AM-IA brings together news, research papers and applications relating to artificial intelligence published around the world into a single feed — and makes them available in your European language at the click of a button.
Scroll to browse · Click to filter
Brazilian researchers also show that enriching the reference annotations changes the scores displayed without changing the ranking of the document search systems.
A model that links each anomaly to its anatomical region and the lesions that often accompany it better locates signs on chest radios, including on unknown bases.
Driven by self-supervision on colored H&E cuts, the model reaches a AUC of 0.91 to distinguish tumour and normal tissue, then groups tumours into four pseudo-stages.
Four controls defined by two clinicians, formally verified and applied to the responses: each verdict is recalculated on the basis of a named rule and sixteen facts.
The SIM-UT method simulates thousands of realistic exchanges with three chatbots deployed in India, South Africa and Kenya, and reveals a clinically dangerous convenience at MedGemma as well as ChatGPT and Claude.
Tested on 86 clinical tasks in 20 languages, models that directly attribute a probability to each answer are fast and well calibrated on the text, but pick up on the images and face the triage nurses.
In a calibrated simulation on a large American hospital, trigger the full capacity protocol over prediction rather than after exceeding the threshold reduces the time spent saturation.
Driven on 15,360 pairs of simulated images, a CARE network divides the reconstruction error against Noise2Void by nearly ten, then improves the contrast on 27 real spheroids marked with 18F-FDG.
Of 1,552 questions from eight IRM sessions, three open models of 8 to 27 billion parameters would rank among the first 600 to 3,000 candidates. Their clinical reasoning often fades.
The multi-agent system, built on GPT-5, mimics a panel of specialists who choose the exams to prescribe. It remains to be validated in real-life conditions, starting with the Beth Israel Deaconess Hospital in Boston.
Out of three cases of Korean clinical examination and 7,480 double-blind evaluations, three out of four models remain slightly below the human expert; Llama 3.1 70B picks up on all criteria.
Out of 8,524 Brazilian patients, images of a Nikon retinograph were twice as often inoperable as those of a Canon. A study shows that this gap has an uneven impact on 12 AI diagnostic tasks.
When trained jointly on the ADNI and PPMI datasets, an interpretable transfer model is better at diagnosing both diseases than separate models and highlights their shared anatomical abnormalities.
The San Francisco-based start-up wants to use AI to decode the electrical signals of brain tumours and counter them with stimulation. A long-term trial on glioblastoma is planned.
Anthropic and Citrini Research let an internal model work for 64 hours on binders against five scorpion toxins. The first step towards an antivenom, without demonstrated neutralization.
By gradually degrading clinical images, researchers identify the point at which a reliable model ceases to be so, and find that it often engages anyway.
In a trial where every serving consumed was weighed, no LLM achieved the accuracy required at the individual level. More surprisingly, comparing photos before and after a meal often underestimates what was eaten.
A table setting out six categories of high-risk formulations shows that specialised models identify fewer high-risk medicines but make their claims with greater certainty.
Six animal species, more than 60 organs and 14,000 studies: a family of numerical pathology models dedicated to preclinical drug safety trials.
Symbolic reasoning, deep learning, similar case search and LLMs: by combining these approaches, FusionRare outperforms GPT 5.5 and DeepRare on 2,528 individual clinical cases, saving doctors 21 per cent of their time.