Puglia Salute

Background: Biorevitalization is a minimally invasive cosmetic procedure that involves the targeted intradermal injection of bioactive substances designed to enrich the skin and stimulate its natural regenerative capabilities. The aim of this single-center, open-label, single-arm, uncontrolled, post-marketing study was to examine the effectiveness and safety of biostimulatory injections using a proprietary, sterile gel containing non-cross-linked hyaluronic acid (Foliage Hydrofil®) for facial skin rejuvenation. Methods: The trial was registered with ClinicalTrials.gov (NCT07010549). The intended target population comprised adults with mild-to-moderate facial photoaging and Fitzpatrick skin types III-IV, without active inflammatory dermatoses or recent aesthetic interventions. Twenty-four Caucasian subjects (21 women, three men; mean age: 44 ± 7.1 years) received three intradermal injections of the study product at three-week intervals. The primary endpoint was the change from baseline in skin capacitance, an objective measure of hydration, assessed via corneometry. Secondary endpoints included investigator and subject assessments of aesthetic improvement, while safety was evaluated through continuous adverse event monitoring. Results: The mean skin capacitance was 23.1 arbitrary units (a.u.) at baseline and increased to 33.2 a.u. at the end of the study, indicating a significant improvement in skin hydration (p < 0.05). Secondary outcomes showed moderate improvements in skin texture and tone, and high levels of participant satisfaction with the treatment. No adverse events were reported. Conclusions: This study demonstrated that the tested non-cross-linked hyaluronic acid gel safely improves skin hydration and patient satisfaction in facial biorevitalization. Although promising, these findings are constrained by the study's uncontrolled, small-scale design; therefore, larger randomized controlled trials are required to confirm these outcomes and evaluate their long-term durability.

In this narrative review we report the main relationships between trace metals and the hemostatic system since this aspect has seldom attracted the attention of the scientific community. A basic aspect to be considered is the importance of maintaining the fine control of all trace metals' levels since they have an important impact on the pathophysiology of the hemostatic system. It is worth noting that poor diet habits are responsible for most trace metal deficiencies, while pollution is responsible for dangerous exposure to them with a consequent negative impact on the general population. This appears of paramount importance in planning the implementation of food and nutrient support to ameliorate the hidden hunger and the quality of life of people especially in developing countries and limiting poisons both in the air and food. As it often happens, when damage to certain mechanisms takes a very long time to appear, no attention is paid to the importance of a systematic prevention to avoid late negative outcomes.

With the rapid growth of video content available, the ability to search for specific moments within videos using textual queries has become increasingly relevant. This is crucial in many scenarios, from surveillance cameras where it may be necessary to find specific events in extensive video streams to searching for exciting movie scenes. However, existing approaches for video Moment Retrieval and Highlight Detection often struggle to effectively align text and video features, limiting their performance. We argue that utilizing recent foundational video models designed for video-text alignment can overcome these limitations. We propose a novel architecture that utilizes such models to test this hypothesis. Combined with our novel Saliency-Guided Cross Attention mechanism and a hybrid DETR architecture, our approach provides significantly improved results. To further enhance our approach, we developed InterVid-MR — a large-scale, high-quality dataset specifically designed for a pretraining stage. Extensive experiments and comparisons with current state-of-the-art methods confirm the effectiveness of the approach, achieving 58.8 mAP on QVHighlights, 60.7 R@1 mIoU on Charades-STA, and 42.4 R@1 mIoU on TACoS. These results highlight the efficiency and scalability of the method for video-language tasks in both zero-shot and fine-tuning scenarios. The dataset and code will be publicly available. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup>

The multimodal management of patients with

A fire-safety retrofit was designed for two existing road tunnels in northern Italy. The design alternatives are compatible with the local geomorphology and the existing portal infrastructures. To assess the risk associated with the current condition and the design alternatives, a risk analysis procedure was implemented based on the FHWA report Prevention and Control of Highway Tunnel Fires. Because no Italian code or recommendation exists on this subject, it was necessary to tune the U.S. document to the Italian and local conditions. This risk analysis allowed both the designer and the client to assess the benefits of the design alternatives with respect to the current condition and to compare the design solutions, thus providing invaluable help in the decision-making process. The two proposed solutions increase the probability of no fatalities from zero (current state) to 64% and 68%, respectively.

