This post hoc secondary analysis of a cluster randomized clinical trial assesses whether implementation of a clinical decision support system is associated with increased antihypertensive treatment intensification and improved blood pressure (BP) control in primary care practices.

This Viewpoint discusses the pros and cons of artificial intelligence (AI) in physician learning and offers ways in which AI systems should be used to support rather than replace clinical judgment.

This Viewpoint explores whether the utility artificial intelligence (AI) tools for non-native English researchers may compromise the epistemic accuracy of scientific writing.

This qualitative study aimed to evaluate the usability, perceived effectiveness, and acceptability of a voice-based generative artificial intelligence assistant for educating patients about intravitreal therapy for wet age-related macular degeneration and supporting treatment understanding and adherence.

This quality improvement study evaluates the physician-rated safety and use of artificial intelligence (AI)–generated hospital course summaries and their association with physician burnout levels.

This cluster randomized clinical trial of patients with chronic kidney disease assesses the effectiveness of a clinical decision support system in improving process measures and clinical outcomes in primary care.

This Viewpoint discusses how overreliance on artificial intelligence (AI) can lead to deskilling and mis-skilling among clinicians still in training and the importance of thoughtful design and implementation into the clinical learning environment.

In this episode of JAMA+ AI Conversations, Associate Editor Yulin Hswen and JAMA Health Forum Editor Sandro Galea discuss the issues surrounding AI’s move from the laboratory into health policy.

This cross-sectional study evaluates the proportion of individuals in Kenya with high probability of left ventricular systolic dysfunction as determined by an artificial intelligence electrocardiogram algorithm vs traditional echocardiography.

The rapid increase in use of ambient scribes and the potential implications for clinical practice are the focus of this installment of the Healthy Dialogue podcast, featuring JAMA Senior Editor Derek C. Angus, MD, MPH, and Vincent X. Liu, MD, MS, chief data officer of The Permanente Medical Group at Kaiser Permanente.

This cohort study of screening mammograms from women 30 years or older explores whether a deep learning model more accurately stratifies risk of breast cancer than breast density–based criteria.

This Viewpoint argues that artificial intelligence can be used to unburden physicians from administrative work that pulls them away from patient care.

This Perspective considers a licensure framework for autonomous clinical artificial intelligence (AI) in light of the clinical care workforce shortage.

This cohort study evaluates a deep learning model’s accuracy in predicting the need for neonatal respiratory support among pregnant women scheduled for cesarean delivery.

This prognostic study describes the development and validation of machine learning models trained on both structured data and manually abstracted variables from unstructured clinical notes to predict major complications after postmastectomy breast reconstruction.

This randomized clinical trial evaluates the efficacy of computer-aided detection–assisted vs standard colonoscopy in routine practice among patients with a high risk for colorectal cancer.

Nature Machine Intelligence, Published online: 13 May 2026; doi:10.1038/s42256-026-01237-5

Torres et al. present ApexGO, a generative approach capable of redesigning peptide antibiotics to better kill drug-resistant bacteria. They validated candidates in laboratory tests and mouse infections and matching or outperforming standard antibiotics.

Nature Machine Intelligence, Published online: 12 May 2026; doi:10.1038/s42256-026-01230-y

Del Dottore et al. present an octopus-inspired soft robotic arm that uses optoelectronic mechanosensors in its suction cups to detect contact forces and infer object positions.

Nature Machine Intelligence, Published online: 11 May 2026; doi:10.1038/s42256-026-01232-w

Hunklinger and Ferruz provide an overview of explainable artificial intelligence methods for protein language models.

Nature Machine Intelligence, Published online: 11 May 2026; doi:10.1038/s42256-026-01231-x

Bode and Chandler discuss how human–machine interactions across the AI life cycle affect human control and decision-making in the military domain.

Nature Machine Intelligence, Published online: 07 May 2026; doi:10.1038/s42256-026-01241-9

A promising foundation model is developed for a range of downstream applications in natural product mining.

Nature Machine Intelligence, Published online: 07 May 2026; doi:10.1038/s42256-026-01235-7

Li and Walsh show that a unified ‘platonic’ geometry emerges across the representations of independent foundation models for learning interatomic potentials. The observation enables cross-model comparison, embedding arithmetic and ground-truth-free diagnostics.

Nature Machine Intelligence, Published online: 06 May 2026; doi:10.1038/s42256-026-01236-6

He et al. comprehensively test the reusability of PanPep, a meta-learning framework for peptide–TCR binding prediction. The authors demonstrate the utility of PanPep in extended testing scenarios, leveraging proposed independent datasets and negative sampling strategies.

