Articles tagged with Artificial Intelligence
A new study investigated people's moral judgments on euthanasia decisions made by AI and robots compared to human doctors. The research found that people generally prefer decisions made by humans over AI in end-of-life care situations.
Aerial robots are limited to manipulating rigid objects, but Lehigh University researcher David Saldaña aims to expand their capabilities with an adaptive controller and reinforcement learning. His research has potential applications in construction, disaster response, and industrial automation.
Researchers from The University of Osaka have demonstrated that vision transformers can spontaneously develop human-like visual attention patterns without specific training. This breakthrough showcases the potential of self-supervised learning for advancing AI applications and modeling biological vision.
The ERQA framework uses a semantic vector database and curated literature repository to improve medical knowledge discovery, achieving high retrieval accuracy in literature-based QA and abstract summarization. The researchers plan to further improve the model by incorporating larger-scale biomedical literature datasets.
The Stuttgart Cluster of Excellence IntCDC will continue to develop innovative construction methods that reduce resource consumption and CO2 emissions through computational design and engineering methods. The cluster aims to evolve its Co-Design approach into a Co-Agency approach, focusing on bio-based building materials, AI technologi...
A team of researchers from UNIGE and UniBE used generative AI to assess emotional intelligence, finding that AIs achieved significantly higher scores than humans. The results also showed that AIs can generate new emotional intelligence tests in record time.
Aurora, a state-of-the-art machine learning model, delivers high-quality forecasts for key environmental systems like air quality, weather, and tropical cyclones. The model's flexibility enables it to be used for various future scenarios, including flood risks, wildfire spread, and renewable energy output.
UT-Battelle has donated $225,000 to purchase advanced manufacturing equipment for Oak Ridge High School's Wildcat Manufacturing iSchool program. The equipment includes AI- and robotics-assisted manufacturing systems, enabling students to earn college credit while gaining experience with the latest innovations in intelligent manufacturing.
Researchers at UCSF have successfully engineered a shapeshifting protein that can change shape in response to signals, potentially leading to breakthroughs in medicine, agriculture, and environmental applications. This achievement marks the first step towards creating stable yet dynamic proteins using AI-augmented protein engineering.
Researchers at UC San Diego develop new approach using artificial intelligence to predict therapy outcomes and match patients with right treatments, improving cancer care and patient outcomes.
OneCareAI applies AI and supercomputing to detect early stroke risk from ECG data obtained with smartwatches, offering a non-invasive solution for personalized risk assessments. The technology has the potential to be extended to other cardiovascular diseases, opening opportunities for scalability.
University of Missouri researchers create digital sentiment map using AI to analyze public Instagram posts, linking emotional tone to real-life features. The tool aims to improve city services, identify areas of concern, and inform emergency response decisions.
While digital and artificial intelligence-based pathology (DAIP) offers improved diagnostic accuracy and equity, its implementation is hindered by financial constraints, technical challenges, and legal/ethical concerns. A balanced assessment of the benefits and pitfalls must be conducted before adopting DAIP.
Triboelectric and piezoelectric nanogenerators convert mechanical energy into electrical energy, enhancing robotic autonomy and efficiency. The technology has the potential to reshape future robotic capabilities, particularly in industrial automation, healthcare, and smart home applications.
Derek Leben's book 'AI Fairness' offers a philosophical framework to evaluate and mitigate biases in AI algorithms. The author argues that principles like autonomy, equal treatment, and equal impact should guide the design of fair AI systems.
The ENABLE-6G project has developed 20 technological breakthroughs that prioritize network intelligence, energy efficiency, localization, and privacy-preserving analytics. The project has generated a range of outputs, including real-world testing platforms, wireless sensing algorithms, and AI-driven network optimization methods.
The application of AI in public health informatics offers potential to revolutionize data collection, disease surveillance, decision-making, and interventions. However, its implementation faces technical, ethical, and operational challenges related to health equity, privacy, security, and bias.
DeepSeek-R1 has been shown to enhance clinical workflows, supporting diagnostic reasoning, treatment planning, and risk assessment. The model has also demonstrated promise in patient engagement and medical education, improving patient adherence and creating interactive educational cases.
Researchers found that artificial intelligence tools can accurately predict disease for patients with typical symptoms but struggle with those exhibiting atypical symptoms. Human oversight is necessary for high-quality patient-centered care when using AI as an assistive tool.
