AI Topic
AI in hospitals, diagnostics, drug discovery, health-tech. Curated and summarized from dozens of sources by AIBriefs. RSS
Anthropic's life sciences team has deep research partnerships with BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines and The Column Group. It has also begun in-house drug discovery, including programs for rare childhood muscular dystrophy, with Claude handling much of the scientific and operational work.
Vera Health secured $28.5M for its clinical AI search platform, per MobiHealthNews. No investors, valuation, or launch details were disclosed in the report.
MobiHealthNews interviews Google about using AI to detect bias in healthcare and to accelerate discovery. The Q&A is the only detail available; no specific models, benchmarks, or figures are disclosed in the source.
C-BRAIN, a consortium of 18 institutions launched in November 2025 and led by Dr. Randall Bateman, is working with NVIDIA on open-source AI tools to mine previously inaccessible Alzheimer's data, including unpublished findings and negative results. It targets 57 million people living with dementia and 10 million new cases a year.
MobiHealthNews reports that a simplified setup approach reduces the barriers healthcare providers face when adopting AI. The item carries no further detail on the tooling, vendors, or measured results.
Guest article by Sully.ai co-founders Ahmed Omar and Amit Kumthekar cites AI scribes saving doctors up to 21 minutes of note-writing daily, and The Permanente Medical Group saving an estimated 15,791 hours across 7,260 physicians in about a year. They call for transparency and consent amid lawsuits alleging conversations were recorded without permission.
In a poll of 348 respondents, 36% said the FDA will never allow AI to prescribe without human intervention, while 34% expected it within 2–5 years. Utah runs a regulatory sandbox testing AI prescribing, including an acne-treatment pilot limited to a predefined list of topical drugs.
MobiHealthNews reports an AI tool improved antibiotic prescribing among rural physicians in Australia and New Zealand. No study size, model name, or effect size was disclosed in the available snippet.
MobiHealthNews flags an event featuring ARPA-H on applying AI to healthcare bottlenecks. No speakers, dates, or program details are disclosed in the listing.
A Johns Hopkins Bayview worker uploaded a spreadsheet of patient data to an outside AI platform the hospital had not cleared, discovered Sept. 21, 2026. Exposed details included names, medical record and account numbers, admission dates, medications and provider names — but no SSNs or financial data.
Suki's Series E financing includes new investors Healthier Capital, Crosslink Capital and Hercules Capital, with existing backers Venrock, Flare Capital Partners, Hedosophia, Philips Ventures and Zoom Ventures participating. Patient encounters supported by Suki grew more than 190% year over year in 2026, and revenue doubled in 2025.
The opinion piece argues AI safety efforts target acute harms like psychosis or suicide while ignoring cumulative harm, such as chatbots displacing human relationships over months. It cites early studies linking heavier companion-chatbot use to reduced socialization and increased loneliness, and notes a dozen states now require companion chatbot operators to disclose their tools are not human.
Two Korean AI medical tools — one for Alzheimer's diagnosis, one for artery analysis — received FDA 510(k) clearance, per MobiHealthNews' Asia briefs roundup.
On 200 biopsy-proven MILK10k lesions, a fine-tuned Inception v3 named the correct diagnosis 62% of the time versus 42% for both frontier LLMs. The dermatologist author says both models called a melanoma a harmless seborrhoeic keratosis and recommended discharge.
EPFL's implantable LogSync system-on-chip records activity across multiple brain regions and extracts oscillation and synchrony biomarkers, then uses machine learning to classify cognitive and emotional states in real time. The work, led by Mahsa Shoaran, was published in the IEEE Journal of Solid-State Circuits and developed under a CHF 6.6 million project.
Anima, a healthcare startup, uses OpenAI's GPT-6 and GPT-Live to cut patient call wait times by 75%, per an OpenAI customer story. The setup expands what care teams can handle, giving them more capacity to respond to patients.
