Articles tagged with Artificial Intelligence
A new open-source framework called MEDS has been developed by a Columbia-led team to streamline and accelerate artificial intelligence research using health data. The framework standardizes data format and provides interoperable tools to support machine learning model development, reducing technical barriers and promoting reproducibility.
Frontiers won two EPIC Awards for its landmark whitepaper on responsible AI use in research and publishing, and a digital campaign highlighting research integrity. The awards recognize the institution's commitment to making science open and trusted.
Researchers developed an all-optical artificial synapse that uses light to mimic neural learning and perform in-sensor image processing. The device shows paired-pulse facilitation and depression, allowing it to both enhance and suppress signals, a requirement for realistic neural behavior.
The symposium aims to consolidate and expand partnerships in strategic areas such as energy transition, artificial intelligence, and health. The event will facilitate cooperation among researchers from São Paulo and the UK, with a focus on joint funding opportunities.
Researchers at Texas A&M University are designing how humans will build and survive on the moon, focusing on sustainable construction using lunar regolith. The institution's efforts aim to reduce costs associated with shipping materials to the moon, making it possible to produce rocket propellant locally.
Researchers at PolyU have pioneered a method to store digital data using engineered proteins, achieving high storage efficiency and capacity. They overcame challenges of variable amino acid sequences and degraded proteins by designing a protein template, successfully expressing and retrieving data.
A new Drexel University study finds that most users see AI chatbots as supplements to human therapy, not substitutes. Users express concerns about emotional dependence, misinformation, and overreliance on the technology.
A deep learning model combines knowledge from different catalyst families to identify a top-performing green hydrogen catalyst. The AI correctly predicted the activity ranking of 12 tested catalysts within a previously unexplored material family.
Researchers at the University of Jyväskylä have developed an AI model that can analyze colorectal cancer samples more efficiently and accurately predict the functioning of the cells' DNA repair mechanism. This could lead to shorter diagnosis times, reduced costs, and improved analysis accuracy.
A recent study found that AI-powered chatbots can provide accurate information in 76.2% of healthcare-related questions, with specialties like obstetrics and gynecology performing best. However, internal medicine, neurology, and dermatology saw lower accuracy rates, with risks of harm associated with incorrect responses.
A new study shows AI can generate hundreds of convincing finance research papers efficiently, but also raises concerns about the potential impact on academic community and meaning of scientific discovery. The study demonstrates how AI can accelerate research paper production while highlighting areas for improvement in peer-review systems.
A new study published in PLOS Digital Health identifies simple food substitutions that can improve meal nutritional quality and lower costs. By analyzing 135,491 meal records, researchers trained an AI model to suggest one to three ingredient swaps, resulting in a 10% improvement in nutritional quality and a 22-34% reduction in costs.
A new framework, SUVA, enables organizations to measure and adjust AI chatbots' social preferences, improving their performance in customer complaints and other human-AI interactions. By understanding an LLM's existing tendencies, organizations can decide whether an available model already fits its values and usage scenarios.
A new transfer learning framework connects two gait analysis tasks, predicting continuous gait cycle percentage and classifying discrete gait phases. The method achieved high F1-scores and efficiency, outperforming traditional approaches.
A robot that catches itself when it falls has been developed using reinforcement learning and artificial intelligence. The system achieved an average success rate of 69.4 percent in arresting the robot's fall and returning it to a stable position.
A new viewpoint article highlights the risk of AI 'collusion' with unreliable human input in mental health settings. The author argues that clinical expertise should be included when designing training data and evaluating AI systems to strengthen safeguards.
A new study published in JAMA Network Open found that approximately 16% of patients with COVID-19 developed long COVID, translating to over 18 million Americans. The study used a novel AI algorithm to identify cases undetected by existing diagnostic systems.
Researchers summarize how AI is accelerating inorganic biomaterial development for various biomedical applications. AI-powered property prediction and inverse design tools are being used to discover effective materials with unique properties.
The Salk Institute has promoted Julie Law to full professor and appointed Talmo Pereira as assistant professor, highlighting their innovative approaches to gene regulation and computational neuroscience. Their research focuses on epigenetics and artificial intelligence techniques, respectively.
Researchers propose a new conceptual framework called Health Elements, which integrates technological factors alongside traditional domains as core drivers of health. The framework highlights the increasingly important role of digital infrastructures in shaping health behaviors and outcomes.
