154092 research outputs found
Sort by
Multi-Modal Machine Learning for Evaluating the Predictive Value of Pelvimetric Measurements (Pelvimetry) for Anastomotic Leakage After Restorative Low Anterior Resection
Get PDFBackground/Objectives: Anastomotic leakage (AL) remains a major complication after restorative rectal cancer surgery, with accurate preoperative risk stratification posing a significant challenge. Pelvic measurements derived from magnetic resonance imaging (MRI) have been proposed as potential predictors of AL, but their clinical utility remains uncertain.Methods: This retrospective, multicenter cohort study analyzed rectal cancer patients undergoing restorative surgery between 2013 and 2021. Pelvic dimensions were assessed using MRI-based pelvimetry. Univariate and multivariate regression analyses identified independent risk factors for AL. Subsequently, machine Learning (ML) models—logistic regression, random forest classifier, and XGBoost—were developed to predict AL using preoperative clinical data alone and in combination with pelvimetry. Model performance was evaluated using F1 scores, with the area under the receiver operating characteristic (ROC-AUC) and precision–recall curves (AUC-PR) as primary metrics.Results: Among 487 patients, the overall AL rate was 14%. Multivariate regression analysis identified distance to the anorectal junction, pelvic inlet width, and interspinous distance as independent risk factors for AL (p < 0.05). The logistic regression model incorporating pelvimetry achieved the highest predictive performance, with a mean ROC-AUC of 0.70 ± 0.09 and AUC-PR of 0.32 ± 0.10. Although predictive models that included pelvic measurements demonstrated higher ROC-AUCs compared to those without pelvimetry, the improvement was not statistically significant.Conclusions: Pelvic dimensions, specifically pelvic inlet and interspinous distance, were independently associated with an increased risk of AL. While ML models incorporating pelvimetry showed only moderate predictive performance, these measurements should be considered in developing clinical prediction tools for AL to enhance preoperative risk stratification.</p
Integration of sensing/acoustofluidic functions and modulation of surface acoustic wave fields on printed circuit board
No full textIn this study, we utilised prototyping printed circuit boards (PCBs) as a surface acoustic wave (SAW) platform to integrate sensing and acoustofluidic functions and study acoustic wavefield modulation. The PCB is a sandwich structure with a woven glass fibre array encapsulated in an epoxy matrix between top/bottom copper layers. On top of this structure, we fabricated SAW devices by sputtering a piezoelectric thin film (using zinc oxide as an example) followed by patterning metal interdigital electrodes. We demonstrated that the thickness of top copper layer on the PCB, relative to the acoustic wavelength, plays a key role in achieving actuation and acoustofluidic functions. On PCB substrates with copper layers of 35 and 135 μm thick, we achieved multiple SAW sensing functions, with SAW wavelengths of 220 and 100 μm, respectively. However, efficient acoustofluidic actuation (e.g., droplet streaming and transportation) was only achieved on the SAW device with a 135 μm thick copper layer but not on that with the copper layer of 35 μm thick, because of epoxy's damping of acoustic energy. Additionally, we observed an intriguing phenomenon of SAW field modulation, and the composite structure of glass fibre array and the epoxy matrix led to spatially differentiated attenuation of acoustic energy. The weakened attenuation in the glass fibres and enhanced attenuation in the epoxy matrix resulted in striped patterns of SAW fields on the substrate, parallel to the glass fibres. This means that the SAW field geometries can be flexibly modulated by designing structured substrates to meet diverse application needs. The multi-functional integration and modulation of SAW field geometry, combined with mature integrated circuit process, make PCB-based SAW devices a promising platform for developing future lab-on-a-chip systems.</p
Mapping phyllosphere microbial communities of temperate European forests using environmental DNA and remote sensing
Get PDFIn this thesis, I set out to investigate the phyllosphere microbial communities of temperate European forests, focusing on both bacterial and fungal assemblages, their drivers, and their potential roles in forest ecosystem functioning. Drawing on environmental DNA (eDNA) metabarcoding, integrated with cutting-edge remote sensing tools, and coupled with measurements of host and environmental variables, , I aimed to bridge gaps in our understanding of microbial diversity patterns and their broader ecological significance. From the outset (Chapter 2), my research demonstrated that European forest canopies harbor diverse bacterial communities that are not only shaped by host species identity but also vary within the same host species and along gradients such as elevation. Demonstrating the close associations phyllosphere bacteria form with their host trees and the influence of the environment.Chapter 3 expanded this perspective to fungal communities (i.e. the mycobiome) in the canopy. I found that elevation was again a key driver of fungal composition and diversity in