University of Udine
Archivio istituzionale della ricerca - Università degli Studi di UdineNot a member yet
91472 research outputs found
Sort by
Reproducibility Companion Paper: AdOCTeRA - Adaptive Optimization Constraints for Improved text-guided Retrieval of Apartments
No full textThis reproducibility Companion paper supports the approach presented in our paper titled ''AdOCTeRA: Adaptive Optimization Constraints for Improved Text-Guided Retrieval of Apartments.'' In that work, we addressed the problem of apartment retrieval using textual descriptions. Specifically, we proposed a novel adaptive approach that leverages the similarity between apartment descriptions in the dataset to enforce different levels of distance-a small distance between similar elements, a slightly larger distance between less similar elements, and an even larger distance between dissimilar elements
Generative artificial intelligence in laboratory medicine: innovations, limitations, and ethical considerations [Intelligenza artificiale generativa nella medicina di laboratorio: innovazioni, limiti e considerazioni etiche]
No full textGenerative artificial intelligence (AI) has the potential to transform Laboratory Medicine by enhancing data processing, report generation, and research productivity. Large language models (LLMs), can generate coherent text, extract critical information from complex datasets, and assist in scientific writing, significantly reducing the time and effort required for these tasks. By streamlining these processes, AI might improve the efficiency and accuracy of clinical operations and research. However, the integration of AI in healthcare raises significant challenges, including the risk of AI producing erroneous outputs, known as “hallucinations,” where the generated content is inaccurate or fabricated, highlighting the need for human oversight. In addition to automating text summarization and document management, AI is powerful for data analysis, lowering technical barriers and making advanced statistical processing more accessible to laboratory professionals. These capabilities are particularly valuable in a field where the ability to process large datasets is critical for research potentially impacting on diagnostics. Despite these benefits, ethical concerns such as potential plagiarism, automation bias, and data privacy risks remain pressing issues, given that AI relies on large datasets, including sensitive patient information. This manuscript critically examines the current state of AI in Laboratory Medicine, addressing its potential to innovate the field, while outlining the ethical and practical considerations that must be addressed to ensure its responsible implementation. The future of AI in Laboratory Medicine depends on developing robust frameworks for its use and training healthcare professionals to leverage these tools effectively, ensuring they augment rather than replace human expertise
Ultrasonographic scores and histopathology correlations in primary Sjögren’s syndrome
Get PDFTitle: Ultrasonographic scores and histopathology correlations in primary Sjögren’s syndrome
Introduction:
Primary Sjögren's syndrome (pSS) is a chronic systemic autoimmune disease characterized by
exocrine glands involvement, leading to sicca symptoms. The glandular damage exposes pSS patients to an increased risk of lymphoproliferative disease.
The diagnostic approach relies on a set of clinical, laboratory, imaging and histopathological features, and salivary gland biopsy is considered the gold standard tool, playing a key role in the classification criteria for pSS.
Even if not included in the pSS classification criteria, salivary gland ultrasonography is emerging as a valuable complement tool to histopathology, allowing the identification of alterations suspicious for pSS disease and for lymphoma. However, the correlation between histopathological and ultrasonographic features remains incompletely elucidated in the literature.
Aim of the study is to analyze the correlation between the ultrasonographic and histological features of major salivary glands in terms of inflammatory infiltrate, fibrosis, and presence of adipose tissue.
Methods: Adult patients with pSS and a clinical indication for major salivary gland biopsy between September 2018 and October 2023 were selected for this study. All patients underwent major salivary gland ultrasonography and performed a major salivary gland biopsy. For each histopathological sample, the evaluation included the presence of lymphoepithelial lesions (LELs), germinal centers (GCs), adipose replacement area and focus score, as well as the presence of Multiple Focal Lymphocytic Sialadenitis/ Lymphoepithelial Sialadenitis (MESA/LESA). Each ultrasound image is assessed according to the Hocevar and OMERACT scoring method.
Results:
63 patients were included in this study. Correlations between ultrasonography and histopathological features were registered. The 0-48 Hocevar score was correlated to the presence of GCs (p=0.009). Aiming at a direct comparison between histopathological findings and the ultrasonographic score of the same major salivary gland, we compare the ultrasonographic score of the biopsied gland with the histopathological results. The gland specific Hocevar score and the OMERACT score both showed correlation with the presence of LELs and GCs (p=0.004 and p<0.001, respectively).
Focusing on patients without pSS complication (i.e. lymphoma patients), we reported significant correlation between Hocevar score and LELs and MESA/LESA (p=0.05). Analysing in this cohort the direct comparison between histopathological findings and ultrasonographic score (Hocevar and OMERACT score) of the same salivary gland, we reported associations with LELs and GCs in both scoring system.
