Articles tagged with Artificial Intelligence
Two new Review articles explore AI's application in early cancer detection, highlighting its potential to enhance diagnostic accuracy and improve treatment selection. The articles emphasize the need for a robust, multi-disciplinary approach to integrate AI into medicine.
Researchers developed an AI algorithm to identify normal mammograms and ran a simulation on patient data, revealing that AI can reduce unnecessary testing while maintaining cancer detection rates. The study suggests that AI can help doctors focus on more questionable scans, reducing false positives and improving workflows.
Zero trust architecture (ZTA) aims to prevent internal attacks by restricting behavior based on resource-based security policies. AI-powered automation and orchestration can help overcome implementation challenges, relieving security personnel from manual tasks.
A new generative AI tool models the infant microbiome, predicting neurodevelopmental deficits. The Q-net model predicts which babies are at risk for cognitive deficits with high accuracy, identifying potential treatment targets.
A new machine-learning approach outperforms human testers in generating diverse prompts that trigger a wider range of undesirable responses from chatbots. The technique provides a faster and more effective way to ensure the safety of large language models.
A pilot study found that ChatGPT provided factual responses to common questions about vaccination and STIs, demonstrating its potential as a tool to reduce vaccine hesitancy and provide helpful advice on sexual health. The AI chatbot's accuracy and conciseness were comparable to professional organizations' guidelines.
A new study by Carey Morewedge and colleagues found that people are more likely to recognize bias in algorithmic decisions than their own. This is because algorithms can codify and amplify human bias, but also reveal structural biases in society. The research suggests ways to increase awareness of biases and correct them.
A team of researchers developed an AI-powered computer vision model to detect Brazilian wild animals on roads and warn drivers in real-time. The system uses roadside cameras and portable computers to identify species such as anteaters, wolves, and tapirs, with the potential to save lives and reduce roadkill.
Climate change causes melting of ice sheet, resulting in loss of about 5,000 meteorites per year. Researchers call for urgent action to preserve the scientific value of meteorites and reduce greenhouse gas emissions.
A new statistical-modeling workflow can quickly identify molecular structures of products formed by chemical reactions, accelerating drug discovery and synthetic chemistry. The workflow also enables the analysis of unpurified reaction mixtures, reducing time spent on purification and characterization.
Researchers at Insilico Medicine developed QFASG, a quantum-assisted algorithm generating novel small-molecule structures from fragments. The tool successfully designed inhibitors for cancer-related proteins, showcasing its potential in accelerating drug discovery and development.
Researchers developed a spring-like device that maximizes muscle contractions to power biohybrid robots. The new flexure design enables predictable and reliable movement, allowing engineers to build muscle-powered robots with increased precision and versatility.
A new AI-based video biomarker is associated with aortic stenosis development and progression, allowing for opportunistic risk stratification across various imaging modalities. The study's findings suggest potential applications on handheld devices for early detection and intervention.
Researchers at Princeton University used language models to optimize partial genome sequences and create more effective mRNA vaccines. The model successfully generated hundreds of new sequences, validating its results through lab experiments, and outperforming benchmarks for vaccine development.
Researchers at Nagoya University developed a framework for evolving AI agents with diverse personality traits using large-scale language models. The study found that AI agents can switch between selfish and cooperative behaviors, mirroring human behavior.
Researchers at the University of Sydney and Queensland University of Technology have developed a new approach to designing cameras that process and scramble visual information. The approach, known as 'sighted systems,' creates distorted images that can still be used by robots to complete tasks but do not compromise privacy.
An interdisciplinary team of researchers developed an AI system capable of deciphering genomic language and understanding its functional and regulatory grammar. The Genomic Language Model (gLM) learns from highly diverse metagenomic data, providing insights into gene functions, regulation, and evolutionary relationships.
A new study published in JAMIA found that large language models (LLMs) respond variably to users' motivations, with better results for committed individuals. LLMs struggle to provide relevant information and support for those hesitant to change behavior, highlighting a major gap in their capabilities.
The digital twin of a railway wireless network uses scenario modeling, radio wave propagation characterization and integrated operation platforms to improve safety and efficiency. The study introduces key technologies such as ray tracing and AI-based super-resolution to achieve accurate deterministic modeling of wireless channels.
Recent advancements in AI and IoT have improved earthquake prediction by identifying patterns in historical seismic data. However, limitations such as computational complexity, data quality, and interpretability remain, requiring a comprehensive approach to integrate diverse datasets.
