Research applying artificial intelligence across science and engineering. Covers machine learning, sensing and quality control, immune and health prediction, and 5G networks.
A novel framework uses AI and social media to identify promising brand collaborations, providing practical guidance for brand managers. The framework has been tested and validated, showing a nearly 90% chance of correctly identifying favorable brand pairs.
A quality improvement study found that AI medical scribes significantly reduced clinician documentation time, with the greatest benefits seen among frequent users. However, the study also found that the tool did not improve same-day note completion or reduce after-hours documentation, limiting its productivity benefits.
Christina Theodoris is recognized for developing novel AI models to revolutionize human cell interrogation and predict gene network biology. Her work has immediate applications for human health, including identifying promising new drug targets for heart disease.
SKKU researchers developed two AI technologies to improve complex tasks and safety, reducing computational costs and dependency on external feedback. The selective-critique framework SAG and reinforcement learning technology VAN-Flow enabled AI systems to accumulate successful experiences and avoid mistaken strategies.
A Rutgers study found that 1 in 3 young adults (18-29 years) use AI for substance-use information, primarily seeking risks and consequences. The study highlights the need for research on how young adults interpret and understand AI outputs on substance use, with implications for public health.
JMIR Publications releases five news and perspectives on digital health, covering AI-driven CPR training, GLP-1 ads' impact on body image, and innovative solutions for postpartum psychosis detection, aiming to improve patient care and awareness of emerging health issues.
The tools, Hallucination Nerd and Protocol Nerd, use AI to identify supported claims in academic citations and suggest relevant experimental protocols for scientific studies. They were validated using research databases and showed promising results in detecting citation hallucinations and identifying effective protocols.
A new AI tool, circMAC, predicts microRNA binding sites on circular RNAs at nucleotide resolution, providing more focused candidate regions for experimental validation. The tool outperforms existing RNA language models and demonstrates its potential in supporting precise studies of RNA-mediated regulation.
The project aims to develop shared cyberinfrastructure that enables realistic, reproducible experiments for AI-enabled robots and cyber-physical systems. The framework will provide research-grade workflows, digital twin environments, and experimental examples to support hands-on training and mentored research.
The study analyzed 7,115 news stories, 13,787 journalistic sources, and 11,266 photographs, finding a 41.60% focus on men, 39.06% on mixed topics, and only 5.33% on women. The report also examines sexist language, biases, and representation in AI media, proposing recommendations for journalists to promote more egalitarian coverage.
Researchers found that IT vulnerabilities in automotive software have shifted to apps and cloud services, posing risks to driver data and car control. Experts advise car manufacturers to secure systems and encourage owners to update software and use strong passwords
Researchers at Chalmers University of Technology developed a new optimisation method for near-optimal large-scale fleet control, accounting for uncertainties in real-world transport systems. The method combines machine learning techniques with advanced optimisation models to predict travel demand and charger availability, improving fle...
A Lehigh-led team is developing an AI platform called UniBrain to simplify brain imaging analysis, reducing the time and expertise required. The platform aims to unify image preprocessing and downstream brain network analysis, integrating information from multiple imaging modalities.
A new materials library was created to improve sodium-ion battery performance, with iron and vanadium linked to higher energy storage and stability. The library could eventually allow AI to identify new battery compositions before they are made, accelerating the development of better sodium-ion batteries.
Tamim Asfour has made significant contributions to the development of humanoid robots that can perform complex tasks and learn from humans. His research has led to the creation of cognitive robots with episodic memory, enabling them to store, retrieve, and describe experiences in natural language.
Global health experts warn that public health institutions must verify the accuracy of AI-generated content to ensure public health information is trustworthy. The authors argue that knowledge is a public good and that sifting truth from AI-generated content is now a core responsibility for public health institutions.
Researchers identified two common drugs, aspirin and lansoprazole, that may prevent early labor when given together during pregnancy. The study used a digital twin system to simulate the immune system of individual pregnant women and found that the combination had a large effect on delaying labor.
A new AI platform, MAESTRO, has been developed to identify promising drug combinations for overcoming treatment resistance in prostate cancer. The platform uses large language models to analyze scientific knowledge and ranks the most promising candidates, providing a clear biological rationale for experimental testing.
The New Jersey Supercomputer Initiative (NJSCI) aims to advance research, innovation, education, and workforce development across New Jersey. The initiative will establish a shared supercomputer to support researchers, students, and businesses, enabling them to tackle complex challenges like cancer research and flood risk mapping.
Oreejoy's AI-powered wearable device uses pressure sensing, signal processing, and artificial intelligence to assess mobility and provide personalized sensory cues for people with neurological disorders. The startup aims to make effective diagnostics and rehabilitation more accessible, improving quality of life for patients undergoing ...
