Archivio Istituzionale della Ricerca - Università degli Studi di Pavia
Not a member yet
135341 research outputs found
Sort by
Startup innovative, piccole e medie imprese innovative e scaleup in Italia
No full textIl contributo mira a verificare l'impatto degli incentivi concessi dalla legislazione italiana alle startup e piccole e medie imprese innovative, verificando quante di esse si sono tramutate in scaleup, raccogliendo significavi finanziamenti di mercat
Global, regional, and national burden of household air pollution, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021
No full textBackground: Despite a substantial reduction in the use of solid fuels for cooking worldwide, exposure to household air pollution (HAP) remains a leading global risk factor, contributing considerably to the burden of disease. We present a comprehensive analysis of spatial patterns and temporal trends in exposure and attributable disease from 1990 to 2021, featuring substantial methodological updates compared with previous iterations of the Global Burden of Diseases, Injuries, and Risk Factors Study, including improved exposure estimations accounting for specific fuel types. Methods: We estimated HAP exposure and trends and attributable burden for cataract, chronic obstructive pulmonary disease, ischaemic heart disease, lower respiratory infections, tracheal cancer, bronchus cancer, lung cancer, stroke, type 2 diabetes, and causes mediated via adverse reproductive outcomes for 204 countries and territories from 1990 to 2021. We first estimated the mean fuel type-specific concentrations (in μg/m3) of fine particulate matter (PM2·5) pollution to which individuals using solid fuels for cooking were exposed, categorised by fuel type, location, year, age, and sex. Using a systematic review of the epidemiological literature and a newly developed meta-regression tool (meta-regression: Bayesian, regularised, trimmed), we derived disease-specific, non-parametric exposure-response curves to estimate relative risk as a function of PM2·5 concentration. We combined our exposure estimates and relative risks to estimate population attributable fractions and attributable burden for each cause by sex, age, location, and year. Findings: In 2021, 2·67 billion (95% uncertainty interval [UI] 2·63-2·71) people, 33·8% (95% UI 33·2-34·3) of the global population, were exposed to HAP from all sources at a mean concentration of 84·2 μg/m3. Although these figures show a notable reduction in the percentage of the global population exposed in 1990 (56·7%, 56·4-57·1), in absolute terms, there has been only a decline of 0·35 billion (10%) from the 3·02 billion people exposed to HAP in 1990. In 2021, 111 million (95% UI 75·1-164) global disability-adjusted life-years (DALYs) were attributable to HAP, accounting for 3·9% (95% UI 2·6-5·7) of all DALYs. The rate of global, HAP-attributable DALYs in 2021 was 1500·3 (95% UI 1028·4-2195·6) age-standardised DALYs per 100 000 population, a decline of 63·8% since 1990, when HAP-attributable DALYs comprised 4147·7 (3101·4-5104·6) age-standardised DALYs per 100 000 population. HAP-attributable burden remained highest in sub-Saharan Africa and south Asia, with 4044·1 (3103·4-5219·7) and 3213·5 (2165·4-4409·4) age-standardised DALYs per 100 000 population, respectively. The rate of HAP-attributable DALYs was higher for males (1530·5, 1023·4-2263·6) than for females (1318·5, 866·1-1977·2). Approximately one-third of the HAP-attributable burden (518·1, 410·1-641·7) was mediated via short gestation and low birthweight. Decomposition of trends and drivers behind changes in the HAP-attributable burden highlighted that declines in exposures were counteracted by population growth in most regions of the world, especially sub-Saharan Africa. Interpretation: Although the burden attributable to HAP has decreased considerably, HAP remains a substantial risk factor, especially in sub-Saharan Africa and south Asia. Our comprehensive estimates of HAP exposure and attributable burden offer a robust and reliable resource for health policy makers and practitioners to precisely target and tailor health interventions. Given the persistent and substantial impact of HAP in many regions and countries, it is imperative to accelerate efforts to transition under-resourced communities to cleaner household energy sources. Such initiatives are crucial for mitigating health risks and promoting sustainable development, ultimately improving the quality of life and health outcomes for millions of people. Funding: Bill & Melinda Gates Foundation
Artificial intelligence-enabled histology exhibits comparable accuracy to pathologists in assessing histological remission in Ulcerative Colitis: a systematic review, meta-analysis and meta-regression