PURPOSE: This observational study aimed to compare the effectiveness of mandibular advancement devices (MAD) and continuous positive airway pressure (CPAP) therapy in patients diagnosed with obstructive sleep apnea (OSA), with a specific focus on changes in heart rate parameters. METHODS: Fifty-two patients treated with MAD were compared with fifty-two patients treated with CPAP. All participants underwent sleep study prior to the initiation of therapy (T0), followed by a control sleep study after a (3-month) predefined treatment period (T1). Changes in cardiac parameters between baseline (T0) and post-treatment (T1) were analyzed both within each group (MAD and CPAP) and between groups. To enhance comparability between patients treated with MAD and those treated with CPAP, a propensity score matching approach was applied based on pre-treatment variables, in order to minimize selection bias and optimize the validity of the between-group comparison. RESULTS: For intragroup comparisons, both the MAD and C-PAP groups demonstrated statistically significant reductions in AHI and ODI following treatment (p < 0.001). Additionally, a modest but significant decrease in mean heart rate was noted in both groups (MAD: p = 0.033; CPAP: p = 0.002). For intergroup comparisons, both treatment modalities resulted in a significant reduction in AHI and ODI values. Regarding heart rate parameters, no statistically significant differences were observed between the two groups after treatment for mean HR (p = 0.350), minimum HR (p = 0.602), or maximum HR (p = 0.942). CONCLUSION: The use of MAD and CPAP leads to significant improvements in heart rate in patients with OSA. However, no significant differences were found between the two treatment modalities in terms of heart rate reduction.

Table understanding is a crucial task in document processing and is commonly encountered in practical applications. We introduce 2Columns1Row, the first open-source benchmark for the table question answering task in Russian. This benchmark evaluates the ability of models to reason about the relationships between rows and columns in tables, employing both textual and multimodal inputs. 2Columns1Row consists of six datasets, 28,680 tables, designed datasets that vary in the complexity of the text within the table contents and the consistency of the values in the cells. We evaluate the models using text-only and multimodal approaches and analyze their performance. Through extensive evaluation, we demonstrate the limitations of current multimodal models on this task and prove the feasibility of a dynamic text-based system utilizing our benchmark. Our results highlight significant opportunities for advancing table understanding and reasoning, providing a solid foundation for future research in this domain.

The rapid advancement of Large Language Models (LLMs) in software engineering has revealed critical limitations in existing benchmarks, particularly the widely used SWE-bench dataset. Recent studies have uncovered severe data contamination issues, e.g. SWE-bench~\cite{jimenez2023swe} reports 32.67\% of successful patches involve direct solution leakage and 31.08\% pass due to inadequate test cases. We introduce \textbf{SWE-MERA}, a dynamic, continuously updated benchmark designed to address these fundamental challenges through an automated collection of real-world GitHub issues and rigorous quality validation. Our approach implements a reliable pipeline that ensures quality while minimizing contamination risks, resulting in approximately 10,000 potential tasks with 300 samples currently available. Evaluation using the Aider coding agent demonstrates strong discriminative power in state-of-the-art models. We report performance across dozen recent LLMs evaluated on tasks collected between September 2024 and June 2025.

This paper introduces \textbf{DRAGOn}, method to design a RAG benchmark on a regularly updated corpus. It features recent reference datasets, a question generation framework, an automatic evaluation pipeline, and a public leaderboard. Specified reference datasets allow for uniform comparison of RAG systems, while newly generated dataset versions mitigate data leakage and ensure that all models are evaluated on unseen, comparable data. The pipeline for automatic question generation extracts the Knowledge Graph from the text corpus and produces multiple question-answer pairs utilizing modern LLM capabilities. A set of diverse LLM-as-Judge metrics is provided for a comprehensive model evaluation. We used Russian news outlets to form the datasets and demonstrate our methodology. We launch a public leaderboard to track the development of RAG systems and encourage community participation.

