdens utilizing hyperlocal data to inform the design of sustainable, healthy, and equitable cities of the future. Immigrants in Dublin and Copenhagen experience more air pollution and less greenery than natives, while in Amsterdam, the reverse is true, according to hyperlocal analysis of nitrogen dioxide and greenspace exposure.

Ultrasonography has been proposed as a complementary tool for the evaluation of udder health, yet limited information exists on its application for measuring the glandular cistern area during mastitis. This study aimed to investigate modifications of the glandular cistern area in clinical (CM) and subclinical mastitis (SCM) compared with contralateral healthy quarters, and to assess the reliability of manual and automated ultrasonographic measurements. A longitudinal study was conducted on 42 Italian Holstein cows, comprising 26 SCM and 16 CM quarters, each paired with contralateral healthy controls. Ultrasound examinations were performed at diagnosis (T0), 24 h (T1), and 5 days (T5). Cisternal areas were measured in transversal and longitudinal sections using both manual and ImageJ-guided methods. Intra- and inter-operator reliability was assessed using Intraclass Correlation Coefficients (ICCs). Statistical analyses included two-way mixed ANOVA with post-hoc Bonferroni correction. Mastitic quarters tended to show smaller cisternal areas compared with contralateral healthy quarters, with significant differences observed between contralateral healthy and CM quarters (p < 0.05), but not between contralateral healthy and SCM or between SCM and CM quarters. Temporal trends differed significantly among groups (interaction effect, p < 0.05). Both manual and automated measurements demonstrated excellent intra- and inter-operator reliability, with ICCs consistently > 0.95 across pathology groups and time points. Ultrasonography of the glandular cistern is a feasible, reproducible, and reliable method under field conditions. Clinical mastitis is associated with a measurable reduction in cisternal area, while changes in subclinical mastitis appear less evident. The high reproducibility of measurements, including automated analysis, supports the use of this technique may contribute to improve the use of ultrasound also for the udder heath.

The agricultural sector remains one of the most hazardous working environments, with viticulture posing particularly high risks due to repetitive manual tasks, pesticide exposure, and machinery operation. This study explores the potential of vision-based Artificial Intelligence (AI) systems to enhance occupational health and safety by evaluating their coherence with human expert assessments. A dataset of 203 annotated images, collected from 50 vineyards in Northern Italy, was analyzed across three domains: manual work activities, workplace environments, and agricultural machinery. Each image was independently assessed by safety professionals and an AI pipeline integrating convolutional neural networks, regulatory contextualization, and risk matrix evaluation. Agreement between AI and experts was quantified using weighted Cohen’s Kappa, achieving values of 0.94–0.96, with overall classification error rates below 14%. Errors were primarily false negatives in machinery images, reflecting visual complexity and operational variability. Statistical analyses, including McNemar and Wilcoxon signed-rank tests, revealed no significant differences between AI and expert classifications. These findings suggest that AI can provide reliable, standardized risk detection while highlighting limitations such as reduced sensitivity in complex scenarios and the need for explainable models. Overall, integrating AI with complementary sensors and regulatory frameworks offers a credible path toward proactive, transparent, and preventive safety management in viticulture and potentially other high-risk agricultural sectors. Furthermore, vision-based AI systems inherently act as optical sensors capable of capturing and interpreting occupational risk conditions. Their integration with complementary sensor technologies—such as inertial, environmental, and proximity sensors—can enhance the precision and contextual awareness of automated safety assessments in viticulture.

