Articles tagged with Artificial Intelligence
Engineers at Rice University and the University of Maryland developed NeuWS, a technology that can undo light scattering effects, enabling full-motion video through various media. The technology measures wavefronts to rapidly decipher phase information, overcoming the 'holy grail problem' in optical imaging.
A team of researchers created a deep learning system capable of detecting natural disasters using images from Twitter and Flickr. The system, trained on over 1.7 million photographs, was able to accurately identify incidents such as floods, tornadoes, and forest fires.
The study uses AI-assisted methods to discover novel deaminase proteins with unique functions through structural prediction and classification, expanding the utility of base editors. New DNA base editors with remarkable features were developed, enabling tailor-made applications for various breeding efforts.
Researchers developed AEGIS, the first technique to detect backdoor attacks in robust machine learning models. It improves the trustworthiness of artificial intelligence by analyzing robust models and detecting mixed input distributions for poisoned classes.
A new AI tool can identify sperm in severely infertile men in seconds, compared to the hours it takes scientists. The algorithm improves antiquated approaches and increases the chance of a couple conceiving their own biological children while reducing stress on sperm and increasing efficiency.
A study published in Radiology found that AI algorithms with high diagnostic accuracy improved radiologists' detection of lung cancers on chest X-rays. High-accuracy AI led to more frequent changes in reader determinations, suggesting increased human trust in AI.
The VIS+AI framework enables seamless communication between human intelligence and artificial intelligence in visual analysis. It incorporates three iterative loops: interaction, execution, and intelligence optimization, allowing AI to adapt to dynamic data analysis processes guided by humans.
A new AI model has been developed to help doctors determine the extent of cancer within the prostate gland, improving the effectiveness of focal therapy. The model was found to be more accurate than magnetic resonance imaging (MRI) in predicting tumor margins, reducing the chance of cancer recurrence.
Researchers found that ChatGPT 4 can accurately identify appropriate imaging services for breast cancer screening and breast pain. The AI model outperformed its predecessor, ChatGPT 3.5, in providing evidence-backed guidelines for radiologists.
A new checklist has been proposed to reduce bias in AI algorithms, promoting more accurate and fair decision-making. The checklist provides a standardized framework for identifying and addressing biases in AI systems.
A research team at the University of Tsukuba has developed a secure AI technology that allows for the shared analysis of personal and identifiable data from multiple organizations. This will enhance the accuracy of AI analysis, particularly in disease prediction and educational effectiveness.
Research highlights how generative AI models encode biases and negative stereotypes in their users, particularly marginalised groups. This can lead to the mass generation and spread of nonsensical information, influencing human beliefs and perpetuating existing inequalities.
An AI model called SCORE-AI achieved human expert level performance in fully automated electroencephalogram interpretation. This technology has the potential to improve diagnosis and patient care in underserved areas and increase efficiency in specialized epilepsy centers.
Researchers developed Precious1GPT, a multimodal transformer-based approach for aging clock development and feature importance analysis. The model utilizes methylation and transcriptomic data to predict biological age and identify disease-related genes, providing a pathway for therapeutic drug discovery.
Researchers applied machine learning to brain responses and achieved near-perfect classification accuracy for songs that may become hits. The approach, called 'neuroforecasting,' uses data from a small group of people to predict population-level effects without needing to measure the brain activity of hundreds of people.
Researchers created a robot inspired by pangolins' ability to curl up into a ball, with a soft layer and hard metal components. The robot can emit heat when needed and transport particles like medicines, making it promising for minimally invasive medical procedures.
A communication robot is being developed to prevent elderly isolation by facilitating self-disclosure through different communication options. The study found that the most appropriate option depends on the recipient, message content, and sender's gender and personality traits, reducing participants' anxiety in self-disclosure.
Researchers have developed DyLiN and CoDyLiN, methods that handle non-rigid deformations and topological changes in 3D structure representation. These advancements enable real-time volumetric rendering and animation with improved visual fidelity and speed.
Generative AI raises fundamental questions about the creative process and human's role in it. Researchers highlight gaps in understanding perceptions of AI-generated content, ownership, credit, labor economics, and media ecosystem impact.
Researchers at Beth Israel Deaconess Medical Center used a generative AI chatbot to solve complex diagnostic cases, finding it accurate in nearly 40% of cases and providing correct diagnoses in two-thirds of challenging scenarios. The study suggests generative AI has the potential to assist medical professionals in diagnosis.