both beech and spruce stands. Moreover, stand-level canopy water content strongly influenced fungal communities, linking host physiological status to the distribution of fungal taxa. This establishes a link between canopy traits, water balance, and fungal community assembly. These findings parallel those in the bacterial domain but highlight some differences as well. While bacterial communities were strongly linked to host species identity, the fungal assemblages also showed strong links to environmental gradients like canopy water content, a common stress indicator.Chapter 4 focused on plant-pathogenic fungi, a subset of the canopy mycobiome with direct implications for forest health. Using host tree traits commonly used indicators of tree vitality (e.g., canopy water and chlorophyll content), I predicted patterns in these pathogenic communities. Lower canopy water content predicted higher percentages of plantpathogenic ASVs and, at times, reduced pathogen diversity—signaling that stressed trees may be more susceptible to a narrower but potentially more harmful group of pathogens. This work supports the notion that pathogen pressure is intimately connected to tree condition: stressed, water-limited trees might be easier targets for fungal pathogens. Such relationships have direct management implications. Monitoring canopy water content (via direct measurement or remote sensing) could provide early indicators of elevated pathogen risk, enabling forest managers to target interventions or conduct closer monitoring under specific stress conditions.In Chapter 5, I moved from local-scale analyses to landscape-level assessments. By integrating eDNA-inferred microbial data with satellite remote sensing (DESIS hyperspectral imagery), I showed that spatial patterns in relative abundance and diversity of canopy fungal plantpathogen communities could be modeled and mapped with high accuracy. Key vegetation indices (e.g., those reflecting canopy water, chlorophyll, and structure) emerged as effective predictors of pathogen abundance and diversity. This step was transformative: it enabled the creation of spatial redictions at 30-meter resolution, effectively scaling up local microbial surveys to entire landscapes.The successful coupling of molecular data with remote sensing highlights the feasibility of a new paradigm: using large-scale earth observation data to predict and monitor microbial community patterns. This approach not only addresses the Wallacean shortfall (gaps in species distribution knowledge) by extending biodiversity survey data into continuous spatial maps but also holds potential for the development of an early warning tool, laying the groundwork for the effective monitoring of environmental microbial communities
AI and the Disruption of Personhood
No full textThe new avatars and bots modeled after humans, the large language models (LLMs) with a “persona,” and the seemingly autonomously acting robots raise the question of whether AI technologies can also possess personhood or at least be part of our personhood. Do we extend our personhood through living or death bots in the digital realm? This article explores the application of the moral concept of personhood to AI technologies. It presents a twofold thesis: first, it illustrates, through various examples, how the concept of personhood is being disrupted in the context of AI technologies. Second, it discusses the potential evolution of the concept and argues for abandoning the personhood concept in AI ethics, based on reasons such as its vagueness, harmful and discriminatory character, and disconnection from society. Finally, the article outlines future perspectives for approaches moving forward, emphasizing the need for conceptual justice in moral concepts
Consensus-based development and practice testing of a generic quality indicator set for parenteral medication administration at home:A RAND appropriateness method study
Get PDFObjectives: Due to nursing shortages, an ageing population and increasing care demand, there is a growing interest in parenteral medication administration at home (PMAaH), comprising the administration of parenteral medication in the home situation of patients. The operational design of such PMAaH care pathways is complex, resulting in many variations of adoptions, showing a need for a quality framework. Although quality indicators (QIs) have been proposed to monitor the quality of specific care pathways, a generic quality framework for all types of PMAaH is lacking. Therefore, this study proposes a generic quality set for PMAaH, which includes structure and process QIs, to benchmark and redesign PMAaH care pathways to ensure high quality.Design: A generic QI set was developed for PMAaH using a systematic RAND appropriateness method adapted at the third phase. This method consisted of a scoping review to identify indicators, an expert panel rating phase including an online questionnaire and subsequent panel meeting to assess the appropriateness of the indicators and a retrospective practice testing to evaluate the feasibility, clarity and measurability of the indicators. After the practice testing, which consisted of an online questionnaire where experts could indicate the implementation state of all indicators in their hospital, a third expert panel adjusted the set to increase the