Hasil untuk "artificial intelligence"

Seungnam Yu, Geegum Lee, Jeongmok Kim

This study presents a VR-based teleoperation framework enhancing collaborative robot stability and manipulability via hand-tracking, adaptive control, and dual-modality haptic feedback. It addresses critical synchronization challenges (singularity avoidance, tracking responsiveness, and workspace constraints), which are especially problematic in first-person VR where kinematic limits are not directly perceivable. The framework employs Adaptive Damped Least Squares (A-DLS) to maintain manipulability near singular configurations, workspace impedance control to enforce boundary constraints, and vibrotactile feedback delivered through a haptic glove to convey both workspace limits (fingertip vibration) and path deviation information (wrist vibration) to operators. Key features include real-time hand-tracking, workspace calibration, and adaptive controls to ensure seamless coordination between virtual and real robot workspaces. Experimental validation through two complementary studies demonstrates the system’s effectiveness. Experiment 1 evaluated singularity management and workspace stability, showing that the A-DLS algorithm maintained manipulability above critical thresholds for 92% of operational time versus 78% without adaptive damping. Experiment 2 assessed trajectory tracking accuracy through a path-following task with 10 participants. Results demonstrate that haptic-enabled control achieves a 24.7% reduction in mean path-following error (from 10.03 mm to 7.55 mm, p = 0.001) compared to haptic-disabled conditions, indicating improvements in both accuracy and consistency. Although haptic guidance modestly increases task time due to higher precision focus, the resulting gains in accuracy and stability make this framework ideal for precision-critical tasks. By ensuring stability near workspace boundaries, the system effectively facilitates VR-based teleoperation for applications like painting, polishing, and contour-following.

Caroline M Godfrey, Ashley A Leech, Kevin C McGann et al.

<h4>Background</h4>Artificial intelligence-based radiomic approaches have been shown to accurately evaluate indeterminate pulmonary nodules. With the expansion of lung cancer screening and utilization of computed tomography imaging, indeterminate pulmonary nodules requiring diagnostic evaluation are increasingly common. Accurate non-invasive characterization may reduce time to cancer diagnosis and decrease invasive procedures for benign disease, but the cost-effectiveness of AI-based methods has not been quantified. We sought to evaluate the cost-effectiveness of AI-assisted clinician evaluation compared to clinician evaluation alone for the cancer risk stratification of patients with indeterminate pulmonary nodules.<h4>Methods</h4>We constructed a decision model assuming guideline-based care from a payer perspective with a lifetime horizon. The base case is a 1.1 cm incidentally discovered IPN in a 60-year-old operative candidate in a clinical population with a 65% malignancy prevalence. Cost per life-year gained (LYG) was the primary outcome. We conducted deterministic sensitivity analyses on all parameters and performed a probabilistic sensitivity analysis. Given clinical variability of malignancy prevalence, we assessed the malignancy prevalence threshold at which utilization of AI would be cost-effective.<h4>Results</h4>AI-supported clinician risk stratification resulted in an increase of 0.03 life years compared to clinician alone. With a 65% malignancy prevalence, AI was cost-effective with an incremental cost-effectiveness ratio (ICER) of $4,485/LYG. When the malignancy prevalence was < 5%, the ICER for AI support exceeded a standard willingness-to-pay threshold of $100,000/LYG.<h4>Conclusions</h4>In clinical settings with a pre-test probability of malignancy exceeding 5%, AI-supported IPN risk stratification is cost-effective compared to clinician assessment alone.

Carina Soler Pons, Ana de Marco García, Ricard Martínez et al.

