Headlines

A multicenter study found that a deep learning model analyzing full ultrasound videos significantly improved diagnostic accuracy for suspicious breast lesions, lifting junior sonographers' accuracy from 66.7% to 89.2%.

Researchers in Algeria have combined convolutional neural networks with a Swin Transformer to detect network intrusions with up to 99.69 percent accuracy by converting traffic data into images.

A new brain-imaging study presented at the ECNP Congress shows that intensive breathwork reduces cerebral blood flow by up to 45 percent and alters communication within the brain's default mode network, producing experiences resembling moderate to high doses of LSD or psilocybin.

Scientists used leaf-extract-derived silver oxide nanoparticles to convert non-edible Madhuca longifolia seed oil into ASTM-compliant biodiesel with a 92 percent yield and a reusable catalyst.

A multi-task deep learning model that fuses pretreatment MRI and whole slide pathology images outperformed standard staging in predicting chemotherapy response and survival for locally advanced nasopharyngeal carcinoma.

When we think about a diamond, we often think about a material whose value comes from its perfection. In my research, however, I am interested in something almost opposite: the tiny imperfections inside diamonds. These atomic-scale defects can give diamonds new optical and electronic properties, and some of them can serve as quantum systems. The challenge is learning how to control these defects without disturbing the others.

Nadella joined other tech moguls and researchers in calling for stronger safeguards and, in some cases, for the pacing of frontier development.

Nadella joined other tech moguls and researchers in calling for stronger safeguards and, in some cases, for the pacing of frontier development.

Researchers at The University of Manchester have created bulk ferroelectric heterostructures in a lead-free bismuth ferrite-barium titanate ceramic, achieving a Curie temperature of 824°C and programmable electromechanical behaviour for high-temperature sensors and actuators.

Scientists have used graphene oxide nanosheets to deliver neuropeptide Y into the rat amygdala, selectively erasing fear memories through Y1 receptor-expressing circuits while leaving anxiety behaviors intact.

© seekingalpha.com. Use of this feed is limited to personal, non-commercial use and is governed by Seeking Alpha's Terms of Use (https://about.seekingalpha.com/terms). Publishing this feed for public or commercial use and/or misrepresentation by a third party is prohibited.

Financial Times: Sources: Nvidia is in talks to acquire or invest more into US open-weights AI startup Reflection AI; the deal may be an acquihire to avoid antitrust scrutiny  —  Donald Trump's administration hopes the company will rival cheap Chinese alternatives such as DeepSeek

Researchers have sprayed beta-phase silver vanadate nanorods onto greenhouse plastic to convert unused green sunlight into red light, significantly boosting root and stem growth in wheat and barley seedlings.

Scientists have developed a powerful new way to calculate the Kondo effect, a strange quantum phenomenon that appears when magnetic atoms are embedded in metals. Instead of relying on simplified models, the method uses the real atomic and electronic structures of materials, allowing researchers to predict their behavior with far greater accuracy.

A strange chemical signature in old Hubble observations may point to a planet that formed after its star had already died. Astronomers found unusually high levels of niobium around the white dwarf HS 0209+0832, suggesting some of the star’s expelled material may have assembled into a new gas giant. TESS data also hints at a Jupiter-sized planet orbiting just 3.7 million miles from the star.

Princeton Plasma Physics Laboratory may have found a dramatically more efficient route to fusion energy by reversing the traditional approach: heating plasma first, then increasing its density. Their calculations could help future reactors reach self-sustaining fusion with far less energy while avoiding hidden problems that make ignition harder.

Depression may not be one biological condition, but several distinct patterns of brain activity hiding under the same diagnosis. Researchers studying 263 people with depression identified five brain profiles, each showing different patterns of communication between brain regions and associations with symptoms such as anxiety, trauma, and substance abuse.

A ZIF-8-gated cascade nanoreactor converts tumor glucose into hydrogen peroxide and carbon monoxide, depleting ATP and suppressing HSP70 so that mild photothermal therapy near 45 degrees Celsius can kill cancer cells through ferroptosis.