likelihood of implementation in practice.Setting: The experts, all healthcare professionals involved in PMAaH processes, were recruited using the snowball sampling technique from three large Dutch, teaching hospitals. Subsequently, a practice testing by self-assessment was conducted in seven large Dutch teaching hospitals.Participants: 17 and seven healthcare professionals with diverse backgrounds participated in the online questionnaire and panel meeting, respectively.Results: The scoping review resulted in 36 QIs for PMAaH. After two expert panel rating rounds (online questionnaire and panel meeting), two indicators were removed: a QI related to travel distance policy since it was irrelevant and redundant, and a QI stating that a clinician should take the lead in a PMAaH-team, which was deemed too restrictive. After the practice testing, two QIs were removed: a QI related to clinical response documentation, which was unclear for the practice testing respondents and already covered by other QIs, and a QI related to survival documentation, which was deemed infeasible and undesirable to measure this differently than other patients by the third expert panel.The final set consists of 32 indicators (of which 15 were structure indicators and 17 were process indicators). The final set predominately includes QIs that are aimed at patient safety but also QIs focusing on the working conditions of the healthcare workers. 17.6% of the QIs are currently fully implemented in general in all seven hospitals. The practice testing revealed that operational QIs are more frequently implemented in practice than systemic QIs and that a structured quality assurance programme is needed in the hospitals.Conclusions: This study proposes a generic quality set for PMAaH that hospitals can use to redesign and benchmark PMAaH care pathways to assure high quality. The practice testing confirmed that there is a need for this structured quality set.</p
Implant survival at four years for hemiarthroplasty and total shoulder arthroplasty in the treatment of atraumatic avascular necrosis of the humeral head:a Dutch Arthroplasty Register study
Get PDFAims: Hemiarthroplasty (HA) and total shoulder arthroplasty (TSA) are often the preferred forms of treatment for patients with atraumatic avascular necrosis of the humeral head when conservative treatment fails. Little has been reported about the survival of HA and TSA for this indication. The aim of this study was to investigate the differences in revision rates between HA and TSA in these patients, to determine whether one of these implants has a superior survival and may be a better choice in the treatment of this condition.Methods: Data from 280 shoulders with 159 primary HAs and 121 TSAs, which were undertaken in patients with atraumatic avascular necrosis of the humeral head between January 2014 and January 2023 from the Dutch Arthroplasty Register (LROI), were included. Kaplan-Meier survival analysis and Cox regression analysis were undertaken.Results: Within four years of follow-up, a total of 15 revisions were required, involving seven HAs (4%) and eight TSAs (7%). This difference was not statistically significant (p = 0.523). Two HAs were revised because of progressive glenoid erosion, and three TSAs were revised for loosening of the glenoid component. The cumulative percentages of revision of HA and TSA were 6% and 8%, respectively (HR 1.1 (95% CI 0.5 to 2.7)).Conclusion: We found no significant difference in short- to mid-term implant survival between the use of a HA and a TSA in the treatment of atraumatic avascular necrosis of the humeral head, without significant glenoid wear
Extended MetaLIRS:Meta-learning for Imputation and Regression Selection Model with Explainability for Different Missing Data Mechanisms
No full textMissing data can cause problems for model decision-making procedures, which is a ubiquitous problem in data science. Imputation is a popular and agile method to handle missing data that enables predictive models to work smoothly. Given the plethora of available imputers and regressors, selecting the appropriate pair for a specific problem can be challenging. Many search-based approaches, such as AutoML tools, require significant processing power and focus on developing the predictive model itself, often bypassing the imputation processes with only basic steps. Models based on a meta-learning approach, combining data characteristics and algorithm performance, can help in imputer and method selection for missing data. Nevertheless, enhancing confidence in resulting models’ decisions and ensuring transparency through explainability is crucial. In this paper, we present a novel meta-learning-based framework for imputer/regressor selection; this framework is called MetaLIRS. With MetaLIRS, we built three recommendation models for MCAR, MAR, and MNAR data mechanisms, respectively, achieving an accuracy ranging from 63% to 67%. These models are designed to select the best-fitting imputer/regressor pair in terms of RMSE values of regressors. The models were learned using 26 datasets with six imputation methods, six regression methods, six missing levels (1%, 5%, 10%, 30%, 50%, and 70%), and 33 meta-features derived from missing data. A limitation of the approach is that due to a group of imputer/regressor pairs with very low occurrence, the so-called “other” class, the recommender occasionally predicts the unhelpful “other”. The main scientific contribution is the MetaLIRS framework which provides interpretability for its recommendations, is resource-efficient for recommendation model outputs, and enables rapid selection of the best imputer/regressor pair for a specific dataset without exhaustive brute-force searching.</p
