Articles tagged with Artificial Intelligence
Researchers developed an AI model that estimates age from chest X-rays and found a correlation between age discrepancy and chronic diseases like hypertension and COPD. The model was validated using data from multiple institutions, showing strong results.
Researchers at Carnegie Mellon University have developed a new AI method that uses transformer models to analyze images of human cells. The technology, called subcellular spatial transcriptomics cell segmentation (SCS), accurately identifies individual cells and their constituent parts, revealing critical information about cellular org...
Researchers propose a hypothesis that astrocytes, non-neuronal cells in the brain, can perform core computation as transformers, providing insights into human brain function and machine learning success. This discovery could spark future neuroscience research and help explain transformer performance across complex tasks.
Neuroscientists at UC Berkeley have successfully reconstructed a recognizable song from brain recordings, capturing the electrical activity of brain regions tuned to music attributes. This breakthrough could enable future brain implants with prosodic content, improving communication for patients with speech disorders.
A new study shows that a rule-based natural language processing tool successfully identified patients with unstable access to transportation, food insecurity, social isolation, financial problems, and signs of abuse or exploitation. The tool performed better than deep learning algorithms in identifying these social determinants of health.
Scientists at Max-Planck-Institut für Eisenforschung developed a machine learning model that enhances predictive accuracy in alloy design, uncovering new corrosion-resistant compositions. The model combines numerical and textual data, enabling the identification of optimal alloy formulas.
Researchers in Africa have developed a roadmap to create better AI-driven tools for African languages, addressing the limited availability of language data. The study identifies key players and themes to consider in designing effective language tools, including writers, linguists, software engineers, and entrepreneurs.
A new AI model, dubbed the 'ChatGPT Chemistry Assistant,' extracts relevant information from scientific papers on MOF syntheses. The tool predicts crystalline states of MOFs based on conditions extracted from texts.
A team of researchers from the University of Cambridge developed a way to incorporate human error into machine learning systems, improving their performance in handling uncertain feedback. However, they found that even with uncertainty accounted for, hybrid systems still perform worse than standalone machine learning models.
Researchers at the University of Queensland believe AI can help feed 10 billion people by 2050. The technology can identify optimal genetic combinations, analyze large-scale images, and predict breed performance.
Researchers at the University of South Australia found that chatbots significantly improved physical activity, diet and sleep. Chatbots were effective across different age groups and more conducive to achieving positive outcomes than speech-based AI. A blended approach with human coaching may be the most beneficial.
Insilico Medicine's inClinico platform uses generative AI to predict Phase II to Phase III clinical trial success with an accuracy of 79%. The tool has been validated in various studies and can provide valuable insights for investors and biotech/pharma companies.
Researchers developed a generative AI tool, AniFaceDrawing, to assist users in creating high-quality anime portraits. The tool uses a sketch-to-image framework and employs stroke-level disentanglement to match raw sketches with latent vectors of the generative model.
A team of scientists has successfully elucidated the structure and function of LITE-1, a biomolecule used by Caenorhabditis elegans to detect danger. The researchers used artificial intelligence to predict the structure of LITE-1, which is a channel protein that forms a pore in the cell membrane allowing charged particles to pass through.
A recent study involving over 500 participants found that humans can only detect speech deepfakes 73% of the time. While training participants showed minimal effects, automated detectors performed better and were comparable to human detection abilities.
A recent study found that Google is more current but biased by advertisers, while ChatGPT is more objective but outdated. The researchers suggest combining both strengths to build a better system.
A randomized trial involving over 80,000 Swedish women found AI-supported mammography analysis to be as accurate as two breast radiologists working together in detecting breast cancer. The technology also reduced the screen-reading workload by nearly half without increasing false positives.
Researchers used NLP tools to analyze court transcripts and found significant differences in questioning tactics for Black and white jurors, indicating attempts to shape responses. The study proves that AI can identify biased jury selection, paving the way for real-time monitoring and diversifying juries.
Researchers at UCLA found that GPT-3 performs well on analogy problems, similar to human undergraduates, but struggles with understanding physical space. The AI model's ability to mimic human reasoning is unclear, raising questions about its underlying cognitive processes.
Researchers developed an AI-based system using wastewater data from US counties to predict COVID-19 hospital admissions. The model achieved accurate predictions, allowing public health officials to prepare for and manage pandemic waves.
