We also have X and podcasts
Daily AI medical research digest in English
A new blood test reliably detects amyloid positivity in adults with Down syndrome, with performance independent of APOE ε4 carrier status—offering a scalable screening approach for Alzheimer’s-related pathology. Original paper: The association between APOE ε4 carrierships and the detection of…
PathPT, a framework combining vision-language foundation models with few-shot prompt-tuning, enables accurate diagnosis of rare cancers using just a handful of labeled examples—addressing a critical bottleneck where expert pathologists and annotated training data are unavailable. Original paper: Boosting pathology foundation…
An AI system matched expert-level performance in leukocyte classification, achieving 95.97% accuracy and ranking second among 16 participants. Original paper: Evaluating AI in leukocyte classification: performance of the AI system against 15 morphology experts. — NPJ digital medicine. 10.1038/s41746-026-02601-w 📄…
A new approach combining physics-based cardiovascular modeling with neural networks delivers accurate, personalized blood pressure predictions from wearable bioimpedance sensors—with minimal training data required. Original paper: Cardiovascular digital twins using a Windkessel physics informed neural network. — NPJ digital medicine.…
A comprehensive analysis of 3,706 individuals reveals that cumulative lifetime exposures—your “exposome”—significantly predict structural brain preservation during aging, with cardiovascular health and lifestyle factors as key drivers. Original paper: Exposome-wide patterns predict brain health in aging. — Nature communications. 10.1038/s41467-026-71271-9…
An AI committee of specialized agents using GPT-4o accurately simulated liver transplant selection decisions, potentially reducing bias and inconsistency in this resource-limited allocation. Original paper: A multiagent large language model-based system to simulate the liver transplant selection committee: a retrospective…
VarNet-T, a deep learning framework, substantially improves identification of cancer mutations from tumor-only sequencing data, achieving up to 33% better accuracy than existing methods and enabling more reliable patient stratification for immunotherapy. Original paper: Improved tumor-only variant calling and mutation…
A multicenter validation of ELDER-ICU, a machine learning model for predicting mortality in elderly ICU patients, reveals critical strategies for successfully deploying clinical AI across diverse international populations. Original paper: Multicenter validation and updating of the ELDER-ICU model for severity…
A radiomic signature of epicardial adipose tissue from routine cardiac CT identifies patients at high risk for heart failure, offering a scalable tool for early risk stratification. Original paper: Early Prediction of Heart Failure From Routine Cardiac CT Using Radiomic…
Researchers developed USDist, a lightweight AI model that accurately diagnoses breast cancer from ultrasound videos and operates on portable devices in resource-limited settings. Original paper: Clinic-aligned Dual Distillation of Video and Image Foundation Models for Automated Breast Cancer US Diagnosis.…
