Generative AI in Cancer Care: Mapping Review Finds Rapid Growth, Weak Evidence
A mapping review of 218 studies finds generative AI tools in cancer care are expanding rapidly but are validated mostly ...
Tag: Large Language Models
AI-Crafted Discharge Nutrition Plans Boost Short-Term Dietary Intake After Gastrointestinal Cancer Surgery
A randomized controlled trial of 90 patients found that AI-facilitated, individualized nutritional guidance improved 30-day energy and protein intake adequacy ...
Older Adults’ Views of Successful Ageing Still Seen Through an Individual Lens, Review Finds
A systematic review of 2012–2023 studies finds researchers interpret older adults' lay perspectives on successful ageing mainly at the individual ...
AI Pipeline Turns Brain MRI Saliency Maps into Diagnostic Reports
A benchmark study couples CNN saliency maps, a cortical atlas and three LLMs to generate interpretable diagnostic reports from brain ...
AI Chatbots Show an Accuracy-Practicality Trade-off in Anesthesia Crisis Scenarios
A benchmark study of OpenAI-o1 and DeepSeek-R1 across 30 anesthesia crisis scenarios found that the most accurate model gave the ...
Drift metrics predict how synthetic clinical text fails on real patient notes
Researchers show that lexical and embedding-based drift measures between synthetic and real clinical text moderately predict how much de-identification models ...
Specialized medical language models for diagnosis reviewed across 75 studies
A systematic review of 75 studies from 2021 to 2025 maps trends in specialized medical language models for diagnosis and ...
Agentic AI speeds and standardizes detection of immune-related adverse events
A multi-phase Nature Cancer study found that an agentic AI system detected immune-related adverse events in clinical notes with retrospective ...
New BASM framework charts a maturity path for AI in clinical biochemistry labs
A new open-access framework called BASM maps five maturity tiers for AI-augmented clinical biochemistry, linking analytics, algorithmic governance, and health-service ...
Structured Memory Scheduling Helps Large Language Models Reason Beyond Retrieval
A new framework called SMS-RAG schedules semantic, structural, and episodic memory during reasoning, outperforming existing retrieval-augmented methods on multi-hop QA ...
CarbonBalance Framework Shifts AI Workloads to Cleaner Power, Cutting Simulated Emissions 41.6%
A new framework combining carbon-intensity forecasting with latency-aware scheduling cut simulated AI workload emissions by 41.6% while keeping critical tasks ...
AI matches expert severity ratings in half of hospital adverse event reports
A single-center study found a large language model exactly matched adjudicated expert severity classifications in 49 percent of 286 hospital ...
ChatGPT-4 Boosts Neurosurgery Residents’ Emergency Decision Accuracy in Simulation Study
A prospective simulation study of 55 neurosurgery trainees found ChatGPT-4 assistance raised decision-making accuracy from 58.0% to 81.0%, with the ...
Large Language Models Are Learning to Reason About Buildings, Review Finds
A systematic review of 76 studies finds large language models evolving from passive text processors into active cognitive partners for ...
AI reasoning models boost diagnostic accuracy but raise safety concerns
A Nature Communications study finds large reasoning models make more accurate primary diagnoses than standard language models but produce over ...
AI Fact-Checks Read Better Than They Argue, and Readers Prefer Them Anyway
A study of 1,498 COVID-19 fact-checks found readers preferred LLM-written explanations over human ones despite the AI texts scoring lower ...
New AI Platform Generates University Course Content Aligned with Learning Standards
A new proof-of-concept study reports that an AI platform called iLeapTutor generated standards-aligned university course content that scored 0.829 on ...
Small On-Device AI Model Trails Cloud Giant on Building Engineering Questions
A preliminary case study found that the compact on-device language model EXAONE 4.0 1.2B matched ChatGPT-4o on basic multiple-choice and ...
Surgeons Give ChatGPT Near-Perfect Marks for Knee Replacement Patient Answers
Three arthroplasty surgeons rated ChatGPT answers to 70 knee replacement patient questions near-perfect, including unanimous maximum scores for all 20 ...
AI Is Redesigning Fashion: What 163 Studies Reveal About Generative Design’s Rise and Risks
A systematic review of 163 publications finds that AI-generated content research in fashion has surged since 2022, shifting from GANs ...