Conclusion:
The ultrasonographic scores of patients with pSS correlate with histopathological features, notably LELs and GCs, which are typical characteristics of pSS. This study suggests the potential utility of SGUS as an alternative for histopathological examination
A Review on the Use of Plastic Waste as a Modifier of Asphalt Mixtures for Road Constructions
Get PDFRising industrialization and population growth contribute to the increasing generation of plastic waste, which poses significant environmental and health challenges. Despite its potential as a resource, plastic waste is often discarded without proper treatment. Repurposing it in road construction offers both economic and environmental benefits, providing a sustainable waste management solution. This paper thoroughly examines various types of plastic waste used in asphalt mixtures, considering both wet and dry processing methods and their impact on bituminous binders and asphalt performance. Overall, incorporating waste plastics into asphalt mixtures has been shown to improve fatigue resistance, rutting resistance, moisture resistance, and high-temperature performance. However, challenges related to compatibility and low-temperature performance persist in plastic-modified asphalt applications. To address these issues, modified approaches, such as the use of chemical additives, have been identified as effective in enhancing the bonding between waste plastics and bituminous binders while also increasing the amount of plastic that can be incorporated. While plastic-modified asphalt shows significant promise, overcoming these challenges through targeted research and careful implementation is essential for its sustainable and effective use in asphalt mixtures, ensuring long-term performance
The role of molecular classification in endometrial cancer: a focus on patients with low-volume metastasis—a retrospective multi-center study and literature review
No full textObjective: The prognosis and optimal management of patients with endometrial cancer with low-volume lymph node metastases remains unclear, particularly, regarding the role of molecular classification. This study investigates the distribution and prognostic role of molecular classes in this cohort of patients, while reviewing existing literature on this topic. Methods: A multi-center, retrospective cohort study involving 8 institutions evaluated 134 patients with endometrial cancer and low-volume metastasis (isolated tumor cells: ≤0.2 mm in diameter or micro-metastasis: >0.2 to ≤2.0 mm). All patients were categorized into 1 of 4 molecular classes: POLE mutated (POLEmut), mismatch repair-deficient (MMRd), p53 abnormal (p53abn), and non-specific molecular profile (NSMP). Recurrence-free survival was estimated using Kaplan-Meier methods and differences between molecular classes were tested using log-rank test. Cox proportional hazards regression models were fit to evaluate the association between molecular class and recurrence after surgery. Results: Of 134 patients, 78 (58.2%) had isolated tumor cells and 56 (41.8%) had micro-metastasis. Molecular classification revealed 78 (58.2%) NSMP, 37 (27.6%) MMRd, 15 (11.2%) p53abn, and 4 (3.0%) POLEmut tumors. No significant differences in recurrence-free survival were observed among molecular classes (log-rank p = .52). At 5 years, recurrence-free survival rates were 84.6% for NSMP, 65.9% for MMRd, and 76.9% for p53abn. No recurrences occurred in POLEmut patients. Predictors of recurrence included micro-metastasis (p = .01) and lympho-vascular space invasion (p = .02), whereas molecular classification was not independently associated with recurrence. The literature review revealed insufficient data on this topic to draw definitive conclusions. Conclusions: Although molecular classification enhances risk stratification in endometrial cancer, its impact should be contextualized in patients with low-volume lymph node metastasis. Pathological factors, such as lymph node metastasis size and lympho-vascular space invasion, continue to represent a key predictor of recurrence. Integrating molecular data with traditional risk factors remains pivotal, especially in intermediate-risk molecular classes, to refine management strategies
Comparison of Body Composition, Basal Metabolic Rate and Metabolic Outcomes of Adults with Prader-Willi Syndrome and Age- and BMI-Matched Patients with Essential Obesity
Get PDFBackground/Objectives: This study compared metabolic syndrome (MetS) features in patients with Prader-Willi syndrome (PWS) to those in age-, BMI-, and gender-matched subjects with essential obesity (EOB). Methods: Thirty-two PWS patients (23 females, 9 males; median age 31.6 years; BMI 42.0 kg/m2) underwent several assessments, including anthropometric measurements, body composition via bio-impedance analysis, basal metabolic rate (BMR) using indirect calorimetry, and blood sampling. Results: Their data were compared to a matched EOB group (23 females, 9 males; median age 31.4 years; BMI 43.5 kg/m2). The study groups (PWS and EOB) were subsequently divided into two subgroups based on the International Diabetes Federation criteria for the definition of MetS. Results showed that individuals with PWS had significantly lower (p < 0.001) body weight (BW, −20.9%), height (−8.9%), fat-free mass (FFM, −23.5%), and fat mass (FM, −19.2%) in absolute terms compared to EOB subjects. However, the relative percentages of FFM and FM were similar. Absolute BMR was 25.5% (p < 0.001) lower in the PWS group; however, this difference disappeared when adjusted for FFM or body weight (BW). Metabolic outcomes were broadly similar between the groups, except for higher fasting glucose (+7.3%) and HbA1c levels (+7.9%), and lower fasting insulin (−29.0%) in PWS patients. Conclusions: Moreover, PWS subjects exhibited higher total cholesterol (+9.6%) and HDL-cholesterol (+19.8%), suggesting a more favourable lipid profile and no extra risk beyond severe obesity