A new approach combines MRI scans and AI tools to measure fluid flow in the brain, shedding light on the glymphatic system’s mechanics. The study reveals two main ways the system washes away particles, with one way moving faster than the other.
A team at Polytechnique Montréal has developed a new material that enables direct light processing on silicon chips, reducing the need for signal conversion and amplification. This breakthrough could help sustain the next wave of AI at scale by giving light a larger role in data processing.
Researchers at Duke University introduce Argus, a 20-eyed robot with no front or back, demonstrating dynamic symmetry and improving performance across various measures. The design surpasses the theoretical maximum of 0.6, enabling robustness, energy efficiency, and resilience to damage.
A team from UChicago PME used AI to generate entire chemical formulations for battery electrolytes, balancing complicated tradeoffs and interactions. The research was published in JACS Au and explores a vast chemical space, generating novel candidates satisfying desired properties simultaneously.
The Association for Computing Machinery announced three technical awards for innovations in global wireless standards, machine learning, and 3D generative AI. Erdal Arikan received the Paris Kanellakis Theory and Practice Award for his discovery of channel polarization and polar codes.
A study from Syracuse University and Ipsos found that AI-generated ads are almost indistinguishable from human-made work, but underperform in actual outcomes. The research suggests that humans and AI should work together for a competitive edge.
Researchers developed an Insect Synergy Circuit that integrates body movement and internal physiological information to guide insect navigation. The system achieved high accuracy in classifying environmental conditions, enabling gentle control over the insect's movements.
A new AI model, MutationProjector, analyzes tumor DNA to predict immunotherapy and chemotherapy outcomes in various cancers. The model identifies biomarkers associated with treatment responses and provides insights into why predictions are made.
A mobile app is being developed to identify disease-carrying insects from their wing patterns using machine learning, aiming to reduce the risk of disease transmission in low-income communities. The technology will enable rapid and accurate identification of vectors, tracing disease clusters and responding effectively to outbreaks.
A new study from the University of Miami Miller School of Medicine suggests that machine learning models using patient-reported outcomes and clinical data can forecast unplanned health care use and elevated symptom burden in cancer survivors. The research offers a potential pathway toward more proactive, personalized survivorship care.
Artificial intelligence can bridge the gap between rich multi-scale data generated by modern exercise biomedicine and urgent clinical need for tailored exercise prescriptions. Four key AI modeling paradigms, including time series learning, multimodal fusion, causal inference, and reinforcement learning, are examined.
A new DGMoE framework enhances EEG-based emotion recognition by modeling individual differences, achieving high accuracy rates on public datasets. The framework's two-stage selection mechanism and graph-convolution-based expert modules improve robustness and generalization to unseen subjects.
A new framework combines a custom-built training dataset with transfer learning to improve the detection accuracy of small and distant objects in omnidirectional videos. The proposed model achieved an overall accuracy of 90%, significantly higher than conventional models for small moving objects.
Researchers from Tohoku University's Advanced Institute for Materials Research use AI and data science to extract valuable insights from decades-old experiments and scientific literature. This approach accelerates materials design and screening in catalysis, solid-state electrolytes, and hydrogen storage research.
Researchers developed a neural network approach that learns to clean co-movement patterns in markets before building portfolios. The method achieved lower volatility and higher Sharpe ratios compared to traditional methods.
Zhe Zhu's doctoral dissertation reveals that trusting AI can increase work engagement and build more sustainable careers. GenAI tools can reshape organisational decision-making and employee experiences of work.
A new Frontiers in Science article explores how AI can accelerate scientific discovery in soil science by creating digital soil twins and trialing climate adaptation strategies. Researchers will discuss the potential of multi-agent AI systems to enable autonomous hypothesis generation, experimental design and data analysis during a fre...
A Virginia Tech study found that AI image generators consistently produce more realistic and recognizable images of larger metropolitan areas than smaller towns. The research raises questions about how generative artificial intelligence tools portray places and whose communities are most visible online.
Scientists have demonstrated that megalibraries can design materials with specific properties, accelerating the traditional trial-and-error approach to rapidly designing and testing materials. The platform generates vast datasets needed to train AI systems to discover next-generation materials.
Researchers developed an AI system called CMR-CLIP that can interpret complex heart scans without manual labels, outperforming existing models by up to 35%. The system learned from over a million images and hundreds of thousands of motion sequences collected from Cleveland Clinic.