A new study reveals that fine-tuned large language models (LLMs) greatly improve error detection in radiology reports, outperforming other NLP tools. The LLMs successfully identify various types of errors, including transcription and directional mistakes, providing a promising solution for medical proofreading.
The article highlights key pitfalls in evaluating AI algorithmic biases in radiology, such as the lack of representation in medical imaging datasets and inconsistent demographic definitions. To address these issues, radiologists recommend collecting and reporting more demographic variables and establishing standard notions of bias eval...
The National Heart Centre Singapore is launching a nationwide AI initiative to reduce hours of specialist analysis for cardiac scans from minutes, aiding early detection and prediction of coronary artery disease. The SENSE project will apply advanced machine learning technology to interpret cardiac imaging scans.
Researchers at KAIST have developed an AI-powered app called AACessTalk that enables meaningful communication between children with autism and their parents. The app uses personalized vocabulary cards and customized prompts to foster mutual engagement in dialogue.
Rice University will offer a Bachelor of Science in artificial intelligence beginning in the fall of 2025. The new major provides a cohesive program covering various approaches to AI and incorporates disciplines outside of engineering.
Neurosymbolic AI proposes an alternative model that requires less computing power, allowing smaller players to enter the field. This approach enables efficient and trustworthy AI systems without unsustainable energy use or gatekeeping.
A survey of 3,000 US residents found that Americans prefer domain experts over non-experts or algorithms for content moderation. Small juries of independent experts were viewed as the most legitimate, while large, diverse juries with minimal qualifications were also seen as acceptable.
A new study by the University of South Australia found that using AI alone in hiring decisions does not necessarily lead to more diverse and inclusive outcomes. The research highlights the importance of organizations having clear diversity guidelines, qualitative goals, and human oversight to support the use of AI tools effectively.
A deep learning model called Sybil can predict future lung cancer risk from a single low-dose chest CT scan. The model demonstrated good performance in predicting cancer diagnosis at both one and six years, including in never-smokers.
The AI system uses evolutionary computing to find optimal policy solutions that balance trade-offs between carbon storage, economic disruptions, and environmental benefits. It recommends nuanced approaches to land-use strategies, such as replacing crop land with forest, rather than simply converting all land into forests.
Training neural networks on simple cognitive tasks improves their ability to handle more complex ones. By adopting the principles of early childhood education, researchers found that recurrent neural networks can be trained faster and with better results.
A team of scientists simulated the movement of microorganisms in liquids without a central control system. They found that simple rules and decentralized control can lead to efficient swimming behavior, potentially enabling nanobots to transport drugs or perform other complex tasks.
Researchers at Duke University developed a novel framework called WildFusion that fuses vision, vibration and touch to enable robots to sense complex outdoor environments like humans do. The system was tested in real-world settings and showed remarkable ability to accurately predict traversability and improve robot decision-making.
A new wearable sensor has been shown to effectively monitor obstructive sleep apnea (OSA) treatment response, offering a bridge between consumer wearables and medical devices. The device provides continuous data collection and connects to an app for patients and healthcare providers to share information and make informed decisions.
A team of San Diego researchers is developing an innovative AI tutor system to provide personalized support to students. The system, based on large language models, will be deployed in eight foundational courses at higher education institutions in San Diego County and assessed for its impact on learning.
Researchers created over 3,800 anatomically accurate digital hearts to investigate how age, sex, and lifestyle affect heart disease. They found that age and obesity cause changes in the heart's electrical properties, which could explain why these factors are linked to a higher risk of heart disease.
Researchers at Virginia Tech have designed a new metallic material alloy with superior mechanical properties, leveraging data-driven frameworks and explainable AI. This breakthrough accelerates the discovery of advanced metallic alloys, offering insights into materials' structure-property relationships.
Researchers found that vision-language models fail to understand negation words like 'no' and 'not', which can cause incorrect diagnoses. This limitation affects multiple choice question answering with negated captions, with accuracy dropping below random chance.
New York University researchers developed a novel process using machine learning to reveal groups of genes governing nitrogen use efficiency in plants like corn. The study aims to help farmers improve crop yields and minimize fertilizer costs.
The new Jean Zay supercomputer in France has increased its AI dedicated resources fourfold to 125.9 petaflops, offering free access for thousands of research projects and start-ups. The upgraded machine is expected to accelerate French and European power in artificial intelligence.
Researchers propose Input-Driven Plasticity model, which integrates past and new information to guide memory retrieval. The model is robust to noise and uses it as a means to filter out less stable memories.