A single-center study found a patient-facing conversational AI showed promising feasibility and safety in primary care, per Nature Medicine. The authors say the results provide a basis for larger-scale evaluation.
Google Research detailed at IDWeek 2026 how open foundation models like MedGemma can be tuned for chest X-ray TB screening at the point of care. TB kills an estimated 1.23 million people annually, and scarce radiologists in high-burden settings leave millions of active cases undetected.
A Reddit user with a jaw that catches, a slight underbite and a chin 4 mm off-center used Claude Code to convert their own CBCT and MRI scans into a 3D x-ray film after finding the written reports hard to picture.
Bristol Myers Squibb led the round, which funds Pathos' expansion from oncology into neurological conditions using new Tempus AI datasets, starting with Alzheimer's. Its Foundry model runs thousands of agents to estimate clinical trial success odds.
A German neurosurgeon reports randomizing 300 patients at first neurosurgical consultation, with half receiving a plain-language version of their MRI report or discharge letter generated with Claude. The author describes it as the first published randomized controlled trial using Claude as the intervention in patient care, published this month in NEJM AI.
Anthropic is inviting principal investigators at academic institutions, nonprofits, and government labs to apply for 8–12 week projects using Claude on causal human biology, structural and computational molecular biology, medicinal chemistry, drug discovery, and clinical science. Anthropic runs its models and harnesses at no cost, and outputs are published with open-source data and code.
MobiHealthNews collects on-site commentary from attendees at the HIMSS AI Forum on healthcare AI adoption. The two-part piece is a conference vox-pop, with no product, model, or benchmark details in the available text.
Eko Health released an AI-enabled stethoscope, per MobiHealthNews. The report gives no pricing, availability date, or detail on what the AI detects.
Biograph's first U.S. study using Caristo Diagnostics' CaRi-Heart found elevated coronary inflammation in 42.9% of asymptomatic, low-risk participants, with 17.9% showing very high inflammation in at least one major artery. About 47% of those with no detectable plaque still had elevated perivascular inflammation, and routine markers (ApoB, HbA1c, hs-CRP, AHA PREVENT-ASCVD) showed no correlation.
In a Science study, Microsoft's multi-agent Diagnostic Orchestrator hit ~83% accuracy on complex NEJM cases versus ~36% for generalist physicians with references, and 80% vs 47% on Beth Israel emergency-department cases. TechCrunch notes most gains trace to the underlying models — best results used OpenAI's GPT-5 and GPT-5.5.
Kevin Weil, who led OpenAI's AI-for-science efforts, launched a startup using AI to speed up a key step of drug development, backed by $215 million.
Ling-3.0-flash-Sante is a healthcare-specialized model built on Ling-3.0-flash, with 124B total parameters and 5.1B active. Its name comes from the French word santé, meaning "health."
Guest piece by Keith Schmidt, MD, proposes a five-business-day tally-sheet audit of one procedure type or clinic location, logging four data points per patient contact, before evaluating chatbots, text assistants, or voice agents. It argues pre-procedure question costs hide across staff time, physician interruptions, and lost procedure slots, so no single report captures them.
The deal puts Cosmetic Physician Partners and Aesthetic Partners on CorralData's platform, covering nearly $2 billion a year in aesthetics transactions. AskCorral, its AI agent, will suggest schedule and follow-up changes drawn from results at similar clinics.
Carbon-X is a decoder-only model built on Qwen3-14B that reads raw DNA in a chat and answers in English, scoring 66.7% average across five DNA benchmarks (self-reported). It matches GPT-6.1 Sol within a point at roughly 1,000x lower inference cost, per Hugging Face.
Yorkshire Ambulance Service is using Microsoft Copilot to combine data from multiple systems and spot patterns across patient pathways. In Rotherham, the share of ambulance journeys ending at a given destination shifted, per Microsoft Research.
Opella and Microsoft are partnering on an AI platform aimed at reviewing health content, per MobiHealthNews. No deal terms, launch date, or model details were disclosed in the report.