Researchers developed a reinforcement learning model that incorporates human decision-making criteria to improve AI accuracy in skin cancer diagnoses. The study showed increased sensitivity for melanoma and basal cell carcinoma, leading to more accurate diagnoses and optimal treatment decisions.
A new reference database, Greengenes2, allows researchers to combine 16S and shotgun sequencing data, improving the reproducibility of microbiome studies. This enables the reuse of millions of samples in older studies, strengthening physicians' ability to draw clinical conclusions.
A new study uses machine learning to analyze data from DrugAge, a database of chemical compounds modulating lifespan in model organisms. The researchers create four types of datasets to predict whether or not a compound extends the lifespan of C. elegans, using features such as compound-protein interactions and Gene Ontology terms.
The article explores how AI will revolutionize leadership, with humans needing to understand AI's capabilities to lead effectively. The author argues that AI will be more effective than average leaders in addressing basic psychological needs of employees.
A team of Lithuanian researchers has created an AI-based system to facilitate the rehabilitation process for stroke patients. The system, which uses wearable equipment and electromyography (EMG) technology, enables patients to track their progress and receive feedback on their exercises.
Researchers developed an AI model powered by deep learning algorithms that outperforms traditional dynamic models in predicting central Pacific El Nino events. The study sheds light on the potential of AI to enhance seasonal forecasts, offering significant advancements in climate prediction.
A new editorial explores the potential of machine learning to enhance early cancer detection in primary care, leveraging extensive patient data and improving risk stratification accuracy. The authors emphasize the need for responsible implementation, collaboration, and validation across diverse populations.
Researchers identified 145 key genetic points controlling skeletal proportions using AI on tens of thousands of X-ray images and genetic sequences. This discovery opens a window into predicting patients' risks of developing conditions like back pain or arthritis.
The use of AI models in healthcare for ethnic minority populations may worsen existing inequalities if not addressed. The researchers argue that the lack of diverse representation in research trials and data can lead to ineffective treatments and guidelines.
Researchers developed a tool called SKILL that enables AI agents to learn 102 distinct tasks by sharing knowledge in parallel, reducing the time needed to master new skills. The technology has potential applications in medicine, education, and other fields where vast knowledge is required.
Scientists will use AI-powered monitoring stations to track species populations and assess the effectiveness of agri-environment schemes in improving biodiversity. The study aims to demonstrate how different farming systems and practices affect species populations, providing crucial evidence for sustainable land management policies.
The study evaluates recent research on artificial intelligence-generated molecular structures from the perspective of medicinal chemists, recommending guidelines for assessing novelty and validity. Insilico Medicine's recommendations aim to improve the process of generating and evaluating novel AI-generated drugs.
In a post-truth era, Americans tend to trust information sources with good intentions over those with malicious intent. When presented with claims of fact, participants classified them as true more often if the source was trying to provide an approximate account rather than a precise one.
A study comparing history of present illness summaries generated by a chatbot and senior internal medicine residents found that both were graded similarly by internal medicine attending physicians. The findings suggest the potential of chatbots to aid clinicians with medical documentation, highlighting their accuracy.
A new study by NYU Tandon School of Engineering and Grossman School of Medicine reveals that chatbots' responses to people's healthcare-related queries are nearly indistinguishable from those provided by humans. Participants were able to correctly identify human-generated responses only 65.1% of the time, while chatbot responses were a...
Researchers from Edith Cowan University developed software to analyze bone density scans, detecting abdominal aortic calcification (AAC) at high accuracy. The AI can predict cardiovascular disease events and late-life dementia, allowing for early interventions.
Researchers at the University of Zurich have used AI to uncover the underlying cellular mechanism behind kidney disease in cystinosis. They identified a causal association between mTORC1 protein regulation and the disease, highlighting a promising therapeutic target.
Researchers create a new privacy metric, Probably Approximately Correct (PAC) Privacy, that enables the addition of minimal noise to protect sensitive data. The algorithm automatically determines the optimal amount of noise needed without requiring knowledge of the inner workings of the model or training process.
Recent AI advancements in healthcare and animal behavior pose unique ethical concerns and policy challenges. Meanwhile, generative AI technologies raise questions about copyright and training data, while task automation via AI innovations could reduce income inequality. These pieces explore the benefits and risks of machine intelligence.
Lero is recruiting 16 top international post-doctoral researchers for a €2.9 million fellowship program focusing on privacy, trust, inclusion and fairness in software expertise. The program will provide discipline-specific skills training and enhance career development.