Researchers converted millions of LendingClub loan records into structure-aware grayscale images and showed that a hybrid convolutional neural network and Random Forest model predicts peer-to-peer credit defaults more accurately and more robustly than conventional tabular machine learning models.

A new study in Medical Oncology reports that diosmin loaded into chitosan nanoparticles reduced tumor burden and protected the kidneys of mice with Ehrlich ascites carcinoma by activating Nrf2 antioxidant signaling and suppressing NF-κB-driven inflammation and apoptosis.

A new PLOS Digital Health opinion argues that tens of thousands of authorised AI diagnostic tools rest on benchmark accuracy figures that say little about real patient benefit, and calls for prospective clinical validation, subgroup performance disclosure, and clinician co-design as prerequisites for deployment.

Most of us recognize basic phases of matter and their properties: solids maintain their shape, liquids flow freely with constant volume, and gases expand to fill their container. Quantum phases, on the other hand, are much harder to categorize, overturning traditional notions of what a phase even is and forcing physicists to rethink their definitions.

CBS News: A look at a 1,700-member Slack run by Medicare agency CMS where Microsoft, OpenAI, and other companies help shape policy on AI apps and medical records access  —  A cadre of elite techies, financiers, and high-level Trump administration officials have spent a year quietly charting an extraordinary …

A new study shows that an interpretable AI model fusing 18F-PSMA-1007 PET/CT and multiparametric MRI can predict prostate cancer aggressiveness before surgery with high accuracy.

Researchers used machine learning to create a CKM-specific biological aging index that outperforms existing measures in predicting mortality and identifying high-risk cardiovascular-kidney-metabolic syndrome.

Researchers developed EntQA, a reference-free, entity-centric metric that scores how faithfully large language models preserve clinically relevant concepts in medical question answering.

A new study shows that ultrafine-grained Ti-6Al-4V substrates refine the crystalline structure of high-entropy (TiZrVCrAl)N coatings, shrinking coherent scattering domains from 14.0 to 8.2 nanometers.

Researchers have developed a dual-stream convolutional neural network that fuses RGB imagery with vectorized Error Level Analysis and CBAM attention to localize image forgeries such as splicing and retouching.

Researchers have created solution-blown polyacrylonitrile nanofiber mats embedded with Congo red dye that change from red to blue within seconds to visually signal the oxidation of heated soybean oil.

InvestorPlace - Stock Market News, Stock Advice & Trading Tips Today, I’ll show you two public companies building the future of brain-computer technology: one with a direct connection to a brain-interface developer, and another that makes research tools I’m following as a more speculative idea. The post Beyond Elon Musk’s Neuralink: Two Public Stocks With a Stake in the Brain Race appeared first on InvestorPlace.

A sweeping review of 7,909 studies shows how food-grade nanoemulsions are transforming nutrient delivery, food preservation, and intelligent packaging while exposing critical gaps in industrial-scale validation and safety evidence.

‘Swarms’ of cyborg cockroaches with cameras and miniature medical injectors could deliver supervised emergency care to people trapped

Boston University's Maria Kamenetska has been named a 2026 Moore Foundation Experimental Physics Investigator, receiving $1.35 million to study light-matter interactions at the nanoscale with ultrafast laser spectroscopy.

Firmus said it had made the decision due to "recent market volatility and prevailing market conditions".

A large experiment around the 2024 UK election found that researching political claims with AI chatbots increased accurate knowledge as effectively as Google search, and 6 to 10 percent faster.

But businesses should not be locked into a single machine and must keep the flexibility to move.

A new computational framework called STARIT converts the spatial positions of mRNA molecules inside cells into image tensors, allowing deep learning models to identify cell states that conventional gene-count analysis overlooks.

Nvidia reportedly discontinues GeForce RTX 5090-series products leaving GB202 for data center and professional graphics boards. GeForce RTX 5080 24 GB rumored to become de-facto flagship from the green camp for a while.

Solid tumors are not merely collections of malignant cells but complex ecosystems comprising tumor cells, stromal cells, and the extracellular matrix (ECM).

Researchers have built an explainable ensemble deep learning framework that fuses four transfer learning models using CMAES optimization to classify brain tumors in MRI scans with up to 99.93 percent accuracy.

