University of Udine
Archivio istituzionale della ricerca - Università degli Studi di UdineNot a member yet
91472 research outputs found
Sort by
Data-driven discovery of delay differential equations with discrete delays
Get PDFThe Sparse Identification of Nonlinear Dynamics (SINDy) framework is a robust method for identifying governing equations, successfully applied to ordinary, partial, and stochastic differential equations. In this work we extend SINDy to identify delay differential equations by using an augmented library that includes delayed samples and Bayesian optimization. To identify a possibly unknown delay we minimize the reconstruction error over a set of candidates. The resulting methodology improves the overall performance by remarkably reducing the number of calls to SINDy with respect to a brute force approach. We also address a multivariate setting to identify multiple unknown delays and (non-multiplicative) parameters. Several numerical tests on delay differential equations with different long-term behavior, number of variables, delays, and parameters support the use of Bayesian optimization highlighting both the efficacy of the proposed methodology and its computational advantages. As a consequence, the class of discoverable models is significantly expanded
Optimization of mechanical processing for sustainability
Get PDFDuring the last few years, European manufacturing has encountered several challenges in obtaining Agile production and reducing environmental impact. The importance of joining the production with a Zero Defects and Zero Waste policy has been a difficult challenge.
In a typical scenario, one or more defects could occur during the production of a manufacturing product, requiring the intervention of a specialized worker to detect the defects and decide whether the product should continue production or be wasted.
The latter possibility increases material waste, and the presence of a specialized worker who needs to stop the production process from detecting the selected area could increase production consumption.
The research objective was to study the possible implementation of an automated detective process that could increase the accuracy of detecting defects without increasing the consumption and time of application. Moreover, the possibility of identifying the defects present in a single piece allows one to direct the defective piece to a different production line or to be repaired without wasting the whole material. In this case, the detection time could increase the possibility of not wasting the material.
The study investigates various information fusion techniques to enhance defect detection accuracy and efficiency. Several strategies are employed, including decision fusion based on Dempster-Shafer's theory with a new measure of uncertainty to improve reliability. The research also integrates Selective Kernel and Multi-Head Attention layers into defect detection and segmentation models to further boost their accuracy and effectiveness.
During the testing phase, industrial datasets provided promising results and additional tests were performed across different domains to validate the effectiveness of the proposed approaches.During the last few years, European manufacturing has encountered several challenges in obtaining Agile production and reducing environmental impact. The importance of joining the production with a Zero Defects and Zero Waste policy has been a difficult challenge.
In a typical scenario, one or more defects could occur during the production of a manufacturing product, requiring the intervention of a specialized worker to detect the defects and decide whether the product should continue production or be wasted.
The latter possibility increases material waste, and the presence of a specialized worker who needs to stop the production process from detecting the selected area could increase production consumption.
The research objective was to study the possible implementation of an automated detective process that could increase the accuracy of detecting defects without increasing the consumption and time of application. Moreover, the possibility of identifying the defects present in a single piece allows one to direct the defective piece to a different production line or to be repaired without wasting the whole material. In this case, the detection time could increase the possibility of not wasting the material.
The study investigates various information fusion techniques to enhance defect detection accuracy and efficiency. Several strategies are employed, including decision fusion based on Dempster-Shafer's theory with a new measure of uncertainty to improve reliability. The research also integrates Selective Kernel and Multi-Head Attention layers into defect detection and segmentation models to further boost their accuracy and effectiveness.