A new study by MIT researchers finds that China's government uses AI-driven facial recognition to suppress dissent, but this approach also spurs the development of better AI tools and other innovative technologies. The study suggests that this 'AI-tocracy' cycle entrenches the regime while stimulating further innovation.
A new report suggests that AI-powered robot companions could alleviate the loneliness epidemic and its associated health risks. Companion robots have already shown promise in reducing stress and promoting physical and mental well-being in socially isolated older adults.
The University of Surrey is involved in six national energy research centres, focusing on sustainable solutions for a greener future. Academics are developing hydrogen and ammonia as alternative fuels and boosting bioenergy production to reduce the UK's energy demand.
A new AI-powered NLP model developed by Emory and Georgia Tech successfully triaged patient messages reporting positive COVID-19 test results with 94% accuracy. The model expedited physician-patient communication, leading to faster access to treatment within a five-day window.
Researchers examine whether robots and AI can replicate the ethical concepts attributed to human nurses, including advocacy, accountability, cooperation, and caring. While AI can inform patients about medical errors and present treatment options, it questions its ability to truly understand and empathize with patients' values.
Researchers found that popular GPT detectors misclassify articles written by non-native English speakers as AI-generated. The study suggests using these detectors with caution and calls for enhancements to address biases. This could have significant implications for students, job applications, and search engines.
Researchers developed an AI model that accurately classifies six types of valvular heart disease using chest radiographs, with high accuracy rates. The model's potential applications include supplementing echocardiography in areas where specialists are scarce and improving emergency care.
The University of Houston's CYBER-CARE research center aims to prevent cyberattacks that could compromise the safe movement of people and goods in the US. The center will focus on four goals, including exploring advanced theories to mitigate the impact of large-scale cyberattacks.
A new study found that deepfake video clips prompted false memories of films in 49% of participants, comparable to text descriptions. Many participants reported remembering the fake remakes as better than originals, highlighting concerns over artistic integrity and shared social experiences.
Researchers demonstrate that AI language models like ChatGPT can generate high-quality fraudulent medical articles with standard sections and references. The study highlights the need for increased vigilance and enhanced detection methods to combat potential misuse of AI in scientific research.
A new study from the University of Montana found that ChatGPT's GPT-4 AI has reached the top 1% in terms of original creative thinking, outperforming human students and national averages. The AI excelled in generating fresh responses, a hallmark of human imagination.
The study reveals five distinct brain types, each suited for its purpose, from a jellyfish's diffuse neural network to the human brain's reflective capabilities. Researchers suggest that autonomous machines can learn from coordination in bees, rapid thinking in birds, and single-mindedness in worms.
Developing a technique to create conductive polymer wire connections between electrodes enables artificial neural networks that overcome the limits of traditional computer hardware. The approach allows researchers to control and train the network using small voltage pulses.
Only 36% of UK employers have invested in AI-enabled technologies like industrial robots and chatbots over the past five years, new research suggests. The survey highlights a growing skills problem, with less than 10% anticipating digital skills training investment in the coming years.
A new method developed by the Max Planck Institute of Animal Behavior has counted Africa's largest bat colony using GoPro cameras and artificial intelligence. The estimate puts the colony at between 750,000 and 1,000,000 bats, making it the largest for bats by biomass anywhere in the world.
A new study by researchers at NYU and the New York Genome Center combines deep learning with CRISPR screens to control human gene expression. The model predicts on- and off-target activity of RNA-targeting CRISPRs, enabling precise gene controls for developing new therapies.
A new study published in NEJM Evidence validates a predictive AI biomarker for androgen deprivation therapy (ADT) benefit in prostate cancer. The study found that personalized use of short-term ADT could improve patient outcomes in men with predominantly intermediate-risk prostate cancer.
A new AI tool, MAFDA, tracks individual fruit flies' complex behaviors and compares them with their genetic backgrounds. This enables researchers to study behavior genetics and gain insights into inherited traits.
Researchers summarize existing compiler technologies in deep learning co-design and propose a new framework, the Buddy Compiler, to address performance bottlenecks in current AI applications. The study highlights the importance of hardware-software co-design in achieving optimal efficiency and effectiveness in deep learning systems.
A new machine learning model developed by University of Pittsburgh researchers uses electrocardiogram (ECG) readings to diagnose and classify heart attacks faster and more accurately than current approaches. The model improves risk assessment, helping patients receive appropriate care without delay.