Multimodal large language models (MLLMs) are currently at the center of research attention, showing rapid progress in scale and capability, yet their intelligence, limitations, and risks remain insufficiently understood. To address these issues, particularly in the context of the Russian language, where no multimodal benchmarks currently exist, we introduce AnonymBench, an open multimodal evaluation framework for Russian-spoken architectures. The benchmark is instruction-based and encompasses default text, image, audio, and video modalities, comprising 18 newly constructed evaluation tasks for both general-purpose models and modality-specific architectures (image-to-text, video-to-text, and audio-to-text). Our contributions include: (i) a universal taxonomy of multimodal abilities; (ii) 18 datasets created entirely from scratch with attention to Russian cultural and linguistic specificity, unified prompts, and metrics; (iii) baseline results for both closed-source and open-source models; (iv) a methodology for preventing benchmark leakage, including watermarking and licenses for private sets. While our current focus is on Russian, the proposed benchmark provides a replicable methodology for constructing multimodal benchmarks in typologically diverse languages, particularly within the Slavic language family.

e15611 Background: In refractory metastatic colorectal cancer (mCRC) treated with trifluridine/tipiracil (T), treatment-related neutropenia has been associated with improved survival. Whether this benefit is linked to objective response patterns in real-world practice remains unclear. This sub-analysis of the ReTrITA study evaluated survival outcomes according to objective response categories in patients stratified by grade 3-4 neutropenia. Methods: ReTrITA is a multicenter Italian real-world observational study including patients with refractory mCRC treated with T. Patients were stratified by the occurrence of grade 3/4 neutropenia (T neut vs T no-neut). Objective response was classified as partial response (PR), stable disease (SD), or progressive disease (PD). Overall survival (OS) and progression-free survival (PFS) were estimated using the Kaplan–Meier method and compared across response categories using log-rank tests (χ²). Results: Overall, 843 patients were included in this substudy, of whom 807 (95.7%) were evaluable for objective response, including 265 in the T neut group and 542 in the T no-neut group. In the T neut group, PR, SD, and PD were observed in 5.9% (n = 16), 28.8% (n = 76), and 63.1% (n = 167) of patients, respectively. Overall survival differed significantly across response categories (log-rank χ² = 57.08, df = 2; p &lt; 0.0001), with median OS of 21.7 months for PR, 20.9 months for SD, and 8.5 months for PD. Progression-free survival showed a similar gradient (log-rank χ² = 82.83, df = 2; p &lt; 0.0001), with median PFS of 10.5, 8.5, and 3.7 months, respectively. In the T no-neut group, PR, SD, and PD were documented in 1.0% (n = 5), 20.3% (n = 110), and 73.4% (n = 398) of patients, respectively. Median OS was 13.0 months for PR, 15.5 months for SD, and 6.5 months for PD, with significant separation of survival curves (log-rank χ² = 75.66, df = 2; p &lt; 0.0001). Median PFS was 7.2, 7.1, and 3.0 months, respectively (log-rank χ² = 119.74, df = 2; p &lt; 0.0001). Across corresponding response categories, patients who developed grade 3–4 neutropenia consistently experienced longer OS and PFS compared with those without neutropenia. Conclusions: In this large real-world cohort of patients with refractory mCRC treated with T, objective response categories were strongly associated with OS and PFS. The occurrence of grade 3–4 neutropenia identified patients with consistently improved survival across response strata, supporting its role as a prognostic and pharmacodynamic marker of treatment benefit beyond radiological response.