During the testing phase, industrial datasets provided promising results and additional tests were performed across different domains to validate the effectiveness of the proposed approaches
L'enigma del lavoro povero
No full textIl contributo riproduce, e aggiorna, il mio intervento in occasione della presentazione dei risultati della ricerca sui “WORKING POOR N.E.E.D.S.: New Equity, Decent work and Skills”. Lo scritto riflette sull’attuale evoluzione del fenomeno del lavoro povero e sulle politiche di contrasto tenendo conto dei bisogni dei lavoratori, a partire da quello di una manutenzione della professionalità per l’occupazione
Adjuvant VaccInation After Conization for the Treatment for CervicAL Dysplasia
No full text: This study aimed to evaluate the role of adjuvant HPV vaccination in women undergoing conization for cervical intraepithelial neoplasia. This prospective study assessed factors influencing recurrence in patients undergoing conization for high-grade cervical dysplasia. After conization, patients were counseled on the potential benefits of vaccination. We compared outcomes between two groups: women who underwent conization with adjuvant human papillomavirus (HPV) vaccination and observation versus conization with observation only. Data from 281 patients were analyzed, comprising 168 (59.8%) patients in the conization-only group and 113 (40.2%) patients in the conization-plus vaccination group. Vaccinated patients were younger than nonvaccinated patients (38 vs. 45 years, P < 0.001). Positive surgical margins were more frequently observed in the vaccinated group compared with the nonvaccinated group (9.7 vs. 3.6%; P = 0.038). Median follow-up was shorter in the vaccinated group, although this difference was not statistically significant (24.9 vs. 27.8 months; P = 0.395). The risk of developing HPV-related lesions was similar between the vaccinated and nonvaccinated groups (P = 0.594, log-rank test). Likewise, the need for reconization did not differ significantly between the groups (P = 0.593, log-rank test). Multivariate analysis showed no significant impact of HPV vaccination on postoperative outcomes [hazard ratio (HR): 0.50, 95% confidence interval (CI): 0.15-1.68) for any lesion; HR: 0.90, 95% CI: 0.47-1.73 for reconization]. This study indicates that adjuvant HPV vaccination does not significantly affect short-term outcomes in women undergoing conization for cervical dysplasia. Ongoing randomized trials will provide more robust evidence to clarify the role of adjuvant vaccination in this setting
Automatic Coding of Clinical Documents: Leveraging Symbolic and Sub-Symbolic Approaches for Enhanced Interpretability and Explainability
Get PDFThe healthcare industry is undergoing an unprecedented surge in the volume of medical data generated on a daily basis. This exponential growth is driven by several factors, including the widespread adoption of electronic health records and applications that continuously generate streams of patient data. Medical data now encompasses a diverse array of structured and unstructured formats. Considering that a significant part of medical data exists in unstructured, free-text form, data processing systems encounter significant challenges in effectively utilizing this information. Thus, assigning standardized meanings to these textual expressions becomes crucial in the fields of epidemiology, statistics, and health informatics. This standardization is critical because it enables the automated processing and analysis of data, provides the requisite information to make informed decisions, and facilitates the implementation of effective public health policies. In the medical field, this is typically achieved by coding and classifying text using appropriate terminologies and classifications.
Traditionally and in many current scenarios, this process has been performed manually by trained professionals, but even when executed meticulously, it is laborious, time-consuming, and prone to human error. To assist practitioners, two main approaches have been utilized to enhance the coding process: symbolic and subsymbolic techniques. Symbolic techniques employ logical reasoning and human-defined rules to produce deterministic outcomes, while subsymbolic techniques employ machine learning and neural networks to manage extensive, noisy datasets and to learn adaptively from patterns within the data.
The aim of this work is multifold and focuses on supporting the automation of clinical coding through symbolic and sub-symbolic approaches and enhancing the interpretability of sub-symbolic methods.
Specifically, we presented the architecture of a novel rule-based system for the automated selection of the so-called Underlying Cause of Death (UCOD), using classification-independent rules. This was followed by a preliminary validation on two datasets of death certificates coded with the International Statistical Classification of Diseases and Related Health Problems, 10th (ICD-10) and 11th (ICD-11) revisions.
Secondly, we addressed the coding of death certificates using sub-symbolic approaches, focusing on converting textual conditions into ICD-10 codes and identifying the UCOD. We proposed a novel method that outperformed state-of-the-art systems for UCOD selection from death certificates, leveraging natural language processing algorithms. Specifically, we compared various techniques applied to tabular and textual data, including logistic regression, random forest, XGBoost, feedforward neural networks with categorical embeddings, and transformers. Through extensive comparative experiments, we found that the fine-tuned Mistral model significantly outperformed other transformer models, particularly in data-limited scenarios.