Why don't you do what you said you would?:Conversational strategies for agents to understand users' reasons in supporting behavior
No full textEffective support from personal assistive technologies relies on accurate user models that capture user values, preferences, and context. Knowledge-based techniques model these relationships, enabling support agents to align their actions with user values. However, understanding values in a single context is insufficient due to the dynamic nature of behaviour. This study explores the use of dialogue strategies to update user models. Participants were randomly assigned to different strategies and they discussed one randomly chosen non-adherence situation with the agent. Then, their emotions, acquired information accuracy, completeness, and dialogue experience were rated. Our findings suggest that multiple-choice dialogues may limit response depth, reducing the perceived completeness of behaviour reasons. In contrast, open-ended questions allow more detailed input but require more time and effort, potentially worsening the dialogue experience. Through inductive coding, we identified key topics, such as individual challenges, priorities, tangible outcomes, and values, essential for constructing personalised user models. We also analyzed conversation paths to improve dialogue-based user model updates in support agents. Further research is needed to refine the relationship between dialogue strategies and self-conscious emotions, considering diverse backgrounds and health goals, while enhancing dialogue design.</p
Unified Feedback Linearization for Nonlinear Systems with Dexterous and Energy-Saving Modes
Get PDFSystems with a high number of inputs compared to the degrees of freedom (e.g. a mobile robot with Mecanum wheels) often have a minimal set of energy-efficient inputs needed to achieve a main task (e.g. position tracking) and a set of energy-intense inputs needed to achieve an additional auxiliary task (e.g. orientation tracking). This letter presents a unified control scheme, derived through feedback linearization, that can switch between two modes: an energy-saving mode, which tracks the main task using only the energy-efficient inputs while forcing the energy-intense inputs to zero, and a dexterous mode, which also uses the energy-intense inputs to track the auxiliary task as needed. The proposed control guarantees the exponential tracking of the main task and that the dynamics associated with the main task evolve independently of the a priori unknown switching signal. When the control is operating in dexterous mode, the exponential tracking of the auxiliary task is also guaranteed. Numerical simulations on an omnidirectional Mecanum wheel robot validate the effectiveness of the proposed approach and demonstrate the effect of the switching signal on the exponential tracking behavior of the main and auxiliary tasks
Agricultural data Privacy:Emerging platforms & strategies
Get PDFBackground: In today's world, grappling with the dual challenges of energy scarcity and climate change, the agricultural and food supply chains are at a crucial juncture for transformation. Data privacy within these sectors is increasingly significant, as the integration of advanced digital platforms becomes essential for improving efficiency and sustainability. Scope and Approach: This systematic literature review (SLR) explores the development and application of privacy-preserving data platforms specifically tailored to the agricultural and food supply chains. The review focuses on the evolving landscape of data architectures, data sovereignty, and advanced privacy-preserving techniques. Techniques such as anonymization, encryption, differential privacy, and federated learning are examined, along with the legal and ethical considerations surrounding data sharing in the context of global energy and climate-related challenges. Key Findings and Conclusions: The review synthesizes findings from a broad spectrum of studies published over the last decade, uncovering significant advancements in privacy-preserving technologies which may show their benefit for the agricultural and food sectors. It identifies dominant research trends and promising future directions for enhancing data security and sustainability. The key findings underscore the vital importance of safeguarding data privacy in these sectors, highlighting the potential of advanced privacy techniques to protect sensitive agricultural data while balancing the need for transparency and operational effectiveness. The review concludes that adopting these innovative privacy methods is crucial for fostering a more sustainable and secure future for agriculture and food systems, contributing to the global Sustainable Development Goals such as SDG 2 (Zero Hunger) and SDG 12 (Responsible Consumption and Production) by improving precision agriculture, optimizing yields (crop and animal) and reducing food waste.</p