136 Background: The ReTrITA study is a large, multicenter retrospective analysis evaluating real-world outcomes of regorafenib (R) and trifluridine/tipiracil (T) in refractory metastatic colorectal cancer (mCRC) across 17 Italian centers (2012–2023). This prespecified subgroup analysis focused on the impact of different initial regorafenib dosing strategies on survival outcomes. Methods: Patients treated with regorafenib (n = 713) were stratified into three groups according to starting dose: ReDOS stepwise escalation (n = 313), fixed reduced dose &lt;160 mg (n = 142), and standard 160 mg dose (n = 258). Primary endpoints were progression-free survival (PFS) and overall survival (OS). Survival was estimated with the Kaplan–Meier method and compared by log-rank tests. Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated using Cox proportional hazards models. Results: Median OS was 7.4 months (95% CI, 6.4–8.9) for ReDOS, 8.3 months (95% CI, 6.3–10.0) for &lt;160 mg, and 6.1 months (95% CI, 6.0–10.1) for 160 mg, with no significant differences (log-rank χ² = 1.95, p = 0.3781). HRs confirmed the lack of OS benefit: 0.90 (95% CI, 0.72–1.11) for ReDOS vs &lt;160 mg and 1.05 (95% CI, 0.87–1.25) for ReDOS vs 160 mg. In contrast, PFS differed significantly across groups (log-rank χ² = 11.77, p = 0.0028). Median PFS was 3.1 months (95% CI, 3.0–35.6) for ReDOS, 3.8 months (95% CI, 3.4–28.1) for &lt;160 mg, and 3.9 months (95% CI, 3.5–66.2) for 160 mg. HRs indicated a reduced risk of progression or death with ReDOS compared with &lt;160 mg (HR 0.83, 95% CI, 0.67–1.02) and 160 mg (HR 0.75, 95% CI, 0.63–0.89). Conclusions: In this large real-world cohort, initial regorafenib dosing strategy did not influence OS, but PFS outcomes significantly favored the ReDOS escalation approach, which was associated with a reduced risk of progression compared with the standard dose. These results support the potential clinical value of gradual dose escalation in optimizing regorafenib therapy in refractory mCRC.

Abstract non disponibile

Istituto di Igiene, Universita di Pavia; Dottorato di Ricerca in Pedagogia delle scienze della salute; Cooperative Incontro di Genova; Lega Italiana per la lotta contro i Tumor, Sezione di Imperia, Italy.

不登校は近年、多くの先進国において深刻な教育的・社会的課題となっている。既存研究では、不安障害や抑うつ、家庭環境の問題、いじめ、人間関係の困難、学業上のプレッシャーなどが主要な要因として指摘されている。しかしながら、これらの要因のみでは十分に説明できない一群の子どもたちが存在する。 本稿では、他者の感情や対人的緊張に対する高い感受性、集団環境における疲弊しやすさ、評価や比較への強いストレス反応を示す子どもたちに着目する。これらの特徴は、Aronによって提唱された感覚処理感受性(Sensory Processing Sensitivity:SPS)の概念と多くの共通点を有している。 本稿は、このような高感受性児における不登校を、個人の病理としてではなく、気質と教育環境との相互作用に起因する構造的ミスマッチとして捉える「教育環境ミスマッチ仮説(Educational Mismatch Hypothesis)」を提示するものである。 さらに、30年以上にわたる東洋医学臨床における観察をもとに、高感受性を持つ子どもたちが競争的・評価中心的な教育環境の中でどのようなストレスを受けやすいのかを検討する。また、OMN™(Oriental Medicine Neurodynamics)の視点から、高感受性を「内向的な情報処理傾向」「対人情報への高い受信性」「持続的な競争刺激に対する低耐性」として機能的に再解釈する。 本稿は、不登校を単なる適応障害や精神病理として理解するのではなく、個人の気質特性と教育環境との適合性という観点から再検討することを目的とする。さらに、教育環境の多様化と個別化に向けた新たな視点を提示し、今後の研究および教育実践への示唆を与えることを目指す。

不登校は近年、多くの先進国において深刻な教育的・社会的課題となっている。既存研究では、不安障害や抑うつ、家庭環境の問題、いじめ、人間関係の困難、学業上のプレッシャーなどが主要な要因として指摘されている。しかしながら、これらの要因のみでは十分に説明できない一群の子どもたちが存在する。 本稿では、他者の感情や対人的緊張に対する高い感受性、集団環境における疲弊しやすさ、評価や比較への強いストレス反応を示す子どもたちに着目する。これらの特徴は、Aronによって提唱された感覚処理感受性(Sensory Processing Sensitivity:SPS)の概念と多くの共通点を有している。 本稿は、このような高感受性児における不登校を、個人の病理としてではなく、気質と教育環境との相互作用に起因する構造的ミスマッチとして捉える「教育環境ミスマッチ仮説(Educational Mismatch Hypothesis)」を提示するものである。 さらに、30年以上にわたる東洋医学臨床における観察をもとに、高感受性を持つ子どもたちが競争的・評価中心的な教育環境の中でどのようなストレスを受けやすいのかを検討する。また、OMN™(Oriental Medicine Neurodynamics)の視点から、高感受性を「内向的な情報処理傾向」「対人情報への高い受信性」「持続的な競争刺激に対する低耐性」として機能的に再解釈する。 本稿は、不登校を単なる適応障害や精神病理として理解するのではなく、個人の気質特性と教育環境との適合性という観点から再検討することを目的とする。さらに、教育環境の多様化と個別化に向けた新たな視点を提示し、今後の研究および教育実践への示唆を与えることを目指す。