Thirdly, we aimed to enhance the interpretability of deep learning models by ensuring proper calibration. We introduced mechanisms to rank instances based on difficulty using variance of gradients. Furthermore, we proposed a model based on text-to-text transformers that generated human-readable explanations, learning from the rule-based system to support the explainability of deep learning models.
Finally, we trained a disease-related language model by creating a pre-training corpus based on ICD-11 and tackled the challenge of linking clinical notes to SNOMED CT by leveraging Large Language Models and Retrieval-Augmented Generation
Influence of Surgical Expertise on Repair of Acute Type a Aortic Dissection in a Standardized Operative Setting
Get PDFBackground/Objectives: The influence of surgeon expertise on patients’ outcomes after repair of acute type A aortic dissection (ATAAD) is not well established. The aim of this paper is to report the results of ATAAD repair performed by expert (ES) and not expert aortic surgeons (NES) in our center. Methods: We have retrospectively divided 199 patients into two groups according to the first surgeon experience (ES = 138 patients and NES = 61 patients), all being members of the aortic team. We evaluated and compared early and long-term outcomes. Results: Although the two groups were comparable in terms of clinical presentation and intraoperative setting, ES performed more aortic arch repairs (40% vs. 26%, p = 0.06) and frozen elephant trunk procedures (15% vs. 3%, p = 0.02), albeit with similar intraoperative times (median cardiopulmonary bypass time of 203 min in ES vs. 201 min in NES, respectively, p = 0.88). The 30-day mortality was the same in the two groups (8%, p = 1), and the postoperative course was similar except for a shorter in-hospital stay in the NES group (10 vs. 17 days, p = 0.04). Conclusions: In our experience, repair of ATAAD could be achieved with similar results between ES and NES. However, NES performed less technically demanding repairs. With standardization of pre-, intra-, and post-operative management, NES can be expected to increase their technical skills and be progressively involved in more complex ATAAD repairs without adversely affecting surgical results
Updates, Applications and Future Directions of Deep Learning for the Images Processing in the Field of Cranio-Maxillo-Facial Surgery
Get PDFThe entry of artificial intelligence, in particular deep learning models, into the study of medical–clinical processes is revolutionizing the way of conceiving and seeing the future of medicine, offering new and promising perspectives in patient management. These models are proving to be excellent tools for the clinician through their great potential and capacity for processing clinical data, in particular radiological images. The processing and analysis of imaging data, such as CT scans or histological images, by these algorithms offers aid to clinicians for image segmentation and classification and to surgeons in the surgical planning of a delicate and complex operation. This study aims to analyze what the most frequently used models in the segmentation and classification of medical images are, to evaluate what the applications of these algorithms in maxillo-facial surgery are, and to explore what the future perspectives of the use of artificial intelligence in the processing of radiological data are, particularly in oncological fields. Future prospects are promising. Further development of deep learning algorithms capable of analyzing image sequences, integrating multimodal data, i.e., combining information from different sources, and developing human–machine interfaces to facilitate the integration of these tools with clinical reality are expected. In conclusion, these models have proven to be versatile and potentially effective tools on different types of data, from photographs of intraoral lesions to histopathological slides via MRI scans
Multivariate Bayesian Global – Local Shrinkage Methods for Regularisation in the High - Dimensional Linear Model
Get PDFThis paper considers Bayesian regularisation using global–local shrinkage priors in the multivariate general linear model when there are many more explanatory variables than observations. We adopt priors’ structures used extensively in univariate problems (conjugate and non-conjugate with tail behaviour ranging from polynomial to exponential) and consider how the addition of error correlation in the multivariate set-up affects the
performance of these priors. Two different datasets (from drug discovery and chemometrics) with many covariates are used for comparison, and these are supplemented by a small simulation study to corroborate the role of error correlation. We find that structural assumptions of the prior distribution on regression coefficients can be more significant than the tail behaviour. In particular, if the structural assumption of conjugacy is used, the performance of the posterior predictive distribution deteriorates relative to non-conjugate choices as the error correlation becomes stronger