New AI tools can accelerate soil science by speeding up early-stage work, improving predictions to support decisions on land-use, carbon, and climate adaptation. The system successfully mimicked key parts of the scientific process, with outputs beyond what's currently being used that strongly align with expert research.
A new study highlights the potential of AI tools in soil science, enabling researchers to better understand soil ecosystems and adapt to climate change. The system successfully generated hypotheses on how soils store carbon and what controls their storage limits, with outputs aligning with expert research.
Researchers at UW Medicine Institute for Protein Design and Skape Bio used AI methods to create on-demand molecules that can toggle GPCRs. The approach enables precise control of GPCR signaling in cells, offering new insights into bodily functions and potential medicines for diseases.
Researchers created an AI model that maps how genes work together inside human cells, helping to understand biology and disease. The Gene Set Foundation Model (GSFM) learns patterns in gene groupings and interactions across thousands of biological contexts.
Researchers developed NSYOLO, an AI framework that automatically segments and analyzes nanoparticles with high precision. The framework achieved a mean Average Precision (mAP@0.5) of 0.957, outperforming baseline models and traditional tools in complex imaging environments.
A recent study found that one-third of college students use generative AI to complete assignments, with 9% using it to cheat. The use of AI in higher education has sparked concerns about the validity of assessments and the credibility of university credentials. Researchers are urging assessment reform to address these issues.
Researchers from ECOG-ACRIN Cancer Research Group will present updated outcomes from the STAMP study on Merkel cell carcinoma, primary results from studies in glioblastoma and non-small cell lung cancer, and emerging approaches in precision oncology. AI-driven insights from the TAILORx breast cancer trial will also be showcased.
A new skin-like computing patch can analyze health data using artificial intelligence in mere milliseconds, directly on the body. It overcomes limitations of wearable devices by running AI computations within the body, improving response time for critical medical applications.
The University of Texas System has invested over $470 million in capital projects supporting UT San Antonio's growth as a top public research university. These investments span research, innovation, infrastructure, technology, and patient care, with a focus on expanding the university's research capabilities and technology infrastructure.
Researchers used top Generative AI models to grade hundreds of undergraduate essays, finding that AI only matched human-awarded degree classification around half the time. The AI systems were overly sensitive to linguistic features, giving out higher marks based on essay length and vocabulary range rather than academic quality.
Scientists at Tohoku University uncovered a hidden rule behind dual-atom catalysts, which follow a previously unknown 'dual-Sabatier optima' pattern. This discovery could accelerate the development of cheaper and more efficient fuel cells.
A new approach enables computers and machines to capture images at higher resolution and faster speed, making it impervious to reflective surfaces. The technology uses a virtual screen created by repurposing the surroundings of specular objects.
The University of Texas at San Antonio's three nominated researchers, Lyle Hood, Amina Qutub, and David L. Roberts, have made significant contributions to various fields, including medical devices, AI in healthcare, and mental health services.
A new AI system, Empirical Research Assistance (ERA), can automatically write scientific software programs that outperform human-written ones. ERA combines a large language model with search strategies to explore and refine thousands of pieces of code, reducing the time required for exploration from months to hours or days.
Researchers developed an AI framework, MouseMapper, to map disease-related changes in the entire mouse body. They found widespread inflammation and previously unrecognized damage to facial sensory nerves caused by obesity, which was also observed in human tissue.
A new AI-powered detection system developed by University of East London researchers combines language analysis with behavioural clues to identify genuine reviews. The hybrid fusion model achieved 93% accuracy on Amazon review data and 91% accuracy on Yelp reviews, outperforming traditional methods.
Shandong University researchers have developed MuSE-Promoter, a deep learning framework that integrates multiple complementary ways of looking at DNA sequences. The method consistently outperforms state-of-the-art tools in challenging cross-cell-line transfer and promoter-enhancer discrimination tasks.
A new report from Brookings Institution highlights the federal government's growing use of AI, but also notes significant disparities and bottlenecks to widespread adoption. Large agencies lead the way, while smaller agencies struggle with workforce capacity and trust issues.
Researchers are analyzing paired original and recurrent breast cancer tumors to identify biological factors driving recurrences years after treatment. The TAILORx and RxPONDER trials have provided a large dataset of clinically annotated tumor samples, enabling the study of late recurrence and potential prevention strategies.
A new model combines text mining and machine learning to extract service-specific aspects and customer actions from online reviews. The model effectively identifies core technical issues and user love for a platform, enabling targeted decisions for improvement. Researchers validated the model using 231,705 online reviews of Roblox.