Ranks distinct AI gain and problem claims from the published record. Scores reward impact, independent source strength, scale, confidence, and recency.
Researchers built a machine learning pipeline to predict retention times in capillary gas chromatography, training six algorithms on 608 data points covering 73 instrument settings and 61 C1-C12 hydrocarbons from RESTEK chromatograms and literature. The adaptive boosting support vector regression model achieved R2 scores of 0.992-0.993 on validation and testing sets.
In a cross-sectional study of 120 diabetes outpatients at a tertiary Endocrinology and Nutrition Department, researchers used the PIIXMED AI-system to analyze rectus femoris ultrasound images and quantify intramuscular fat percentage (FATi). By publication date 2026-08-24, they reported that higher FATi was associated with adverse metabolic profiles and microvascular complications.
Researchers built stacked deep learning ensembles to detect periapical lesions and stratify them by radiographic size on intraoral radiographs. Using 146 cropped and augmented images, five CNN backbones were combined with MLR and XGBoost meta-learners and evaluated on an internal hold-out test set.
Researchers adapted the individualized polysocial risk score, originally built for type 2 diabetes, to a disease-agnostic cohort of 17,857 adults at University of Florida Health. Using 13 individual-level social determinants of health, they trained XGBoost and logistic regression models to predict all-cause hospitalization within one year, testing multiple fine-tuning levels and sampling strategies.
A cross-sectional survey of 454 final-year dental students in the UAE, Jordan, Malaysia, Oman, and Brazil examined LLM use, motivations, and safeguards. Published August 6 2026, it found ChatGPT predominated at 95.9%, with 39.2% using LLMs several times per week and 28.6% daily for tasks like understanding complex concepts and summarising lecture notes.
A systematic review and diagnostic meta-analysis of 20 studies, 14 in quantitative synthesis, evaluated machine learning models to predict clinical pregnancy or live birth after assisted reproductive technology. As of the August 2026 publication, pooled sensitivity was 0.737 and specificity 0.789 with a DOR of 10.49 and acceptable discrimination on SROC, but heterogeneity was very high.
Researchers compared large language models and clinicians on 34 real-world glaucoma cases, with glaucoma specialists scoring responses on medical accuracy, key-point recall, and logical completeness. AI models produced structured reasoning with weighted mean scores overlapping attending ophthalmologists and exceeding some residents.
Published October 22, 2025, this PLOS One scoping review examined literature on AI and large language models like ChatGPT in medical education. It found potential for personalized learning alongside a set of ethical challenges, synthesizing 50 studies from three major databases covering 2010 to August 2024.
ace