Participants meditated inside an MRI tube, with and without psilocybin.

New films about Zuckerberg, Musk, Altman and AI arrive as public concern about the technology and its leaders rises.

New films about Zuckerberg, Musk, Altman and AI arrive as public concern about the technology and its leaders rises.

A fully automated deep learning system called CFAPS segments carotid plaques on multi-contrast MRI and classifies their vulnerability in under a second, matching expert radiologists and predicting stroke-free survival across four clinical centers.

Researchers synthesized multi-walled carbon nanotube/gallium selenide nanocomposites by pulsed laser ablation in liquid and used DFT calculations to reveal how interfacial charge redistribution dramatically reshapes the hybrid's anisotropic optical properties.

For the better part of a century, people have been trying to understand dark matter. This invisible substance

Live Science spoke with Steve Winter ahead of the release of "Big Cats."

Physicists in the Department of Physics at the University of Oxford have used a hybrid quantum computer, made up of qubits and quantum oscillators, to observe the Aharonov–Bohm effect in a quantum simulation. The effect is a quantum phenomenon in which a particle can acquire a measurable phase by traveling around a magnetic flux, even though it never passes through a region where a magnetic field is present.

A new ternary ZIF-8/NiFe2O4/magnetic graphene oxide nanocomposite removes up to 96 percent of tetracycline from water, recovers magnetically and retains over 92 percent of its capacity after five reuse cycles.

A national discrete choice experiment of 211 Chinese urologists found that surgeons' stated preferences for robot-assisted versus laparoscopic prostatectomy shifted dramatically depending on whether the hypothetical patient was financially well-resourced or constrained, with cost becoming the dominant factor only for financially constrained patients.

A prospective study found that an AI-enhanced 12-lead ECG score accurately identified angiographically significant coronary artery disease in patients undergoing elective angiography.

A new computational study shows that the magnetic field induced by a flowing Cu–TiO2/water hybrid nanofluid inside a parabolic solar collector's receiver tube can be systematically strengthened or damped, opening a potential new control channel for solar-thermal systems.

Researchers used Millettia ferruginea leaf extract to synthesize a zinc oxide/bentonite nanocomposite that showed enhanced antibacterial activity against four bacterial strains.

Researchers at the University of Illinois Urbana-Champaign and the National Energy Technology Laboratory showed that coal-derived carbon nanodots enable the growth of ultrathin, high-quality dielectric layers on two-dimensional semiconductors, clearing a longstanding obstacle for next-generation electronics.

China has unveiled a smart robot prototype designed to assist with wheat breeding in the latest step in its ongoing drive to advance agricultural technology and bolster food security. The robot made its public debut at a recent seminar on wheat breeding, state-run China Science Daily reported on Friday. Traditional wheat breeders rely on a wealth of accumulated experience to identify and select strains with superior traits from tens of thousands of hybrid offspring. But this requires manual...

Researchers at Institute of Science Tokyo developed pyrene-modified polyionic complex vesicles whose membrane stiffness can be tuned to program their fluorescence response to hydrostatic pressure, enabling optical pressure sensing in aqueous and biological environments.

Nobel Physics Prize for IceCube For the first time since 1991, the Nobel Prize in Physics belongs to

A study of real Lyft, Waymo, and adaptive cruise control trajectories finds that automated vehicles follow more safely than humans but induce shorter headways and higher collision risk in the human drivers behind them.

At events marking China’s National Day in Africa on October 1, robots provided the entertainment. In Luanda, Angola, dancing robots amused guests and ambassador Zhang Bin called the industry “a small snapshot of the Chinese economy’s move towards innovation and higher-quality development”. In Brazzaville, Congo, nearly 400 guests – including Congolese officials and Chinese business representatives – watched on as a robot dog performed. Guests also interacted with a robot dog at a Chinese New...

Researchers have built an AI framework that combines live video of floc formation, real-time turbidity sensing and physics-based equations to predict the optimal coagulant dose in water treatment before conventional sensors can react.

Belgian researchers built a model-scale wind turbine called TWIST and showed that a refined continuous actuator line simulation accurately reproduces its turbulent wake in a wind tunnel.