The fragmentation and decentralization of medical data, including radiological imaging, continue to challenge large-scale observational research across Europe. Artificial Intelligence (AI) applied to big datasets is transforming diagnosis and treatments towards precision medicine across many diseases, yet the lack of findable, accessible, and interoperable datasets still limits model development, validation, and final clinical translation. The European Federation for Cancer Images (EUCAIM) project was launched in 2023 to address these challenges by establishing a secure centralized and federated infrastructure for the secondary use of large-scale oncological imaging and related clinical data. By consolidating fragmented datasets, EUCAIM lays the groundwork for harmonized data governance and trusted cross-border sharing. Implementing a robust documentation framework is essential to ensure regulatory compliance, safeguard data integrity, and support secure data flows across institutional and national boundaries, fully aligned with European regulations and ethical standards. EUCAIM builds on the AI for Health Imaging (AI4HI) initiative (Predictive In-silico Multiscale Analytics to support cancer personalized diagnosis and prognosis, empowered by imaging biomarkers - PRIMAGE, Accelerating the lab to market transition of AI tools for cancer management - CHAIMELEON, Novel pan-European imaging platform for artificial intelligence advances in oncology - EuCanImage, An AI Platform integrating imaging data and models, supporting precision care through prostate cancer’s continuum - ProCancer-I, A multimodal AI-based toolbox and an interoperable health imaging repository for the empowerment of imaging analysis related to the diagnosis, prediction and follow-up of cancer - INCISIVE and integrates over 94 partners and more than 180 stakeholders spanning medical imaging, high performance computing, data standardization, innovation, and legal compliance. This large collaborative ecosystem reinforces EUCAIM’s role as a reference for General Data Protection Regulation (GDPR) and European Health Data Space Regulation (EHDSR) adherence. This publication presents the real-world experience of integrating imaging and clinical data from a reference university hospital into the EUCAIM infrastructure. It outlines the procedural, ethical, and legal challenges encountered, and details the strategies implemented to ensure compliance with data protection regulations, including privacy, security, and ethical standards. These insights offer a practical framework for future large-scale oncological imaging datasets harmonization and AI development, contributing to scalable, reproducible, and legally compliant research that strengthens Europe’s capacity for trustworthy AI-driven oncology solutions.

Bianca Pilla, Jennifer Stone, Zoe Jordan

Abstract Background Understanding the alignment and contributions of research to the United Nations Sustainable Development Goals (UN SDGs) is essential for guiding global progress toward these critical targets. Several SDG mapping approaches have been proposed and tested by organisations and researchers but have not produced consistent results. With its capacity to analyse vast datasets and identify patterns, AI-powered search functionality has been presented as an innovative mechanism for tracking, analysing and reporting SDG research to assess progress towards targets and facilitate evidence-based decisions. This study aimed to assess the reliability of automated mapping approaches utilised by online research databases in mapping published evidence syntheses to the UN SDG-3, Good Health & Wellbeing. Methods This study mapped systematic and scoping reviews published in JBI Evidence Synthesis to SDG- 3, Good Health and Wellbeing. Four unique raters independently assessed 204 evidence syntheses based on relevance to SDG-3. These four raters included AI in three established databases and a manual ‘human’ assessment. Inter-rater reliability was assessed using Light’s Kappa. Results Concurrence occurred for 52% of publications. Inter-rater reliability indicated ‘minimal agreement’ among the four raters in mapping the 204 evidence syntheses to SDG-3. Discrepancies in the publications mapped to SDG-3 across the four raters may be explained by the different taxonomies used by the databases, different machine learning algorithms, and constantly evolving search strategies. Our results indicate significant room for improvement in achieving greater consensus among approaches, including AI algorithms designed to identify such publications. Conclusion The findings of this study point to the need for SDG mapping tools that are practical and effective, as the 2030 agenda deadline nears and progress on numerous targets lags. Identifying evidence across the global ecosystem is critical, but AI’s reliability to contribute to this has yet to be established with confidence. It signals to the scientific community, policymakers, funders and others the importance of critically reflecting on how research is captured and aligned to the SDGs.

Mecit Kantarcı, Volkan Kızılgöz, Ramazan Terzi et al.

PURPOSE: This study aimed to evaluate the effectiveness of artificial intelligence (AI) in diagnosing focal nodular hyperplasia (FNH) of the liver using magnetic resonance imaging (MRI) and compare its performance with that of radiologists. METHODS: In the first phase of the study, the MRIs of 60 patients (30 patients with FNH and 30 patients with no lesions or lesions other than FNH) were processed using a segmentation program and introduced to an AI model. After the learning process, the MRIs of 42 different patients that the AI model had no experience with were introduced to the system. In addition, a radiology resident and a radiology specialist evaluated patients with the same MR sequences. The sensitivity and specificity values were obtained from all three reviews. RESULTS: The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the AI model were found to be 0.769, 0.966, 0.909, and 0.903, respectively. The sensitivity and specificity values were higher than those of the radiology resident and lower than those of the radiology specialist. The results of the specialist versus the AI model revealed a good agreement level, with a kappa (κ) value of 0.777. CONCLUSION: For the diagnosis of FNH, the sensitivity, specificity, PPV, and NPV of the AI device were higher than those of the radiology resident and lower than those of the radiology specialist. With additional studies focused on different specific lesions of the liver, AI models are expected to be able to diagnose each liver lesion with high accuracy in the future. CLINICAL SIGNIFICANCE: AI is studied to provide assisted or automated interpretation of radiological images with an accurate and reproducible imaging diagnosis.