Disclosure: G. Scozia: None. D. Menghini: None. D. Fintini: None. S. Cianfarani: None. V. Russo: None. M. Mainardi: None. N. Gianni: None. G. Furini: None. G. Scabia: None. M. Maffei: None. M. Manco: None.Background: In pediatric obesity, altered energy regulation and impaired inhibitory control over food intake are linked to dysfunction in the dorsolateral prefrontal cortex (DLPFC). Lifestyle interventions can potentially restore these brain-metabolic pathways. Leptin, brain-derived neurotrophic factor (BDNF), and eotaxin (CCL11), a chemokine involved in immune signaling, may modulate neural responses to metabolic challenges. Objective: To assess the impact of an 8-week intensive health behavioral treatment (IHBT) on DLPFC activation during oral glucose tolerance testing (OGTT), and to explore the relationship with insulin sensitivity and circulating levels of leptin, BDNF and eotaxin in children with obesity. Methods: Twenty-seven children with obesity (mean age 11 years) underwent functional near-infrared spectroscopy (fNIRS) during OGTT at baseline (T0) and post-IHBT (T1). The IHBT included healthy diet, structured physical activity, and cognitive training. Changes in cerebral blood flow (total Hb, tHb) and oxygenated hemoglobin (O₂Hb) were measured in the DLPFC. Insulin sensitivity was assessed via the Matsuda Index and HOMA-IR; serum eotaxin, leptin, and BDNF levels were also measured by commercial assays. Results: After IHBT, there was a significant increase in bilateral DLPFC activation during OGTT:Left DLPFC: ΔO₂Hb ↑, p < 0.01Right DLPFC: ΔO₂Hb ↑, p < 0.05ΔtHb bilateral ↑, indicating enhanced cortical perfusionImprovement in DLPFC activity correlated with metabolic indices:Positive correlation with Matsuda Index (r = 0.54, p < 0.01)Negative correlation with HOMA-IR (r = -0.47, p < 0.05)Left DLPFC activation at T1 showed sex-related differences, with greater increase in males (p < 0.05)Baseline eotaxin levels were positively associated with DLPFC activation changes at T1 (r = 0.41, p < 0.05), while leptin and BDNF showed no significant associations. Conclusion: An 8-week lifestyle intervention significantly enhanced DLPFC activation during glucose challenge, suggesting improved neural regulation of food-related behavior in children with obesity. The association with eotaxin implicates a potential neuroimmune mechanism whereby inflammatory signaling may influence prefrontal responsiveness to metabolic inputs. Eotaxin may serve as a biomarker of neuroplastic adaptation to behavioral interventions targeting obesity.Presentation: Monday, July 14, 2025

Abstract Radiotherapy (RT) is a cornerstone treatment for nearly 50% of cancer patients, but its curative potential and safe dosing are constrained by cumulative toxicity to surrounding healthy tissues. Delivering RT at ultra-high dose rates (FLASH-RT) represents a transformative strategy, as it appears to maintain tumor control, while sparing normal tissue. Melanoma is among the most radioresistant tumors, and skin and muscle are invariably exposed during RT, also in case of deep seated tumors. Here, we compared electron FLASH-RT and conventional RT (CONV-RT) in melanoma-bearing and naïve mice assessing tumor control, tissue integrity, and systemic homeostasis over the medium to long term. Both modalities achieved comparable tumor suppression. However, CONV-RT induced persistent skin damage, dermal fibrosis, muscle dysfunction and systemic inflammatory-metabolic alterations, while FLASH-RT largely spared normal tissue and systemic balance. Bulk RNA sequencing revealed striking differences: FLASH induced minimal transcriptional disruption in skin and muscle, whereas CONV-RT triggered thousands of differentially expressed genes, including massive activation of fibrosis, inflammation, cell death-related pathways in skin, and broad dysregulation of genes linked to muscle function, remodeling and the unfolded protein response. Histological and ultrastructural analyses corroborated the findings, showing reduced immune infiltration in the skin and preserved tissue architecture both in skin and muscle following FLASH. In conclusion our study not only confirms the protective nature of FLASH but also provides novel mechanistic insights into the cascade linking local injury to systemic dysfunction under CONV-RT, reinforcing the translational potential of FLASH to expand the therapeutic window of radiotherapy. One Sentence Summary Comparative analysis of FLASH and conventional radiotherapy in a murine model of melanoma and naïve mice - Therapeutical efficacy, local and systemic effects.

