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.
In the second of a three-part podcast episode, Wendie Berg, Manisha Bahl and Amy Patel discuss how to handle disagreements with adjunctive AI in breast imaging. They also cover whether to include AI assessments in PACS and whether patients should pay for the use of adjunctive AI.
Read the original →A new deep learning model embeds Turing-inspired reaction-diffusion mathematics and attention mechanisms into a ResNet50 backbone. It classifies dermoscopic images of melanoma and benign nevi.
Read the original →Researchers in Ethiopia have built an interpretable deep learning ensemble that identifies malaria parasite species from thin blood smears. It is reported to reach near-perfect accuracy and provides transparent visual evidence.
Read the original →A two-center Chinese study shows that a machine learning model can combine radiomics features from ultrasound and DCE MRI. The model can non-invasively distinguish luminal from non-luminal breast cancers.
Read the original →The Faculty of Pharmacy at Universitas Airlangga (UNAIR) held the International Graduate Student Conference on Pharmaceutical Sciences (IGSCPS) on Saturday, September 18, 2026.
Read the original →Sanford Burnham Prebys scientist Lukas Chavez has won the 2026 AWS Children's Health Innovation Award. The award will fund Leitstern, an AI-powered system that helps clinicians rapidly identify personalized treatments.
Read the original →Leaders from healthcare organizations in New York, West Virginia and Boston shared thoughts on encouraging AI use among late adopters. They spoke on a panel in New York City.
Read the original →Munich start-up calliora has raised $6 million from BHI and redalpine. The money is for an A.I. platform that catches hospital billing errors before stricter audits begin in 2027.
Read the original →Researchers at Sylvester Comprehensive Cancer Center have published a framework for building AI tools. The tools would personalize evidence-based support for cancer survivors between clinic visits.
Read the original →A Severance Hospital team built an AI model using eight pre-surgery data points to predict acute kidney injury risk. It reached an AUROC of 0.809 for any-stage injury and 0.908 for moderate-to-severe (stage 2-3) injury.
Read the original →The UK government confirmed on October 6, 2026 that it will accept all 44 recommendations of the National Commission into the Regulation of AI in Healthcare. It also set out how they will be delivered across the UK.
Read the original →Nigerian AI researcher Mary-Brenda Akoda has won the $100,000 Nigeria Prize for Science and Innovation. She developed a technology she says can reduce MRI scan time by up to 90 per cent.
Read the original →An interactive skills webinar is aimed at PhD students, postdoctoral researchers and academic scientists who want to move into biotech. The description says AI, automation and data-driven approaches are reshaping scientific roles and industry expectations. It adds that entering biotech often requires more than a strong publication record or technical expertise.
Read the original →Italfarmaco and VeriSIM Life will combine AI and HDAC expertise in a preclinical drug discovery deal. The deal is worth over USD 100 million.
Read the original →Twist Bioscience shares have traded up by 10.19 percent after news of upbeat synthetic DNA demand. Traders are reacting to a new antibody data deal with Eli Lilly's AI/ML drug discovery unit.
Read the original →The UK has accepted all 44 AI regulation recommendations. The MHRA has opened applications for the third phase of its AI Airlock.
Read the original →Medical AI company Coreline Soft has secured official AI software certification from the Japan Radiological Society for its lung nodule detection software, AVIEW Lung Nodule CAD. The certification is described as strengthening its foothold in the Japanese medical market.
Read the original →Ministers have approved all 44 recommendations made by the National Commission into the Regulation of AI in Healthcare, which was led by NHS doctors. The government says NHS patients are to benefit from safe, effective and trusted AI technology while robust safeguards are maintained.
Read the original →Internal medicine physician Dr. Kiran Raj Pandey and AI expert Bishesh Khanal have launched the podcast series 'Dialogues in Health AI'. According to Dr. Pandey, the series serves as a platform for researchers to connect with international knowledge production on health policy.
Read the original →Five academic publishers have struck a deal with AI company Redpine: BMJ Group, Sage, IOP Publishing, IGI Global Scientific Publishing and Wanfang Data. The deal provides AI agents with licensed access to 20 million peer-reviewed research papers.
Read the original →Making AI more culturally fluent could help people feel understood. But feeling understood and receiving safe mental-health guidance are not the same thing.
Read the original →GENSHI AI will exhibit at Medical Japan Tokyo. It will display on actual hardware its medical AI platform MedLocal, which runs on NVIDIA DGX Spark. The press release is dated October 6, 2026.
Read the original →A lecturer at Niigata University of Health and Welfare worked with a lecturer at Yamaguchi University Hospital's radiation therapy department. Using publicly available breast cancer clinical trial MRI data, they examined how to evaluate AI that captures differences in treatment effect between individual patients.
Read the original →