A new approach integrating genetic and blood-based gene expression data substantially improves Alzheimer’s disease detection, offering practical early identification in at-risk populations. Original paper: Integrating polygenic and transcriptional risk scores for detecting Alzheimer’s disease. — Alzheimer’s & dementia : the journal of the Alzheimer’s Association. 10.1002/alz.71408 📄 Read the article

A fully automated deep learning model called PlaqueSegNet can measure coronary plaque volume with accuracy rivaling expert radiologists and invasive ultrasound, while reducing analysis time from 20 minutes to 11 seconds per patient. Original paper: A Fully Automated Deep Learning Model for Quantifying Coronary Plaque at Coronary CT Angiography. — Radiology. 10.1148/radiol.251967 📄 Read the article

A machine learning framework automatically identifies prognostically distinct patient subgroups across cancer types by optimizing directly for survival heterogeneity, without prior knowledge of established risk factors. Original paper: Unsupervised risk factor identification across cancer types and data modalities via explainable artificial intelligence. — NPJ digital medicine. 10.1038/s41746-026-02663-w 📄 Read the article

A multi-center study reveals a critical clinical AI paradox: the highest-performing deep learning model for kidney injury prediction failed in real-world deployment, while a less accurate alternative proved most reliable. Original paper: Deep learning models for acute kidney injury prediction: multi-center external validation and evaluation under simulated continuous monitoring conditions. — NPJ digital medicine. 10.1038/s41746-026-02722-2 📄 Read the article

An AI-powered agentic workflow safely generated hospital discharge summaries in active clinical use, with high safety profiles and measurable benefits to clinician well-being. Original paper: Physician-Reported Safety Outcomes of AI-Generated Hospital Course Summaries. — JAMA network open. 10.1001/jamanetworkopen.2026.16556 📄 Read the article

While antiretroviral therapy is essential for child survival with HIV, a new meta-regression study identifies three additional protective factors: improved water access, pneumococcal vaccine coverage, and maternal education. Original paper: Risk and Protective Factors Associated With HIV-Related Mortality in Children Receiving Antiretroviral Therapy: A Model-Based Meta-Regression. — JAMA Pediatrics. 10.1001/jamapediatrics.2026.1235 📄 Read the article

A new preoperative risk calculator can identify veteran surgical patients at high risk for protracted recovery or permanent loss of independence, enabling surgeons to provide more realistic counseling about long-term functional outcomes beyond the standard 30-day recovery paradigm. Original paper: Predicting Protracted Recovery and Long-Term Loss of Independence after Major Surgery in Veterans. — Annals of surgery. 10.1097/SLA.0000000000007080 📄 Read the article

Researchers developed a machine learning model that predicts major complications after breast reconstruction with 83% accuracy, supporting individualized risk assessment and surgical decision-making. Original paper: Machine Learning Model to Predict Postmastectomy Breast Reconstruction Complications. — JAMA network open. 10.1001/jamanetworkopen.2026.7232 📄 Read the article

A validated AI algorithm can reliably detect left ventricular systolic dysfunction from standard ECGs in Kenya, potentially transforming cardiac screening where echocardiography access is limited. Original paper: Artificial Intelligence Electrocardiogram and Left Ventricular Systolic Dysfunction in Kenya. — JAMA Cardiology. 10.1001/jamacardio.2026.0908 📄 Read the article

A new deep learning model demonstrates substantially higher accuracy than radiologist-assessed breast density for identifying women at risk of breast cancer, offering a more precise pathway for supplemental screening decisions. Original paper: A Deep Learning Breast Cancer Risk Model for Precise Supplemental Screening. — JAMA network open. 10.1001/jamanetworkopen.2026.10559 📄 Read the article