A study suggests that groups of artificial intelligence language models can self-organise into societies, reaching consensus on linguistic norms, and are prone to tipping points in social convention. Collective biases emerge between agents through interactions, a blind spot in most current AI safety work.
A new AI-driven tool can forecast acute child malnutrition in Kenya up to six months in advance with high accuracy, enabling timely interventions. The model integrates clinical data and satellite information to identify emerging risk areas, providing a game-changing solution to address public health emergency in the country.
A new special report warns of LLM cybersecurity threats in radiology, highlighting the importance of implementing security measures to prevent malicious use. The report recommends robust security measures and best practices for safe integration of LLMs into healthcare.
UCF students created an AI system, AIMS, to track surgical staples during robotic surgeries, reducing waste and improving sustainability. The technology was developed as part of the college's Senior Design capstone course and has been successfully tested in operating rooms at Orlando Health.
A new deep learning model, ENDNet, significantly enhances subgraph matching accuracy by identifying and neutralizing extra nodes that interfere with the matching process. This improves performance in pattern recognition tasks across various fields, including drug discovery and natural language processing.
Researchers, Mollie Brewer and Kevin Childs, are exploring how coaches use data and technology to maximize player performance and safety. They found that coaches analyze data from wearable sensors to determine when players need rest or additional training, and that AI-powered analytics can improve team strategy.
A new study uses explainable machine learning to predict metastasis and mortality in children with medulloblastoma, identifying key immune factors driving risk. The model provides a transparent approach to prognostic accuracy and supports personalized clinical decision-making.
JMIR Mental Health is seeking rigorous research on AI-powered therapy bots and virtual companions for mental health. The journal aims to explore their effectiveness, limitations, and place within broader care models.
A new AI model can predict protein location in human cells with high accuracy, enabling faster diagnosis of diseases like Alzheimer's and cancer. By combining protein sequence analysis with computer vision, the model can pinpoint proteins' locations at the single-cell level.
A study analyzing clinical notes found stigmatizing language was more common in Black patients' charts, with 54.9% experiencing bias. Hispanic and Asian/Pacific Islander patients also showed disparities in documentation, highlighting the need for culturally sensitive practices in perinatal care.
New AI tools are preserving marginalized cultures by recording endangered languages, creating interactive learning tools, and preserving oral histories. The authors highlight the importance of culturally aware AI systems that can adapt to regional traditions and perspectives.
A new AI-based method accurately sorts cancer patients into groups with similar characteristics before treatment and similar outcomes after treatment. The approach showed better performance than standard statistical and machine learning methods in predicting treatment outcomes from health record data.
A study found that large language models generalize language patterns in a surprisingly human-like way through analogy, relying on stored examples and drawing analogies when dealing with unfamiliar words. The models behave as if they have formed a memory trace from every individual example of every word encountered during training.
The Human Exposome is a global scientific effort to understand the environmental factors that underpin disease and health. The Exposome Moonshot Forum aims to chart this exposome, providing usable metrics and data points for targeted public health interventions.
Researchers developed an AI system called Image2Reg that can identify genetic perturbations in cell images, potentially leading to new drugs. The system uses machine learning and molecular networks to analyze patterns in chromatin structure, revealing links between genes and their functions.
A new generative AI model, DiffSMol, has been developed to generate realistic 3D structures of small molecules with promising drug properties. The study achieved a 61.4% success rate in creating novel molecules, outperforming prior research attempts.
A Carnegie Mellon University researcher critiques an article on large language models' ability to analyze sensitive discourse, including hate speech. The commentary highlights the need for a nuanced approach to human-AI collaboration in critical discourse analysis.
A new AI model, EVE-Vax, uses evolutionary data to predict viral proteins and design future-proof vaccines against rapidly mutating viruses like SARS-CoV-2. By analyzing millions of years of genetic information, the model can create brand-new proteins with the same immune response as real viruses.
The TUCAN6-CM project aims to transform next-generation mobile networks through efficient integration of communication, computing, and sensing technologies. Advanced solutions will enhance quality of life, sustainability, and safety in smart cities and industrial environments.
Researchers at Pohang University of Science & Technology (POSTECH) have developed an achromatic metagrating that handles all colors in a single glass layer, eliminating the need for multiple layers. This breakthrough enables vivid full-color images using a 500-µm-thick single-layer waveguide.