Every story in this issue with a short summary, so you can decide what to read before leaving the site. Summaries are written by AI (Claude) only from each publisher's own description and checked by Jev for anything not in that description. They are not full translations. Follow "Read the original" for the full story.
The AI-powered NeuroDR software reportedly removes bone, soft tissue and veins from CTA scans. It then provides 3D MIP views and curved planar reformatted views of carotid and vertebral arteries.
Read the original →A validated deep learning framework segments the diaphragm on routine 2D-cine and 3D thoracic MRI in seconds. It achieved a normalized surface Dice above 0.9.
Read the original →Clarius Mobile Health announced 30 specialized AI models integrated into its handheld ultrasound offering.
Read the original →New labor-backed laws about AI in the workplace are about to take effect in California. It is not clear how many employers realize that software they already use may violate the new rules.
Read the original →An independent evaluation by researchers at the University of Technology Sydney and St Vincent's Hospital Sydney found the tool is associated with improved antibiotic prescribing across rural and remote hospitals in Western Australia. The tool, Clinical Branches, also known as Antimicrobial Pathways (AMPS), gives clinicians step-by-step prescribing guidance based on individual patient characteristics and local antimicrobial resistance patterns.
Read the original →University of Johannesburg researchers published the approach in Scientific Reports. It uses an AI algorithm, routine patient intake data and light-weight computing to estimate risk of suicide-related behaviour after admission to a hospital emergency room. The model was built with historical open-source patient data from the USA and is intended to assist ER staff rather than replace them.
Read the original →Rural and critical access hospitals report under 50 percent adoption of even basic predictive tools, compared with 86 percent among top-tier system-affiliated hospitals. The description says the gap widens each year without a bridge, and fastest for the hospitals with the least room to absorb it.
Read the original →Northrim Horizon announced this week that it has acquired Cressy Door and Fireplace. The company is a Seattle-based provider of installation, repair and maintenance services.
Read the original →A qualitative interview study of hospital staff identified six ethical themes in AI-assisted clinical practice. The researchers proposed a lifecycle-based, multi-actor governance framework.
Read the original →The government has announced a new review of artificial intelligence and its potential uses for the NHS workforce in England.
Read the original →symplr has launched Time and Scheduling, an AI-powered workforce platform that combines predictive scheduling and timekeeping. The company says it cuts nurse manager administrative tasks by 50%.
Read the original →Scientists are the biggest converts in Endpoints Signal's Q3 survey. 78% of them now say they expect AI to boost drug discovery within a year.
Read the original →Clinical Maestro founder Anca Copaescu says AI could help pharmaceutical companies identify contractor performance issues, financial risks and clinical trial problems earlier.
Read the original →Colossal Biosciences uses the AI platform GenoMorph to identify genetic targets for dodo bird de-extinction. The platform pinpoints DNA for distinctive traits.
Read the original →Anthropic's Nasdaq IPO plan, with the ticker ANT, meets its drug discovery push. The description cites Claude Science, six research partners and four named pharma firms.
Read the original →Fixed AI algorithms are already embedded in over 1200 FDA-cleared medical devices, and pharmacists routinely manage and dispense software-driven health care technology. Understanding these tools helps pharmacy teams connect digital health innovations with everyday patient care.
Read the original →AdvaMed encouraged the U.S. Food and Drug Administration to build on its existing regulatory framework for generative AI-enabled medical devices rather than create new frameworks. The comments respond to an FDA discussion paper (Docket No. FDA-2026-N-7874) seeking early input on assessing risks and evaluating such devices before marketing and in post-market monitoring.
Read the original →The Ontario company has received its second FDA clearance in 12 months. The technology helps clinicians recognize a warning sign of a potentially life-threatening air leak around the lung.
Read the original →The authors developed and tested AMHART, a clinician-informed tool to evaluate AI chatbot responses to adolescents disclosing anxiety or depression. Applied to four widely used chatbots, it showed greater coverage for anxiety than depression and for moderate than severe presentations. It also identified gaps in coping guidance.
Read the original →A qualitative study of 45 stakeholders in Shenzhen finds concerns about safety, regulatory clarity, responsibility allocation and cross-sector cooperation in medical AI governance.
Read the original →The article says shadow AI happens daily in GP surgeries, outpatient clinics, pharmacy back offices and finance departments. It says this occurs because the tools are free, instant and useful, and sanctioned alternatives are often slow, clunky or absent. The healthcare sector has spent a decade building governance around the AI it chose to buy.
Read the original →This edition of AI Prognosis covers a new investigation into how Epic's market dominance is affecting health tech innovation and AI adoption.
Read the original →The article argues that Ada is assembling infrastructure the rest of the healthcare AI industry will need. It says the patent announced in April is less a product launch than a statement about where the safety boundary in clinical AI should sit.
Read the original →Clinical AI tools are suggesting more treatments for patients than human providers do. This adds to the ongoing problem of healthcare costs.
Read the original →The previous installment looked at how far AI diagnosis is accepted from the physician's side. This installment turns to the patient's side, where the act of a patient handing their own medical data to AI is workable in some countries and not in others. The difference affects the development of medical AI itself.
Read the original →Rapid advances in generative AI are changing daily life, and medicine is no exception, with uses spreading across clinical care, research and medical education. Atsushi Otsuka, chair professor of dermatology at Kindai University Faculty of Medicine, works as a clinician while also using generative AI. He discusses the current state of medicine and AI, and the possibilities and challenges he sees as a physician.
Read the original →US IT company Oracle announced on the 27th a new service called Oracle AI Security Mesh, which uses AI to diagnose and fix vulnerabilities in corporate systems. Developed with OpenAI, it gives access to an AI model built specifically for cybersecurity measures.
Read the original →