Sofia Morandini, Federico Fraboni, Mark Hall et al.

The integration of AI technologies in aerospace manufacturing is significantly transforming critical operational processes, impacting decision-making, efficiency, and workflow optimization. Explainability in AI systems is essential to ensure these technologies are understandable, trustworthy, and effectively support end-users in complex environments. This study investigates the factors influencing the explainability of AI-based Decision Support Systems in aerospace manufacturing from the end-users' perspective. The study employed a Closed Card Sorting technique involving 15 professionals from a leading aerospace organization. Participants categorized 15 AI features into groups—enhances, is neutral to, and hinders explainability. Qualitative feedback was collected to understand participants' reasoning and preferences. The findings highlighted the importance of user support features in enhancing explainability, such as system feedback on user inputs and error messages with guidance. In contrast, technical jargon was consistently perceived as a hindrance. Transparency of algorithms emerged as the highest-priority feature, followed by clarity of interface design and decision rationale documentation. Qualitative insights emphasized the need for clear communication, intuitive interfaces, and features that reduce cognitive load. The study provides actionable insights for designing AI-based DSSs tailored to the needs of aerospace professionals. By prioritizing transparency, user support, and intuitive design, designers and developers can enhance system explainability and foster user trust. These findings support the human-centric development of AI technologies and lay the groundwork for future research exploring user-centered approaches in different high-stakes industrial contexts.

Dawn Yi Xin Lee, Chun En Yau, Maeve Pin Pin Pek et al.

Background: Socioeconomic status (SES) is a well-established determinant of cardiovascular health. However, the relationship between SES and clinical outcomes in long-term out-of-hospital cardiac arrest (OHCA) is less well-understood. The Singapore Housing Index (SHI) is a validated building-level SES indicator. We investigated whether SES as measured by SHI is associated with long-term OHCA survival in Singapore. Methods: We conducted an open cohort study with linked data from the Singapore Pan-Asian Resuscitation Outcomes Study (PAROS), and the Singapore Registry of Births and Deaths (SRBD) from 2010 to 2020. We fitted generalized structural equation models, calculating hazard ratios (HRs) using a Weibull model. We constructed Kaplan–Meier survival curves and calculated the predicted marginal probability for each SHI category. Results: We included 659 cases. In both univariable and multivariable analyses, SHI did not have a significant association with survival. Indirect pathways of SHI mediated through covariates such as Emergency Medical Services (EMS) response time (HR of low-medium, high-medium and high SHI when compared to low SHI: 0.98 (0.88–1.10), 1.01 (0.93–1.11), 1.02 (0.93–1.12) respectively), and age of arrest (HR of low-medium, high-medium and high SHI when compared to low SHI: 1.02 (0.75–1.38), 1.08 (0.84–1.38), 1.18 (0.91–1.54) respectively) had no significant association with OHCA survival. There was no clear trend in the predicted marginal probability of survival among the different SHI categories. Conclusions: We did not find a significant association between SES and OHCA survival outcomes in residential areas in Singapore. Among other reasons, this could be due to affordable healthcare across different socioeconomic classes.

Yasunari Miyagi, Toshihiro Habara, Rei Hirata et al.