Giuseppe Giusti, poeta del dissenso
No full textIl contributo vuole delineare l’atteggiamento ironico e satiro presente nella produzione lirica del poeta, che utilizza il genere per contribuire a creare un senso critico e una possibilità di discussione aperta, nella quale lui stesso non si pone come il poeta-vate, ma come uomo fra gli uomini, disposto a mettersi in discussione, e ad aprire un discorso problematico e di lettura critica verso la realtà che lo circonda
Nonlinear Oscillations of a Bistable Composite Shell: Experimental Studies
No full textThe dynamics of a cantilever bistable composite shell is investigated in the paper. The shell shape is derived from a region of a cone surface and is composed of eight composite layers. The proposed shell’s geometry and the asymmetric configurations of the layers give rise to the structure with bistability. A prototype of the shell is tested experimentally to detect nonlinear dynamic phenomena in the vicinity of each equilibrium point (in-well dynamics) and to study global (cross-well) oscillations. The experimental campaign shows strong softening resonance characteristics around I-shape and almost linear resonance curves for C-shape. Upon increasing the amplitude of excitation reversible snap-through and nonlinear chaotic oscillations are observed
Impersonating the Crowd: Evaluating LLMs' Ability to Replicate Human Judgment in Misinformation Assessment
Get PDFLarge Language Models (LLMs) are increasingly used to replicate human decision-making in subjective tasks. In this work, we investigate whether LLMs can effectively impersonate real crowd workers when evaluating political misinformation statements. We assess (i) the agreement between LLM-generated assessments and human judgments and (ii) whether impersonation skews LLM assessments, impacting accuracy. Using publicly available misinformation assessment datasets, we prompt LLMs to impersonate real crowd workers based on their demographic profiles and evaluate them under the same statements. Through comparative analysis, we measure agreement rates and discrepancies in classification patterns. Our findings suggest that while some LLMs align moderately with crowd assessments, their impersonation ability remains inconsistent. Impersonation does not uniformly improve accuracy and often reinforces systematic biases, highlighting limitations in replicating human judgment
Effect of Combined Proteinuria and Increased Renal Resistive Index on Chronic Kidney Disease Progression: A Retrospective Longitudinal Study
Get PDFIntroduction: An increased renal resistive index (RRI) and proteinuria can predict an estimated glomerular filtration rate (eGFR) decline in patients with chronic kidney disease (CKD) of various causes. This study hypothesized that the RRI and proteinuria interact to determine disease progression in patients with CKDs of unknown origin. Patients and Methods: One hundred and fifty six patients (age 76.0 ± 8.1 years, 63.5% males) were analyzed for anthropometric, kidney morphology, blood pressure, 24 h urinary protein excretion, and RRI. The CKD-EPI equation was used to calculate the eGFR at baseline and after a two-year follow-up. Patients with an elevated (≥0.80) or normal (<0.80) RRI and significant (≥150 mg/day) or physiological (<150 mg/day) proteinuria were evaluated for the likelihood of at least a 30% drop in the eGFR or the onset of end-stage kidney disease (endpoint). Results: Hypertension and diabetes were the predominant cardiovascular risk factors (90.4%). Fifty patients (32%) met the endpoint. Elevated RRIs (odds ratio, OR, 4.28; 95% confidence interval, CI, 1.82–10.6; p = 0.001) and significant proteinuria (OR 3.59, 95% CI 1.59–8.48, p = 0.003) were independent predictors of the endpoint in a multivariate logistic model. Patients with an elevated RRI and significant proteinuria were more likely to meet the endpoint (R1P1: 65.2%) compared to those with only proteinuria (R0P1: 39.5%, p = 0.043) or both normal factors (R0P0: 10.9%, p < 0.001) but not to those with only an elevated RRI (R1P0: 42.3%, p = 0.094). Continuous RRIs (partial correlation r = −0.245, p < 0.001) and 24 h urinary protein excretion (partial r = −0.226, p = 0.003) were inversely and independently correlated with eGFR% change. R1P1 showed a higher eGFR% reduction (−38.0% ± 20.4%) compared to R0P1 (−25.3% ± 19.0%, p = 0.043) and R0P0 (−8.8% ± 25.1%, p < 0.001) but not to R1P0 (−29.6% ± 21.0%, p = 0.192). Conclusions: An increased RRI and proteinuria were independent predictors of disease progression. When interaction was considered, the negative effect of an elevated RRI on CKD progression was evident in both proteinuric and non-proteinuric patients, whereas the negative effect of proteinuria on disease progression was only significant in patients with no elevated RRIs