No full textBackground and aims: Achieving histological remission is a desirable emerging treatment target in Ulcerative Colitis (UC), yet its assessment is challenging due to high inter- and intra-observer variability, reliance on experts, and lack of standardisation. Artificial intelligence (AI) holds promise in addressing these issues. This systematic review, meta-analysis, and meta-regression evaluated the AI's performance in assessing histological remission and compared it with that of pathologists. Methods: We searched Medline/PubMed and Scopus databases from inception to September 2024. We included studies on AI models assessing histological activity in UC, with or without comparison to pathologists. Pooled performance metrics were calculated: sensitivity, specificity, positive and negative predictive value (PPV & NPV), observed agreement, and F1 score. A pairwise meta-analysis compared AI and pathologists, while sub-meta-analysis and meta-regression evaluated heterogeneity and factors influencing AI performance. Results: Twelve studies met the inclusion criteria. AI models exhibited strong performance with a pooled sensitivity of 0.84 (95% CI 0.80-0.88), specificity 0.87 (0.84- 0.91), PPV 0.90 (0.87-0.92), NPV 0.80 (0.71-0.88), observed agreement 0.85 (0.82- 0.89), and F1 score 0.85 (0.82-0.89). AI models demonstrated no significant differences with pathologists for specificity, observed agreement and F1 score, while they were outperformed by pathologists for sensitivity and NPV. AI models for the adult population were linked to reduced heterogeneity and enhanced AI performance at meta-regression. Conclusion: AI shows significant potential for assessing histological remission in UC and performs comparably to pathologists. Future research should focus on standardised, large-scale studies to minimise heterogeneity and support widespread AI implementation in clinical practice
Post-earthquake structural damage detection with tunable semi-synthetic image generation
No full textIn the aftermath of an earthquake, conducting rapid structural safety assessments is essential. A Deep Learning-based damage detector capable of automatically analyzing videos from Unmanned Aircraft Systems (UAS) surveys would be highly beneficial for this purpose. Despite significant advancements in object detection using Deep Convolutional Neural Networks (DCNNs), developing an effective post-earthquake damage detector remains challenging due to the scarcity of large, annotated image datasets. In this work, we present a system to create a large number of images where artificial damage instances are applied to real-world three-dimensional (3D) models of buildings and bridges. We defined such images as semi-synthetic. The proposed method relies on the definition, made by human experts, of meta-annotations from which a variety of damage instances can be generated in a controlled way. Semi-synthetic images are designed to augment real-world datasets, enhancing the training process of a DCNN-based damage detector. This semi-synthetic image augmentation can be iteratively refined to target the most critical cases. Experiments conducted on the ‘Image Database for Earthquake damage Annotation’ (IDEA) dataset shown that a detector trained on a combination of real and semi-synthetic images performs better than one trained on real images alone. A damage detector trained using the proposed strategy was then incorporated into a system that analyzes and tracks multiple damage instances in UAS-acquired videos, generating concise summaries of the findings. The effectiveness of the system was validated by the analysis of post-earthquake UAS videos and the production of reports that were reviewed by structural engineering experts
Stability of radiomic features in magnetic resonance imaging of the female pelvis: A multicentre phantom study
No full textThe potential of Generative Pre-trained Transformer 4 (GPT-4) to analyse medical notes in three different languages: a retrospective model-evaluation study