Abstract Purpose To investigate the usefulness of an original dual artificial intelligence (AI) system, in which the first AI system eliminates the background of sliced tomographic blastocyst images, then the second AI system predicts implantation success using three‐dimensional (3D) reconstructed images of the sequential images and conventional embryo evaluation parameters (CEE) such as maternal age. Methods Patients (from June 2022 to July 2023) could opt out and there was additional information on the Web site of the clinic. Implantation and non‐implantation cases numbered 458 and 519, respectively. A total of 10 747 tomographic images of the blastocyst in a time‐lapse incubator system with the CEE were obtained. Results The statistic values by the dual AI system were 0.774 ± 0.033 (mean ± standard error) for area under the characteristic curve, 0.727 for sensitivity, 0.719 for specificity, 0.727 for predictive value of positive test, 0.719 predictive value of negative test, and 0.723 for accuracy, respectively. Conclusions The usefulness of the dual AI system in predicting implantation of blastocyst in handling 3D data with conventional embryo evaluation information was demonstrated. This system may be a feasible option in clinical practice.

Marcelino Campos, Juan Carlos Galán, Mario Rodríguez-Domínguez et al.

ABSTRACTThe epidemiology of sexually transmitted infections (STIs) is complex due to the coexistence of various pathogens, the variety of transmission modes derived from sexual orientations and behaviors at different ages and genders, and sexual contact hotspots resulting in network transmission. There is also a growing proportion of recreational drug users engaged in high-risk sexual activities, as well as pharmacological self-protection routines fostering non-condom practices. The frequency of asymptomatic patients makes it difficult to develop a comprehensive approach to STI epidemiology. Modeling approaches are required to deal with such complexity. Membrane computing is a natural computing methodology for the virtual reproduction of epidemics under the influence of deterministic and stochastic events with an unprecedented level of granularity. The application of the LOIMOS program to STI epidemiology illustrates the possibility of using it to shape appropriate interventions. Under the conditions of our basic landscape, including sexual hotspots of individuals with various risk behaviors, an increase in condom use reduces STIs in a larger proportion of heterosexuals than in same-gender sexual contacts and is much more efficient for reducing Neisseria gonorrhoeae than Chlamydia and lymphogranuloma venereum infections. Amelioration from diagnostic STI screening could be instrumental in reducing N. gonorrhoeae infections, particularly in men having sex with men (MSM), and Chlamydia trachomatis infections in the heterosexual population; however, screening was less effective in decreasing lymphogranuloma venereum infections in MSM. The influence of STI epidemiology of sexual contacts between different age groups (<35 and ≥35 years) and in bisexual populations was also submitted for simulation.IMPORTANCEThe epidemiology of sexually transmitted infections (STIs) is complex and significantly influences sexual and reproductive health worldwide. Gender, age, sexual orientation, sexual behavior (including recreational drug use and physical and pharmacological protection practices), the structure of sexual contact networks, and the limited application or efficiency of diagnostic screening procedures create variable landscapes in different countries. Modeling techniques are required to deal with such complexity. We propose the use of a simulation technology based on membrane computing, mimicking in silico STI epidemics under various local conditions with an unprecedented level of detail. This approach allows us to evaluate the relative weight of the various epidemic drivers in various populations at risk and the possible outcomes of interventions in particular epidemiological landscapes.

Mikkel Andreas Kvande, Sigurd Løite Jacobsen, Morten Goodwin et al.

Agricultural development is one of the most essential needs worldwide. In Norway, the primary foundation of grain production is based on geological and biological features. Existing research is limited to regional-scale yield predictions using artificial intelligence (AI) models, which provide a holistic overview of crop growth. In this paper, the authors propose detecting several field-scale crop types and use this analysis to predict yield production early in the growing season. In this study, the authors utilise a multi-temporal satellite image, meteorological, geographical, and grain production data corpus. The authors extract relevant vegetation indices from satellite images. Furthermore, the authors use field-area-specific features to build a field-based crop type classification model. The proposed model, consisting of a time-distributed network and a gated recurrent unit, can efficiently classify crop types with an accuracy of 70%. In addition, the authors justified that the attention-based multiple-instance learning models could learn semi-labelled agricultural data, and thus, allow realistic early in-season predictions for farmers.