A new research project at University of Leicester will develop a digital 'twin' of the UK using artificial intelligence and big data to inform environmentally-friendly land use. The project aims to reduce emissions from cattle and sheep farming, which contribute around 10% of the UK's emissions.
A new AI-based technique measures brain fluid flow with unprecedented accuracy, revealing pressures and three-dimensional flow rates. This breakthrough could lead to the development of new treatments for Alzheimer's, small vessel disease, strokes, and traumatic brain injuries.
Research by Francis de Véricourt and Huseyin Gurkan found that trust in machines' decision-making ability is key to effective learning. They discovered that biased learning occurs when humans override algorithmic decisions without observing the machine's correctness, leading to incorrect usage of machines in decision making.
A machine-learning study has found that individual characteristics, including age and weight, determine which drug combinations most effectively reduce COVID-19 recurrence rates. The study used real-world data from a hospital in China and identified unique treatment combinations for different demographic groups.
Lero is recruiting 16 world-class researchers to work on privacy, trust and inclusion in software development through the €2.9 million SyMeCo programme. The programme will attract outstanding international candidates from computer science and related fields.
Researchers developed MOVA to evaluate the pathogenicity of missense variants based on their structural position. The new method showed superior performance in predicting pathogenicity, especially for hotspot mutations in ALS-causing genes like TARDBP and FUS.
The Mori3 robot is a polygon shape-shifting robot designed for space travel, capable of morphing into any 3D object and adapting to various environments. Its versatility makes it an ideal candidate for communication and external repairs in spacecraft.
A new study published in JMIR Cardio reveals the potential of AI-QCA in analyzing coronary angiography, providing accurate measurements of coronary stenotic lesions similar to IVUS. This technology has the potential to improve patient outcomes and support clinical decision-making.
Research by the American Psychological Association found that employees who frequently interact with AI systems are more likely to experience loneliness, insomnia, and increased after-work alcohol consumption. However, these effects may be mitigated by equipping AI systems with social features or limiting their use.
Researchers developed Eye2Gene, an AI system that accurately identifies genetic causes of inherited retinal diseases from retinal scans. The system achieved higher accuracy than human experts in identifying causative genes, with a ranking score above 70% in over 70% of cases.
Researchers have developed a new AI model that can quickly screen large libraries of potential drug compounds against target proteins. The ConPLex model uses language analysis to match potential drugs with proteins without needing to calculate molecular structures, enabling fast screening of over 100 million compounds per day.
MethaneMapper is an artificial intelligence-powered hyperspectral imaging tool that can detect real-time methane emissions and trace them to their sources. With a performance accuracy of 91%, it has the potential to revolutionize the way we monitor oil and gas operations and curb climate change.
Researchers developed an AI model to estimate best-corrected visual acuity from fundus photographs in patients with diabetic macular edema, achieving accuracy within one to two lines on a standard chart. The study's findings support the potential of AI in clinical settings, with opportunities for further improvement.
ChatGPT achieved high scores in clinical reasoning and demonstrated rapid response times to evolving questions. However, its lack of local ethnic knowledge and limited medical references hindered its performance in highly contextual situations.
Researchers found that manual labor's perceived meaning can predict the adoption of autonomous products. Consumers tend to reject these products due to a loss of meaning in life. Companies can highlight alternative sources of meaning to reduce negative effects on adoption.
ChatGPT consistently provided accurate, evidence-based answers in public health questions, beating rival AI assistants. Meanwhile, Amazon Alexa and Google Assistant struggled to recognize or refer users for help.
Researchers at NYU Grossman School of Medicine have developed an AI tool called NYUTron that can accurately estimate patients' risk of death, length of hospital stay, and other factors important to care. The tool achieved impressive results in predicting readmission rates, improving upon standard methods by up to 7%.
Researchers at TU Delft and EPFL collaborated with ChatGPT to design a tomato-harvesting robot, showcasing the potential of AI in robotics. The study highlights the need for verification and validation of LLM output to prevent misinformation and bias.
A study by the University of California - San Diego found that ChatGPT provided evidence-based responses to 91% of public health questions, but only referred users to resources in 22% of cases. The researchers emphasize the importance of promoting human-powered critical resources like life-saving hotlines.
Researchers at KAUST developed smart digital image sensors that can recognize images with high accuracy, using a charge-trapping 'in-memory' sensor sensitive to visible light. The devices have an extremely long-lived retention time of up to 10 years and can perform optical sensing, storage, and computation.