Engineers at the University of Cincinnati developed a system to detect damaged drones using sound recordings. The system can alert operators before takeoff, preventing crashes and ensuring public safety. The technology aims to enable fully autonomous drones that can detect damage and return to base before catastrophic failure.
A new study by NYU Tandon School of Engineering found that 86% of AI-generated advice fell short on accuracy, completeness, or safety, and 60% omitted critical safety warnings. Researchers tested four AI systems, including IPV-specific chatbots, and found that most responses were rated unsafe or incomplete.
Researchers at the University of East Anglia found that tiny changes in pupil size and blinking patterns can reveal an individual's mental workload during virtual reality tasks. This could lead to wearable systems that automatically adapt therapy in real-time, offering personalized pain relief for millions living with chronic pain.
Engineering students at the University of Cincinnati are developing an autonomous drone system that uses virtual reality for warehouse inventory management. The system, which uses photorealistic models and LIDAR technology, increases accuracy and efficiency in inventory tracking and leak detection.
The FullStaQD project has developed a reference architecture for quantum software that integrates applications and hardware, breaking down quantum software into three layers: applications, system functions, and hardware. This blueprint provides a foundation for the joint development of larger software systems by multiple teams.
Researchers developed an AI tool, DROID-RV, to estimate right ventricle size and function from echocardiogram videos. The tool accurately predicted right ventricle function and was associated with cardiovascular outcomes, including heart failure, atrial fibrillation, and mortality.
A new study reveals AI chatbots successfully identify redundant medical interventions in most cases, but stumble in gray areas where human consensus is divided. The study found that AI models often echo human disagreements, and struggle to convey restraint in social interactions with patients.
A new paper argues that human creativity's value lies in the transformative experiences and personal development it requires, shifting the focus from the finished product to the experiences that shape artists and cultures. The research highlights how AI can enrich human experiences and encourage curiosity, exploration, and new perspect...
Researchers at Binghamton University have demonstrated that hidden trigger words inserted into LLMs can cause robots to go off-script and perform dangerous actions. To combat this, they developed a defense mechanism called PROTEA, which evaluates task plans for signs of an attack and prevents harmful actions from taking place.
Duke University has launched a national hub to advance trustworthy AI for aging and dementia care, receiving $25 million in funding. The initiative aims to move promising AI from research prototypes into real-world solutions for older adults and people living with dementia.
A UCLA-led study found that misunderstanding an opponent's intentions can fuel conflict, but sharing a visual boundary can help stop it. Researchers created a two-person territorial-conflict computer game to explore this phenomenon.
A research team at POSTECH developed an AI technology that analyzes genetic information from cancer patients to predict their response to immunotherapy. The study, published in npj Digital Medicine, outperformed biomarkers in predicting treatment response across four cancer types and six patient cohorts.
Researchers at Korea University developed an AI-guided strategy to discover potential antiviral compounds against viruses like HIV-1 and SARS-CoV-2. The study combined AI-guided molecular design with experimental validation to identify potential inhibitors of a host enzyme involved in viral replication.
A team at Kyushu University analyzed 1,623 annual securities reports from 340 Japanese food and beverage companies, finding that firms place greater emphasis on social issues, while environmental disclosure remains relatively limited. Disclosure priorities shifted after 2022, with greenhouse gas emissions showing the sharpest increase.
Frontiers will provide AI-powered data curation for 1,000 Horizon Europe datasets and award 100 FAIR Data Publication Grants to early-career researchers. This initiative aims to make research data easier to discover, understand, and reuse, supporting the European Strategy on AI in Science.
The PROSPERO project seeks to overcome current telecommunications fragmentation by designing a unified framework for network and computing as a single logical resource. The goal is to achieve higher network reliability and reduce costs, enabling new business models and monetizable services.
Researchers at The University of Tokyo developed a contact-engineered MoS₂ memtransistor with brain-inspired memory, enabling analog synaptic behavior and temporal ECG signal processing. The device demonstrated gradual long-term potentiation and depression, spike-amplitude-dependent plasticity, and timing-dependent conductance changes.
A new study reveals that afforestation and reforestation efforts face significant social and institutional challenges, including poverty, food insecurity, and weak governance. This threatens to limit the global potential for forests to capture CO2, with the biggest hurdles facing tropical countries.
Scripps Research has received a significant award of in-kind compute power from the Jen-Hsun and Lori Huang Foundation to advance the adoption of AI into biomedicine. The grant will support projects building a virtual human immune system and charting the
Researchers developed AI tools to estimate biological age and predict cancer survival risk using facial photographs, with improved accuracy and robustness across photo variations, and lifestyle factors influencing facial aging, including smoking and heavy alcohol use.