Fation Fera, Christos Spandonidis

This research focuses on enhancing the preventive maintenance strategies currently employed for induction motors within ship propulsion systems, advocating for a shift towards a predictive maintenance model. It introduces a real-time monitoring framework that continuously observes the induction motor, providing essential support to maintenance personnel. The motor operates under a range of environmental and operational conditions, including temperature fluctuations, rotational speeds, and mechanical loads. These variations can obscure the current time series data, potentially masking signs of actual damage and hindering effective damage detection. To tackle this issue, the proposed framework utilizes artificial intelligence (AI) technology, specifically the well-established autoencoder, in conjunction with the Mahalanobis statistical distance. This approach accounts for the diverse operating conditions during the training phase, allowing it to model complex, non-linear relationships and effectively differentiate between normal and anomalous states. The framework is integrated into a decision support platform designed for real-time operations in maritime settings, offering a sophisticated system architecture that aims to align advanced damage detection methodologies with the maritime industry’s need for real-time, user-friendly solutions.

M. S. Ganachari

Rohan Reddy Kalavakonda, Junjun Huan, Peyman Dehghanzadeh et al.

This paper introduces Fusion Intelligence (FI), a bio-inspired intelligent system, where the innate sensing, intelligence and unique actuation abilities of biological organisms such as bees and ants are integrated with the computational power of Artificial Intelligence (AI). This interdisciplinary field seeks to create systems that are not only smart but also adaptive and responsive in ways that mimic the nature. As FI evolves, it holds the promise of revolutionizing the way we approach complex problems, leveraging the best of both biological and digital worlds to create solutions that are more effective, sustainable, and harmonious with the environment. We demonstrate FI's potential to enhance agricultural IoT system performance through a simulated case study on improving insect pollination efficacy (entomophily).

Leon Witt, Armando Teles Fortes, Kentaroh Toyoda et al.

Blockchain technology and Artificial Intelligence (AI) have emerged as transformative forces in their respective domains. This paper explores synergies and challenges between these two technologies. Our research analyses the biggest projects combining blockchain and AI, based on market capitalization, and derives a novel framework to categorize contemporary and future use cases. Despite the theoretical compatibility, current real-world applications combining blockchain and AI remain in their infancy.

Gustavo A. Mesías-Ruiz, Gustavo A. Mesías-Ruiz, María Pérez-Ortiz et al.

Crop protection is a key activity for the sustainability and feasibility of agriculture in a current context of climate change, which is causing the destabilization of agricultural practices and an increase in the incidence of current or invasive pests, and a growing world population that requires guaranteeing the food supply chain and ensuring food security. In view of these events, this article provides a contextual review in six sections on the role of artificial intelligence (AI), machine learning (ML) and other emerging technologies to solve current and future challenges of crop protection. Over time, crop protection has progressed from a primitive agriculture 1.0 (Ag1.0) through various technological developments to reach a level of maturity closelyin line with Ag5.0 (section 1), which is characterized by successfully leveraging ML capacity and modern agricultural devices and machines that perceive, analyze and actuate following the main stages of precision crop protection (section 2). Section 3 presents a taxonomy of ML algorithms that support the development and implementation of precision crop protection, while section 4 analyses the scientific impact of ML on the basis of an extensive bibliometric study of &gt;120 algorithms, outlining the most widely used ML and deep learning (DL) techniques currently applied in relevant case studies on the detection and control of crop diseases, weeds and plagues. Section 5 describes 39 emerging technologies in the fields of smart sensors and other advanced hardware devices, telecommunications, proximal and remote sensing, and AI-based robotics that will foreseeably lead the next generation of perception-based, decision-making and actuation systems for digitized, smart and real-time crop protection in a realistic Ag5.0. Finally, section 6 highlights the main conclusions and final remarks.

Apoorv Singh

While advancing rapidly, Artificial Intelligence still falls short of human intelligence in several key aspects due to inherent limitations in current AI technologies and our understanding of cognition. Humans have an innate ability to understand context, nuances, and subtle cues in communication, which allows us to comprehend jokes, sarcasm, and metaphors. Machines struggle to interpret such contextual information accurately. Humans possess a vast repository of common-sense knowledge that helps us make logical inferences and predictions about the world. Machines lack this innate understanding and often struggle with making sense of situations that humans find trivial. In this article, we review the prospective Machine Intelligence candidates, a review from Prof. Yann LeCun, and other work that can help close this gap between human and machine intelligence. Specifically, we talk about what's lacking with the current AI techniques such as supervised learning, reinforcement learning, self-supervised learning, etc. Then we show how Hierarchical planning-based approaches can help us close that gap and deep-dive into energy-based, latent-variable methods and Joint embedding predictive architecture methods.