Hydroethanolic extracts of coriander seeds (CE), garlic bulbs (GE), and their combination (CGE), were chemically profiled using HS-SPME/GC-MS and LC-Q-TOF/MS-MS to assess volatile composition and to study phenolic molecules, respectively. Biological assays were conducted through invitro and invivo experiments to measure the EC50 of the antioxidant assays and the MIC/MBC/MFC values of the antibacterial/antifungal activities. Antioxidant combination Index (CI) and fractional inhibitory concentration index (FICI) values were further recorded. The acute oral toxicity, analgesic, and gastroprotective activities were evaluated invivo on Wistar rats and Swiss albino mice. Caffeoyl quinic acid, feruloyl-quinic acid, and caffeic acid derivatives (quercetin, apigenin, and luteolin-O-glycoside) together with monoterpene linalool, α-pinene, and γ-terpinene were found to be highly present in CE. Organosulfur compounds (allicin, S-allyl-l-cysteine, allin, γ-Glutamyl-S-allylcysteine, and allyl sulfide) were predominant in GE. All the profiled compounds were co-present in CGE. Invivo assays responded in a dose-dependent showing better activity mainly at 200 mg kg-1. Chromatographic analysis profiled various phenolic acids, flavonols and derivatives, monoterpene, and organosulfur compounds in the assessed extracts and their combinations. Bioassays' responses were found to be dose-dependent with better scores recorded with CGE. Thus, a synergetic effect was significantly highlighted.

Distributed multiple-input multiple-output (MIMO) radar systems deployed on mobile platforms are gaining strategic importance across a wide range of civil and military applications. These systems offer high sensitivity, robustness to shadowing, and improved target localization capabilities. In recent years, photonic technologies have demonstrated the capability to coherently distribute the radar signals via optical fiber or free space optical links, thereby enabling the development of coherent distributed radar configurations. This coherence enables effective data fusion and centralized processing from the spatially distributed radars. Thereby, coherent MIMO processing can be achieved across widely separated radars, as in the case of radar mounted on a swarm of drones. The design of a distributed MIMO radar system is inherently challenging due to the large number of design parameters involved, especially for mobile sensor platforms. Stepping from our recent experimental demonstrations of coherent photonics-assisted distributed MIMO radar, this paper analyzes the impact of the main system design parameters, i.e., radar signal bandwidth, number of radar peripherals, and baseline geometry and dimension, on the key performance indicators, i.e., spatial resolution and sidelobe levels of the MIMO ambiguity function. It also proposes an innovative approach to improve system performance. This approach aims to minimize hardware complexity by exploiting a combination of meta-heuristic algorithms for architecture optimization and signal post-processing. The former step allows optimization of spatial resolution, while the latter reduces the detrimental effects of sidelobes, which are typical in distributed MIMO radars. The proposed analysis focuses on a specific application of a photonics-assisted radar constellation mounted on a swarm of drones operating in the X and Ka bands. In particular, the proposed optimization method is applied to 3 × 3 MIMO system, demonstrating the capability to achieve 2 cm resolution in both range and cross-range. The detection capability of the enhanced system is validated when the target is located 50 m away from baseline, using cell-averaging constant false alarm rate detection criterion. A target detection probability of 0.996 is achieved at false alarm probability of 10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−6</sup>. These results confirm the effectiveness of the proposed approach in enabling high-performance, low-resource drone-based MIMO radar surveillance systems.

This paper explores the need for new systems to detect and monitor cyberattacks in Connected Vehicles (CVs). Sensor health in CVs is vital, as prediction errors and communication issues can weaken the sensor network. Intrusion Detection Systems (IDS) for CVs must be continuously updated to meet changing needs and be robust against adversarial attacks. We developed a new Label Flipping system against Deep learning-based IDS (LFD-IDS) to help cloud operators understand unusual vehicle sensor data. LFD-IDS specifically targets detecting and explaining sensor data manipulation from poisoning attacks. We proposed two label-flipping attacks based on Bootstrapping and Bagging and a defensive strategy using a multi-layer deep neural network. Our LFD-IDS achieves at least 90% accuracy in identifying cyberattacks.