The research team trained the core image-generating model from scratch on Earth observation data to learn the features of diverse terrain and urban areas. The AI can restore cloud-covered images and convert radar images into images that resemble ordinary photographs with high accuracy.
A real-time wildlife detection and driver warning system was successfully tested on a 14-mile stretch of Route 460 in Southwest Virginia, showing drivers reduced speeds by an average of 3 mph when approaching activated signs. The system, which utilizes thermal cameras and AI software, demonstrated an overall accuracy of 97 percent for ...
Researchers developed an AI-assisted method to optimize woven textile patterns for sound absorption, leveraging symmetry and physics-based simulations. The approach identified promising designs in under a minute, with predicted performance matching detailed simulations within 5%.
Researchers at Tohoku University developed a method to predict sewer pipe corrosion by analyzing sulfate-reducing bacteria in wastewater. The method, which combines ecological modeling and machine learning, identified pipe sections at higher risk of corrosion and improved the accuracy of inspection records.
Researchers have developed a unified robot AI system that can understand human instructions, predict future scenes, and carry out tasks. The technology achieved an average success rate of 78.4% in real-world robot manipulation settings, outperforming comparison models.
Cog, a new journal published by AAAS and SJTU, aims to bridge neuroscience, psychology, and artificial intelligence to understand cognition, emotion, and behavior. The journal welcomes interdisciplinary research with clear mechanistic insights and encourages studies bridging fundamental discovery and translational applications.
Researchers developed a framework to detect abnormal lung sounds in children using smartphones, achieving high accuracy and outperforming traditional methods. The framework can provide an assistive tool for home-based respiratory screening, improving remote clinic consultations and community healthcare.
A new study designs an informed consent framework for AI-related clinical trials, highlighting the need for transparency and disclosure. Meanwhile, a Canadian physician-led campaign supports the Safe Social Media Act, aiming to protect children's mental health. Additionally, a promising autonomous oxygenation device shows promise in re...
Researchers propose a framework for Third-generation metabolomics that combines standardization, automation, and intelligence to enhance reproducibility and data quality. The framework aims to transform metabolomics from fragmented procedures into a more intelligent and reproducible research process.
Researchers at the University of Tokyo developed a new AI model, MatterGen-e, which designs electronic structure and crystal structure together for new materials. The model produces substantially more valid and novel structures than a previous model, and provides density of states for newly generated materials.
The Simons Foundation is launching a $15M initiative to support mathematicians in navigating the AI era, leveraging AI tools and shaping collective standards. The foundation will provide funding, resources, and community-driven initiatives to help researchers adapt to changes in the field.
Researchers Dr. Hutan Ashrafian and Dr. Ahmad Guni discuss the expanding capabilities of medical AI, including predictive and preventive applications, and the need for technical and health-system advances to improve health outcomes. They will explore the safeguards needed for responsible and equitable adoption of increasingly capable A...
The $5 billion DOE Genesis Mission aims to create an AI ecosystem to accelerate nuclear reactor deployment and reduce operating costs. University of Tennessee's Professor Nick Brown leads a team developing AI-based tools to evaluate reactor design and safety, with the goal of deploying a commercial-grade platform in 2027.
Researchers used AI to analyze nearly 150,000 Reddit posts about unapproved GLP-1 weight loss drug retatrutide, identifying potential side effects that differ from clinical trial results. The study found increased appetite to be the most frequently reported symptom, which contrasts with gastrointestinal complaints identified in Phase 2...
UT MD Anderson Cancer Center has made significant advancements in cancer care, including the development of a covalent Werner helicase inhibitor, which showed promising antitumor efficacy in a Phase 1 trial. Additionally, a multicenter clinical trial found that tile-based radiation therapy improved tumor control and lowered the risk of...
JAMA AI will publish research on AI capabilities, clinical applications, and evaluation of their safety and effectiveness. The journal will focus on generative and agentic AI, clinical decision support, and AI-enabled diagnostics and therapeutics.
A new study characterizes the use of OpenEvidence, the most widely used AI-powered clinical platform in the US, among 952,000 clinicians. Most queries were related to treatment and management decisions, with frequent requests for information outside primary specialties.
Researchers developed a tool to analyze gene regulation, revealing new insights into biological variability and potential for genetic medicines. The tool, called FIRE, uses machine learning to identify regions crucial for gene function, suggesting new therapeutic targets for hereditary diseases.
Artificial intelligence can match human doctors in certain tasks, but human doctors are essential for making complex decisions about patient care. AI may identify a patient's medical condition, but human doctors must consider individual circumstances and make choices that work in the patient's best interest.
A University of Toronto research team is developing an AI-powered peptide drug discovery tool with $2.3M in funding from Google.org. The OpenPep project will create an open access database to predict peptide-protein binding, accelerating peptide medicine discovery.