A University of Minnesota team developed a new superconducting diode that is more energy efficient and versatile than past models. The device can process multiple electrical signals at once and has gates to control the flow of energy, which could enable faster quantum computers for industry use and enhance AI performance.
Researchers developed a new AI framework that significantly improves the ability to analyze team communication, enabling adaptive training technologies to facilitate effective team collaboration. The framework performed substantially better than previous AI technologies in classifying dialogue and following information flow.
A recent study published in Radiology found that AI algorithms performed better than the standard Breast Cancer Surveillance Consortium (BCSC) risk model for predicting five-year breast cancer risk. The AI models extracted hundreds of additional mammographic features, leading to improved predictive performance.
Researchers develop a novel voice spoofing detection method combining physiological and physical features, achieving improved performance over existing methods. The proposed method outperforms single systems in terms of EER and t-DCF scores.
Scientists using popular computational tools to interpret AI predictions are picking up too much 'noise' when analyzing DNA. Researchers have found a way to fix this by applying a new line of code, leading to more reliable explanations and potentially unlocking the next breakthrough in health and medicine.
A team of University of Colorado School of Medicine researchers explores the impact of chatbot avatars on patient experience, trust, and compassion in healthcare. They found that chatbots may reinforce social stereotypes and biases, but also have potential to increase trust among underserved patient groups if designed with respect for ...
Researchers developed an AI-powered mass spectrometric technique to determine a polymer's monomeric sequence. This breakthrough may lead to improved materials performance and tackle plastic recycling issues.
Researchers updated their protein localization prediction model, MULocDeep, to provide more targeted predictions for biological discoveries. The tool helps researchers design more effective experiments and advance scientific discoveries related to drug development and treating diseases like epilepsy.
A new Danish research project, CP-SENS, aims to develop a digital twin platform for the mechanical and construction industries. The project will provide companies with access to intelligent IT systems tailored to their needs, enabling them to adopt digital twins without significant financial investment.
Researchers at UCLA Health highlight the impact of test results on lung cancer screening adherence and the potential unnecessary use of axillary ultrasound in breast cancer diagnostics. Meanwhile, a new AI system helps check endotracheal tube placement with high accuracy and user satisfaction.
FES2023 brings together leading researchers and industry experts to share groundbreaking research on AI-based energy solutions, smart grids, and renewable energy systems. The conference covers key themes including planning and operation of energy systems, electricity markets, and demand response.
Researchers at North Carolina State University have developed a new methodology called Patch-to-Cluster attention (PaCa) that addresses the challenges of vision transformers. PaCa improves ViT's ability to identify, classify, and segment objects in images while reducing computational demands and enhancing model interpretability.
Researchers develop AI-based method to quantify cracking patterns in reinforced concrete structures, enabling more accurate and efficient assessments of structural damage. The approach uses graph theory and machine learning algorithms to create a unique 'fingerprint' for each set of cracks, allowing for quick and consistent evaluations.
A WVU researcher is developing a data-driven approach to enhance roadway maintenance operations, incorporating AI analysis of real-time data from sensors on consumer cars and roadside observation stations. The goal is to minimize time and resources spent on less affected roadways, reducing winter maintenance costs and time.
Researchers developed an AI system, Geneformer, to predict how disruptions in human gene connections cause disease. The model, trained on data from thousands of genes, can identify potential drug targets for diseases like heart disease and cancer.
Researchers at Ohio State University developed an AI framework called G2 Retro to automatically generate chemical reactions for molecules. The framework was shown to cover a vast range of possible reactions and accurately predict the best synthesis routes, offering more efficient drug design options.
A new study published in The Neuroradiology Journal introduces an artificial intelligence computer program that can accurately identify changes in brain structure resulting from repeated head injury. This AI tool uses machine learning to process magnetic resonance imaging (MRI) scans and distinguish between the brains of male athletes ...
A deep learning-based framework called EMGSense enables accurate wearable EMG device usage through AI self-training techniques. It achieves average accuracy of 91.9% in gesture recognition and 81.2% in activity recognition, outperforming state-of-the-art approaches.
Researchers at the Beckman Institute for Advanced Science and Technology have developed a new framework for super-resolution ultrasound using deep learning, reducing processing speeds from minutes to seconds. The new technology enables real-time blood flow visualization, overcoming challenges faced by conventional methods.
A deep learning system accurately classifies autism spectrum disorder and predicts symptom severity based on behavioral responses to social cues. The method offers a digital measurement of joint attention, but further validation is needed.