Alzheimer’s disease is complex, presenting and progressing in so many different ways in individual patients that it’s difficult

A comprehensive BMC Medicine review finds that artificial intelligence and digital health tools show strong potential to transform cardiometabolic care, provided the field addresses real-world validation, algorithmic transparency, privacy, and equitable access.

A Stowers Institute scientist will help lead a $28.6 million ARPA-H effort to generate billions of protein aggregation measurements and train AI to predict the molecular changes behind neurodegenerative disease.

Researchers at Huzhou University have developed ESM, a deep reinforcement learning system that adapts video streaming bitrates to improve perceived quality by nearly 10 percent while cutting modeled energy consumption by over 14 percent.

Most of today's computer chips — also sometimes known as conventional integrated circuits — use electrons to deliver their signals via semiconductor materials and metallic components. In contrast, photonic integrated circuits use faster signal carriers, such as photons or quantized "light particles."

A new deep learning framework called MedFusionNet combines diffusion-based synthetic MRI generation with a fusion of three neural architectures to classify pediatric dyslexia and spelling deficits with 96 percent accuracy.

Photoelectrochemical water splitting offers a clean path to produce hydrogen fuel directly from sunlight and water. Quantum dots—nanoscale semiconductor crystals—are highly promising for this application due to their tunable bandgaps and strong light absorption. However, most high-performance quantum dots rely on toxic heavy metals such as cadmium or lead, limiting their practical deployment.

Solid tumors are not merely collections of malignant cells but complex ecosystems comprising tumor cells, stromal cells, and the extracellular matrix (ECM). The tumor microenvironment (TME) fosters immunosuppression through tumor-associated macrophages (TAMs), myeloid-derived suppressor cells (MDSCs), regulatory T cells (Tregs), and cancer-associated fibroblasts (CAFs), while physical barriers such as dense ECM and abnormal vasculature hinder drug penetration.

A hybrid deep learning framework combining two complementary LSTM models with Bayesian ensemble stacking predicts winter road surface temperature more accurately than ten rival models across one, three, and six hour horizons.

A KFF Health News probe found tech firms lobbying officials in a CMS-run Slack. What it means for seniors using Medicare-promoted health apps.

A multicenter Chinese study of 3,148 Parkinson's disease patients developed and validated interpretable survival prediction models and deployed the best-performing Cox model as a web-based clinical tool.

Researchers have developed a cable-driven skating training robot controlled by a novel fractional-order active disturbance rejection strategy that significantly reduces trajectory tracking errors compared with conventional PID control.

A scoping review finds that GPT-4o and DeepSeek-R1 have surpassed the passing threshold on China's national medical licensing examination, while warning that inconsistent evaluation methods still cloud the field.

A two-center study trained a deep learning radiomics model on routine breast ultrasound images to predict sentinel lymph node metastasis preoperatively, achieving moderate discrimination across cohorts while the authors stress that larger prospective validation is needed before clinical use.

Author(s): Refaat Ismail, Ashish Kakkar, and Anatoly DymarskyThe authors reformulate the NP-hard problem of finding a quantum stabilizer code’s minimum distance into a Quadratic Unconstrained Binary Optimization (QUBO) framework compatible with classical, quantum, and hybrid solvers. Testing the approach with quantum annealing and hybrid optimization, they identify hardware connectivity and embedding as the main limitations of current implementations. [Phys. Rev. A 114, 042610] Published Fri Oct 09, 2026

Author(s): Soren S. Plueger, Bethany R. Wilcox, Steven J. Pollock, and Gina PassanteThrough think-aloud interviews, this study explores how students learn quantum computing, identifying three key cognitive resources they activate and occasionally misapply when reasoning about the CNOT gate. [Phys. Rev. Phys. Educ. Res. 22, 020145] Published Fri Oct 09, 2026

Researchers in China have combined an online-learning radial basis function neural network with super-twisting sliding-mode control to achieve chattering-free, high-precision trajectory tracking for the UR10 collaborative robot arm under friction, parameter uncertainties, and external disturbances.

A virtual reality experiment with ninety older adults in Ghent, Belgium, found that more greenery and less traffic independently increased self-reported feelings of relaxation while walking through simulated urban streets.