A new study in the American Journal of Preventive Medicine found that large language models (LLMs) like ChatGPT-4 and Google Bard can generate incorrect or incomplete medical advice, highlighting the importance of consulting healthcare professionals for accurate information. The study assessed 56 questions posed to LLMs and found that ...
Researchers developed an AI model to detect viable tumor cells in osteosarcoma patients, improving prognosis predictions. The model showed comparable detection performance to pathologists and reduced inter-assessor variability, enabling timely assessment.
A new study found that AI systems produce much less CO2e per page of text generated compared to human writers and illustrators. DALL-E2 emits approximately 2,500 times less CO2e than a human artist for illustration tasks. The authors suggest collaboration between AI and human efforts for the most beneficial outcome.
In a study published in JAMA Internal Medicine, a chatbot (GPT-4) performed better than internal medicine residents and attending physicians in processing medical data and demonstrating clinical reasoning. However, the bot's accuracy was lower, while it made more incorrect diagnoses.
An international study led by the University of Granada used artificial intelligence to show that our personalities alter the expression of our genes. The findings suggest that certain outlooks on life are conducive to a healthy and long life, while others lead to stress and short life.
A survey of 204 oncologists found that while patients should understand AI models, most agreed that patients should consent to their use. The study highlights the need for rigorous assessments of AI's impact on care decisions and patient autonomy.
Researchers have developed a new open-source algorithm called Conditional Variational Diffusion Model (CVDM) that improves the quality of images by reconstructing them from randomness. The CVDM is computationally less expensive than established diffusion models and can be easily adapted for various applications.
Researchers from the University of Washington created an AI algorithm to analyze infant poses using limited training data. By leveraging generative AI, they were able to produce high-quality results, enabling parents to monitor their babies' daily activities and detect potential health issues early.
Researchers used Google Street View to analyze built environment factors and found they can predict 63% of coronary heart disease variation between neighborhoods. Features like green spaces and walkable roads are associated with lower risk, while poorly paved roads with higher risk.
A team of Columbia engineers created Emo, a robotic face that makes eye contact and uses AI to anticipate and replicate human smiles. The robot can predict facial expressions and execute them simultaneously with humans, reducing disingenuous interactions.
Researchers found that neural networks use a similar path to chart their way from ignorance to truth when presented with images, despite varying network designs and training recipes. This commonality holds the potential for developing more efficient image classification algorithms, reducing the computational power required by AI systems.
Researchers developed Umwelt software that allows blind and low-vision users to build customized data representations without an initial visual chart. The system incorporates three modalities: visualization, textual description, and sonification.
A new study suggests that artificial intelligence can accurately locate lesions in the brain after a stroke, identifying which side of the brain is affected and specific brain regions. The AI model achieved high sensitivity and specificity rates, but its accuracy depends on the quality of available health history information.
Researchers developed a real-time temperature reconstruction technique for HIFU treatment, enabling accurate monitoring and planning. This approach uses deep learning to transform ultrasonic images into temperature images in just a few milliseconds.
Researchers at Uppsala University found that ChatGPT-4 can produce administrative medical notes comparable in quality to those written by doctors, but at a speed of ten times faster. This could have major implications for the healthcare industry.
A new MIT-derived algorithm corrects coarse climate model predictions by 'nudging' them toward more realistic patterns, leading to more accurate forecasts of extreme weather events. The approach uses machine learning and dynamical systems theory to improve the resolution of large-scale climate models.
Researchers found that AI models using language patterns from past studies are less predictive for depression among Black people on Facebook. The study suggests that considering intersection of race and social media can improve depression detection, and calls for more data to learn depression patterns in Black individuals. Better under...
A study using AI models to analyze Facebook posts found that words associated with depression were more predictive of severity in white participants than in Black participants. Researchers highlight the importance of including diverse pools of data to ensure accuracy and tailor effective healthcare interventions.
Researchers have identified two plant compounds with potential as GLP-1 agonist weight loss pills, derived from common plants and potentially offering fewer side effects than current medications. The discovery was made using high-performance artificial intelligence techniques to analyze over 10,000 compounds.
Researchers developed Smart-CKD, a computer-aided diagnostic tool integrating ultrasound data and selected clinical variables to assess renal fibrosis progression in CKD patients. The tool has excellent predictive accuracy and high clinical application value, providing a cost-effective solution for guiding patient management.