No full textBackground: Patient notes contain substantial information but are difficult for computers to analyse due to their unstructured format. Large-language models (LLMs), such as Generative Pre-trained Transformer 4 (GPT-4), have changed our ability to process text, but we do not know how effectively they handle medical notes. We aimed to assess the ability of GPT-4 to answer predefined questions after reading medical notes in three different languages. Methods: For this retrospective model-evaluation study, we included eight university hospitals from four countries (ie, the USA, Colombia, Singapore, and Italy). Each site submitted seven de-identified medical notes related to seven separate patients to the coordinating centre between June 1, 2023, and Feb 28, 2024. Medical notes were written between Feb 1, 2020, and June 1, 2023. One site provided medical notes in Spanish, one site provided notes in Italian, and the remaining six sites provided notes in English. We included admission notes, progress notes, and consultation notes. No discharge summaries were included in this study. We advised participating sites to choose medical notes that, at time of hospital admission, were for patients who were male or female, aged 18–65 years, had a diagnosis of obesity, had a diagnosis of COVID-19, and had submitted an admission note. Adherence to these criteria was optional and participating sites randomly chose which medical notes to submit. When entering information into GPT-4, we prepended each medical note with an instruction prompt and a list of 14 questions that had been chosen a priori. Each medical note was individually given to GPT-4 in its original language and in separate sessions; the questions were always given in English. At each site, two physicians independently validated responses by GPT-4 and responded to all 14 questions. Each pair of physicians evaluated responses from GPT-4 to the seven medical notes from their own site only. Physicians were not masked to responses from GPT-4 before providing their own answers, but were masked to responses from the other physician. Findings: We collected 56 medical notes, of which 42 (75%) were in English, seven (13%) were in Italian, and seven (13%) were in Spanish. For each medical note, GPT-4 responded to 14 questions, resulting in 784 responses. In 622 (79%, 95% CI 76–82) of 784 responses, both physicians agreed with GPT-4. In 82 (11%, 8–13) responses, only one physician agreed with GPT-4. In the remaining 80 (10%, 8–13) responses, neither physician agreed with GPT-4. Both physicians agreed with GPT-4 more often for medical notes written in Spanish (86 [88%, 95% CI 79–93] of 98 responses) and Italian (82 [84%, 75–90] of 98 responses) than in English (454 [77%, 74–80] of 588 responses). Interpretation: The results of our model-evaluation study suggest that GPT-4 is accurate when analysing medical notes in three different languages. In the future, research should explore how LLMs can be integrated into clinical workflows to maximise their use in health care. Funding: None
Measurement of HΛ3 production in Pb–Pb collisions at sNN=5.02 TeV
No full textThe first measurement of HΛ3 and H ̅Λ ̅3 differential production with respect to transverse momentum and centrality in Pb–Pb collisions at sNN=5.02 TeV is presented. The HΛ3 has been reconstructed via its two-charged-body decay channel, i.e., HΛ3→3He+π−. A Blast-Wave model fit of the pT-differential spectra of all nuclear species measured by the ALICE collaboration suggests that the HΛ3 kinetic freeze-out surface is consistent with that of other nuclei. The ratio between the integrated yields of HΛ3 and He3 is compared to predictions from the statistical hadronisation model and the coalescence model, with the latter being favoured by the presented measurements
Investigating electrospun shape memory patches as Gentamicin drug delivery system
No full textGoal of this study is to apply electrospinning technology for manufacturing polymeric drug delivery systems with
thermal sensitive shape memory behaviour. The hypothesis is to obtain electrospun patches loaded with an
antibiotic drug to be implanted into the human body by minimal invasive technique, as local anti-infective
therapy for surgical site infections treatment. Polylactide-co-polycaprolactone 70:30 (PLLA-co-PCL 70:30) was
selected as temperature responsive polymer due to its Tg◦ value (32–42 ◦C) in the range of body temperature.
Gentamicin (GS) was selected because used in last-line therapy against multidrug-resistant bacteria with several
side effects upon its systemic administration.
After been manufactured, the electrospun patches underwent thermalshape memory thermal treatment (SMT)
applying different thermal conditions and they were characterized before and after SMT by SEM, DSC, mechanical
testing and antibacterial effect on S. Aureus clinical strains.
The results show that shape memory property of PLLA-co-PCL 70:30 patches is maintained both after GS
loading and SMT at 60 ◦C that did not affect both nanofiber morphology and drug release. A change in copolymer
conformation due to electrospinning occurs and GS loading, as highlighted by changes in patches thermal
behaviour. The matrices mechanical properties address their application to internal surgical wounds, mainly soft
tissues. Patches antimicrobial effect gave promising positive results for Gentamicin-susceptible strains, including
a clinical isolate