Stylish upcycling, with a retro flare, at Spain's Museo de Historia de la Computación.

One damaged data center has supercomputers used for training Yandex’s AI model.

A Houston study of nearly 8,000 pediatric appointments found that while social risk data did not improve AI no-show prediction, unconfirmed appointments, inactive patient portals, transportation needs, and housing instability emerged as modifiable targets for intervention.

Researchers in Vietnam have developed WGPose-LL, a deep learning method that uses paired well-lit images to guide human pose estimation in extremely dark conditions, setting new performance baselines on the ExLPose benchmarks.

Researchers have woven naturally occurring halloysite nanotubes into a polymer-ceramic electrolyte, creating fast lithium-ion channels that boost conductivity, suppress dendrites, and extend battery life.

Researchers in China have developed a Sparrow search algorithm-optimized attention-LSTM neural network that predicts step error in five-axis finishing machining with over 99 percent accuracy and far greater speed than traditional geometric methods.

A new study shows that a corona discharge in a flame can either promote or suppress the formation of titanium- and silicon-containing nanoparticles below 10 nanometers depending on precursor concentration, with preferential negative charging explaining the suppression.

Every black hole hides a question at its center. Einstein's theory predicts that whatever falls in is crushed into a singularity, a point of infinite density where the theory itself breaks down. Most physicists expect quantum gravity to replace that point with something finite. But the center lies hidden behind the horizon. How could we ever learn what is there?

InvestorPlace - Stock Market News, Stock Advice & Trading Tips Jonathan Rose takes us beyond Musk's company to explore two publicly traded stocks he's watching. One has a direct business relationship with an OpenAI-backed brain-interface developer, potentially positioning it to supply an important piece of this emerging technology. The other is a more speculative research-tools company with a very different set of opportunities and risks. The post Beyond Elon Musk’s Neuralink: Two Public Stocks With a Stake in the Brain Race appeared first on InvestorPlace.

Hugh Langley / Business Insider: Sources: as Google prepares to roll out its Gemini 4 Argon, staff are testing a new version internally named Carbon; one staffer says it “feels like Opus 5.5”  —  As Google prepares to roll out its new Gemini 4 ‘Argon’ model to the public, staff inside the company are testing better versions …

While some leaders warn of mass unemployment, Bezos predicts productivity gains could let people work less without sacrificing their ability to support a family.

NASA has taken the next step in the process of developing a new all-glass telescope for the LISA (Laser Interferometer Space Antenna) mission, a space observatory designed to detect ripples in space-time called gravitational waves.

DEEP Robotics tested waterproof DR02 humanoid robots directing traffic and helping tourists in Hangzhou, China, raising questions about safety and reliability. The post China Tests Humanoid Robots to Direct Traffic in Heavy Rain appeared first on TechRepublic.

The automaker has been criticized for misleading drivers with the feature’s name. Now it’s renaming it for European regulators.

The European Union’s Digital Markets Act transparency website lists Huxe as Apple’s newest acqui-hire, and at least the fourth AI-related deal disclosed this year. Here are the details. more…

Nikon announced that it had disqualified the winner of its Small World in Motion competition over how A.I. was used in his submission.

A novel nanovesicle-based platform developed at Institute of Science Tokyo can measure hydrostatic pressure by converting pressure-induced molecular changes into fluorescence signals.

Researchers at Shiv Nadar University have developed a dual-path AI framework that combines a residual gated autoencoder with patchwise nearest-neighbor reconstruction to detect and localize visual anomalies without labeled defect examples.

Researchers have developed an AI-driven scheduler that predicts congestion in edge-cloud environments before it happens, cutting response times and protecting critical workloads.

© seekingalpha.com. Use of this feed is limited to personal, non-commercial use and is governed by Seeking Alpha's Terms of Use (https://about.seekingalpha.com/terms). Publishing this feed for public or commercial use and/or misrepresentation by a third party is prohibited.

A systematic review of 21 preclinical studies finds that curcumin and its nanoformulations modulate key cancer signaling pathways and enhance chemotherapy sensitivity in ovarian cancer models, though clinical confirmation remains pending.