Researchers at University of Missouri are developing software that allows drones to fly independently, perceiving and interacting with their environment while achieving specific goals. This technology has the potential to assist in mapping and monitoring applications, such as 3D or 4D advanced imagery for disaster response.
Researchers have developed an AI model called SyntheMol that can design new antibiotics against deadly bacteria like Acinetobacter baumannii. The model identified six potent antibacterial compounds with non-toxic properties, offering a potential solution to the global crisis of antibiotic resistance.
Researchers at Klick Labs developed an algorithm to detect deepfakes with 80% accuracy by analyzing speech pause patterns, offering a solution to the growing problem of AI-generated content. The study's findings suggest that vocal biomarkers can distinguish between real and fake voices, providing a novel approach to flagging deepfakes.
The concept of AI art emerges as a fusion of human senses, offering rich audio-visual experiences. AI technology enhances creative abilities, enabling anyone to become an artist, while also improving production efficiency in industries like movies and games.
Researchers explore ophthalmic image-based AI for diagnosing systemic diseases, leveraging the eye's unique characteristics to assess overall health. The article discusses two primary modes of ocular image analysis and highlights the potential of emerging AI technologies like blockchain and large language models.
Researchers at WVU are developing an AI tool to reduce medication errors that lead to hospital readmissions, aiming to improve patient safety and reduce healthcare costs. The tool will analyze patient records and identify high-risk patients, alerting pharmacists to potential issues.
Researchers used ChatGPT to design a university field course, finding it an effective tool for planning educational trips. The AI model was also adaptable for other industries, such as environmental impact studies and business trips.
Researchers at Georgia Tech have developed a universal approach to controlling robotic exoskeletons that requires no training, calibration, or adjustments. The system uses deep learning to autonomously adjust assistance levels for walking, standing, and climbing stairs, reducing user effort and metabolic expenditure.
Oxford researchers call for a more considered approach to AI development for children, highlighting four main challenges in applying existing ethical principles. The study recommends ensuring fair digital access, transparency, and safety, while involving children in system development.
Artificial intelligence has been developed to spot COVID-19 features in lung ultrasound images, combining computer-generated images with real scans to identify signs of disease. The tool holds potential for developing wearables that track illnesses like congestive heart failure and monitor fluid buildup in patients' lungs.
A mobile app using AI analyzes images of suspected skin lesions and provides guidance on whether they are melanoma or not. The app was tested in primary care and showed promising results, accurately diagnosing melanoma in all cases and determining harmless lesions with a high probability.
Researchers have made significant breakthroughs by harnessing AI in metamaterials research, leading to faster device development and more precise data analysis. This convergence of AI and metaphotonics has the potential to transform various domains, including diagnosis, environmental monitoring, and security.
Researchers at Mass General Brigham have developed two AI foundation models, UNI and CONCH, to advance computational pathology. These models excel in diagnostic accuracy, prognostic insights, and predicting therapeutic responses, with universal applications in anatomic pathology.
Researchers developed a new training tool to help AI programs account for human dishonesty, especially in mortgage and insurance applications. The modified AI was better able to detect inaccurate information from users.
A new pilot study explores the effectiveness of an artificial intelligence-supplemented Cognitive Behaviour Therapy for Perfectionism (CBT-P) intervention for young people with anxiety and depression. The research suggests that AI tools could provide strong guidance through a therapy program, while parent support is also crucial in hel...
A review published in Intelligent Computing outlines the strengths of automatic approaches to designing metaheuristics, which can lead to more successful outcomes and reduce redundant, metaphor-based algorithms. The authors encourage research that relies on automatic design, utilizing modular software frameworks and configuration tools.
Researchers developed two AI-powered tools to identify healthcare-associated infections and found they accurately identified cases even in complex scenarios. However, the tools struggled with missing or ambiguous information, highlighting the need for human oversight.
Researchers at UT Austin develop tools using AI and biosensors to harness microbes for faster drug production, potentially creating a reliable supply of galantamine. The innovative approach uses genetically modified bacteria to produce a chemical precursor of the medication.
A new AI-powered tool developed by the University of Copenhagen and 3Shape predicts how teeth will move, allowing orthodontists to ensure braces are neither too loose nor too tight. The tool uses scanned imagery of teeth and bone structures to simulate how braces should fit, reducing trial and error.
Researchers develop new assessment method to evaluate post-editing effort of machine translations, highlighting the importance of human translators in the process. The study suggests complementing automated systems with a new program to choose tools that increase translation efficiency.