A prospective multi-center study found non-sonographer operators using low-cost probes plus AI blind-sweep interpretation matched trained sonographers on gestational age and fetal presentation. WHO estimates 250,000+ preventable pregnancy-related deaths yearly, and two-thirds of people lack access to diagnostic imaging.
NVIDIA's video describes AI linking doctor-patient conversations, lab results, operating room planning, supplies, and recovery support into one connected workflow for care teams.
Miami-based OpenEvidence, founded by Daniel Nadler in 2022, says physicians processed over 42 million prompts last month; it raised $250M in September at a $15B valuation. The ad-powered platform booked $100M in ad buys for 2025 and is on track for $400M–$500M by end of 2026.
MRO says Prodigy hits 97% accuracy abstracting unstructured EHR text into CDISC SDTM format, refusing to return results it deems low quality. It works alongside a human abstractor rather than autonomously, automating the 40% of abstractor time traditionally spent finding and copying EHR data.
OpenAI's venture arm and Peter Thiel's Founders Fund joined a $45M seed round for Coherence Neuro, which is building a brain implant to detect and treat cancer. CEO Ben Woodington cites "growing synergy" between AI labs and brain-computer interface makers seeking novel neural data.
RFD4-Proteina combines programmable protein design with search capabilities, released as an open model at GTC Berlin. Siemens Healthineers is also using NVIDIA intelligence for adaptive radiotherapy with agentic co-planning and robotic control.
Radical Numerics used its Omnii-1 model to engineer viruses targeting an antibiotic-resistant infection; lab tests cleared the patient's bacteria before compassionate-use administration.
SimBioSys received FDA clearance for an AI tool that predicts breast cancer recurrence risk, per MobiHealthNews. No launch date, pricing, or clinical validation figures were disclosed in the report.
OpenAI is acquiring Orchestra Bio to bring AI-powered operations to life sciences R&D, aiming to help biopharma teams accelerate drug development. No financial terms were disclosed.
NVIDIA's Cosmos3-H-Dreams model is presented as a generative simulation system for clinical surgery, per a Hugging Face blog post. The post frames it as bringing surgical scenarios to life through synthetic video rather than real footage.
Google Research's AMIE is an AI system patients chat with before appointments, securely collecting information so doctors can spend more time on collaborative care. The post is framed as a Q&A on findings from a clinical study of the system.
Anthropic's internal research model designed minibinders against scorpion neurotoxins over 64 hours, working mostly autonomously from one prompt and four follow-ups. Adaptyv Bio and Twist Bioscience tested 211 designs; 17 met binding criteria. All 211 designs and data are being released.
More than 300 million people use ChatGPT for health and wellness questions each week, per a16z's interview with OpenAI's Karan Singhal. He says the longer-term goal is AI that helps people navigate health proactively while making trustworthy medical intelligence more widely accessible.
Surgery Mode brings intraoperative 3D cancer mapping to robotic and non-robotic prostatectomies, plus new case-management and workflow tools. Avenda cites peer-reviewed research predicting Unfold AI could halve positive surgical margins, and says it outperformed MRI at assessing extraprostatic spread.
Scenario analysis uses evidence-based diagnostic performance as the driving force and clinician trust and adoption as the major uncertainty. The piece argues healthcare needs a workforce that understands AI well enough to use it appropriately.
A Singapore-developed AI system is being used to detect early psychosis in young people, per MobiHealthNews. The report gives no model details, accuracy figures, or deployment numbers.
XRlabs closed an $8.4M round to fund its AI surgical platform, per MobiHealthNews. No investors, clinical partners, or regulatory status were disclosed in the report.
UK digital health firm Numan replaced single-turn static-fixture evals with per-category, simulated-conversation evals for Nu, its AI health coach, after averaging one score hid failures. One low-score cluster traced to the rubric, not the agent: it demanded empathy, exploration and advice within two turns.