Recent progress in Natural Language Processing (NLP) has driven the creation of Large Language Models (LLMs) capable of tackling a vast range of tasks. A critical property of these models is their ability to handle large documents and process long token sequences, which has fostered the need for a robust evaluation methodology for long-text scenarios. To meet this requirement in the context of the Russian language, we present our benchmark consisting of 18 datasets designed to assess LLM performance in tasks such as information retrieval, knowledge extraction, machine reading, question answering, and reasoning. These datasets are categorized into four levels of complexity, enabling model evaluation across context lengths up to 128k tokens. To facilitate further research, we provide open-source datasets, a codebase, and a public leaderboard associated with the benchmark.

Background: Hamstring injuries are the leading cause of time lost in professional soccer. Among these, tendon-related lesions of the biceps femoris (BF) are associated with longer recovery and higher reinjury risk. Emerging magnetic resonance imaging (MRI) findings suggest the existence of morphological subtypes with different prognostic implications. Purpose: To describe an MRI pattern of proximal BF tendon injury-the "Bunny Tail lesion"-and report its management and outcomes in elite soccer players. Study Design: Case series; Level of evidence, 4. Methods: Over 2 competitive seasons (July 2023-June 2024), all acute hamstring injuries in a Serie A club were prospectively recorded and retrospectively reviewed. Players were eligible if MRI showed focal peritendinous edema at the distal tip of the proximal tendon of the BF long head, with increased intratendinous signal and preserved tendon continuity. Three musculoskeletal radiologists independently classified lesions using the British Athletics Muscle Injury Classification (BAMIC). Interobserver agreement was assessed with Fleiss κ. Rehabilitation followed a structured, football-specific program with standardized return-to-training and return-to-play (RTP) criteria. Results: Eight professional players met inclusion criteria. All lesions showed a well-circumscribed ovoid peritendinous edema surrounding the distal extremity of the proximal BF tendon, with preserved structural continuity. Lesions were variably categorized as BAMIC 2b to 2c, with one graded 3c by a single radiologist. Interobserver agreement for combined grade and tissue classification was very low [Fleiss κ close to 0]. All athletes were managed conservatively and returned to training after 14.6 ± 3.0 days and to play after 26 ± 6 days. No reinjuries occurred during a mean follow-up of 395 ± 202 days. Conclusion: This small series describes a consistent MRI pattern of proximal BF tendon injury characterized by focal ovoid peritendinous edema at the distal tendon tip, preserved continuity, and relatively rapid RTP despite predominantly tendinous BAMIC classification. This pattern may represent a less severe variant within the spectrum of proximal BF tendon injuries.

Membership Inference Attacks (MIAs) aim to determine whether a specific data point was included in the training set of a target model. Although there are have been numerous methods developed for detecting data contamination in large language models (LLMs), their performance on multimodal LLMs (MLLMs) falls short due to the instabilities introduced through multimodal component adaptation and possible distribution shifts across multiple inputs. In this work, we investigate multimodal membership inference and address two issues: first, by identifying distribution shifts in the existing datasets, and second, by releasing an extended baseline pipeline to detect them. We also generalize the perturbation-based membership inference methods to MLLMs and release FiMMIA — a modular Framework for Multimodal MIA. Our approach trains a neural network to analyze the target model's behavior on perturbed inputs, capturing distributional differences between members and non-members. Comprehensive evaluations on various fine-tuned multimodal models demonstrate the effectiveness of our perturbation-based membership inference attacks in multimodal domains.

An experience in managing a lodging house for young psychotic girls is described. Model is referred to the relational theory. Meetings with families and educators, institutional crises and results are considered. Results pointed out the importance of managing on the Department territory in order to treat and rehabilitize psychotic patients.