Ranks distinct AI gain and problem claims from the published record. Scores reward impact, independent source strength, scale, confidence, and recency.
Researchers evaluated 7T susceptibility-weighted imaging to map glioma vascular microenvironment features for preoperative differentiation. In 218 histologically confirmed cases, they extracted vascular topology, density and intensity after Frangi filtering and 3D skeletonization and trained machine learning models to distinguish IDH-mutant versus wildtype and WHO Grade 1-2 versus 3-4.
Researchers introduced EndoVLM, a vision-language assistant tailored for gastrointestinal endoscopy, using ConvNeXt as a hierarchical visual encoder instead of Vision Transformers and a three-stage fine-tuning schedule to align the projector, adapt the visual backbone, and improve instruction following. By August 2026 publication, the model was tested on Kvasir-VQA and externally validated on Gastrovision.
On August 26, 2026, a peer-reviewed study reported a proof-of-concept deep learning approach to distinguish two diagnostically challenging thyroid neoplasms, NIFTP and IEFVPTC, using gross pathology photographs from 87 patients. Using frozen pretrained CNN backbones and traditional classifiers with nested cross-validation, the prespecified EfficientNet plus Random Forest model achieved a pooled AUC of 0.788.
A 2026 review in Therapeutic Innovation & Regulatory Science assessed artificial intelligence for real-world evidence in oncology from a statistical perspective for regulatory decisions. It described RWE as limited by data quality, population selection, treatment characterization, outcome assessment, and statistical methodology, and discussed machine learning and generative AI combined with causal inference frameworks as approaches that may address those challenges.
Published February 24, 2025, this Nature Communications review examines how artificial intelligence is used to model and understand extreme weather and climate events including floods, droughts, wildfires, and heatwaves. It reports that AI has improved weather forecasting, model emulation, parameter estimation, and prediction of extremes, while also discussing methods to identify and explain events more effectively.
This February 2025 systematic review of 103 papers examined how Cognitive Load Theory, Educational Neuroscience, and AI/ML combine in adaptive learning. It found that systems using EEG, fNIRS and other physiological signals to feed CNN, RNN and SVM models can automatically manage cognitive load and dynamically adapt learning pathways for K-12 and adult learners.
On March 12, 2025, an umbrella review in the Journal of Medical Internet Research synthesized 18 reviews from 274 screened records on AI in nursing. It found consistent reports of potential advances in patient care and clinical workflows alongside an urgent push to update nursing curricula with AI-driven tools and ethics training.
Published March 24 2025 in the BMJ, this methods article describes PROBAST+AI, an updated assessment tool for prediction models built with regression or artificial intelligence methods. It splits assessment into model development and model evaluation, each organized around participants and data sources, predictors, outcome, and analysis domains.
ace