Nature Machine Intelligence, Published online: 05 May 2026; doi:10.1038/s42256-026-01227-7

Thiemann et al. introduce TrajCast, a neural network that bypasses force calculations to directly predict atomic trajectories, enabling time steps up to 30 times longer while accurately reproducing physical properties of molecules and materials.

npj Digital Medicine, Published online: 29 May 2026; doi:10.1038/s41746-026-02820-1

Effectiveness of AI and rule-based conversational agents for depression, anxiety and stress: A meta-analysis

npj Digital Medicine, Published online: 28 May 2026; doi:10.1038/s41746-026-02784-2

Guideline for secondary use of health records within Norwegian and EU regulatory frameworks

npj Digital Medicine, Published online: 27 May 2026; doi:10.1038/s41746-026-02796-y

Reassessing negative 24 h pH impedance tests for hidden gastroesophageal reflux disease using multi feature anomaly detection

npj Digital Medicine, Published online: 27 May 2026; doi:10.1038/s41746-026-02803-2

A clinical neuroimaging platform for rapid, automated lesion detection and personalized post-stroke outcome prediction

npj Digital Medicine, Published online: 27 May 2026; doi:10.1038/s41746-026-02782-4

Personalised thrombo-embolic risk prediction after endometrial cancer surgery: an explainable AI approach using SHAP

npj Digital Medicine, Published online: 27 May 2026; doi:10.1038/s41746-026-02721-3

Multicenter randomized trial of a digital therapeutic game for executive function in children with ADHD

npj Digital Medicine, Published online: 27 May 2026; doi:10.1038/s41746-026-02792-2

Deep chemical structure graph learning deciphers the lipotoxicity code of hypertriglyceridemic pancreatitis

npj Digital Medicine, Published online: 27 May 2026; doi:10.1038/s41746-026-02771-7

Consistency in causal reasoning for large language models in scenarios of HIV antiretroviral treatment, drug interactions, and side effects

This study introduces ChexGen, a generative vision-language foundation model that provides a unified framework for text-, mask-, and bounding box–guided synthesis of chest radiographs, and demonstrates its applications in training data augmentation, data-efficient learning, and bias detection and mitigation.

This Review Article examines how large language models could reshape informed consent in clinical research by making it clearer, more accessible, and more responsive to participants’ needs through plain-language revision, translation, and comprehension support. It also highlights the accuracy, bias, and oversight safeguards required for responsible use.

This Perspective examines the emerging legal risks of ambient artificial intelligence documentation tools in clinical care, with a focus on consent, data transmission, and state-level recording laws. It highlights gaps in current regulatory frameworks and outlines operational and policy strategies to mitigate liability and ensure ethical, transparent implementation.

This Policy Corner clarifies key concepts and differences across various applications of AI in health care to help clinicians, patients, managers and policy-makers better understand, apply, manage and govern these technologies in practice.

Automation bias from large language models presents a potential patient safety risk, as physicians may overtrust plausible sounding but erroneous artificial intelligence outputs. This randomized clinical trial quantified that risk, revealing a 14-percentage-point decrease in diagnostic accuracy among physicians trained in AI literacy, even when their consultation of the AI model was entirely voluntary.

This Editorial discusses the automation bias that was found in the study by Qazi et al., “Automation Bias in Large Language Model–Assisted Diagnostic Reasoning among AI-Trained Physicians.”

Health care delivery organizations are facing a surge of artificial intelligence vendor offerings but often lack the information needed to evaluate these systems responsibly. This Perspective introduces the Health AI Partnership (HAIP) AI Vendor Disclosure Framework, a structured approach that clarifies what information vendors should provide, helping health systems make informed procurement decisions; strengthen transparency; and promote safety, efficacy, and fairness.

This article introduces EchoNext-Mini, an open dataset of 100,000 electrocardiograms with curated structural heart disease labels and an accompanying convolutional neural network model for detecting structural heart disease from electrocardiogram data. The authors describe the dataset’s development and demonstrate that the EchoNext-Mini model achieves strong performance, providing an open resource to support further research.

This Perspective explores the rapid integration of artificial intelligence into health care, highlighting how safety-net hospitals and underserved patients risk being left behind. It argues that deliberate policies, equitable access, and clinician-led governance are essential to ensure AI strengthens care without deepening existing disparities.

Artificial intelligence now plays a central role in drug discovery. In antibody research especially, models support sequence design, paratope prediction, affinity maturation, and developability screening. Yet as these systems move from benchmark tasks to real discovery programs, a familiar pattern is emerging: models often perform well within known training space, then weaken when asked to ...;

ISTANBUL, Turkey — Modern breast augmentation is evolving as patients increasingly move away from exaggerated cosmetic trends and seek more natural-looking, personalized results. According to Dr. Leyla Arvas, MD, an Istanbul-based plastic surgeon internationally recognized for her work in aesthetic surgery and body contouring, modern breast aesthetics is no longer defined simply by implant size or ...;

Elon Musk recently brought back one of his long-standing ideas during the OpenAI trial: that as artificial intelligence becomes more powerful, humans may need a closer, more direct connection to machines. He calls this future “human–AI symbiosis.” The phrase sounds ambitious, but the underlying idea is straightforward. If AI begins to outpace human cognition in ...;

Which AI mental health chatbots do psychologists believe are safe for daily use? Find out about the benefits and risks. Discover the benefits and risks of using them. Learn about research showing their effectiveness and the impact of using an AI mental health bot  on mental health. What ‘Safe’ Actually Means for AI Mental Health ...;

Modern medical facilities require healthcare software as their fundamental operational base. Hospitals clinics and startups use digital platforms to manage patient data while they automate their operational activities which leads to improved clinical performance.  Healthcare applications must satisfy mandatory requirements which will be established in 2026. The systems should implement HIPAA and GDPR regulations while ...;

In 2024, Douglas Ford found himself in a position no one wants to be in: waiting nine hours in an emergency room while a life-threatening condition went untreated. For the Harvard-trained scientist, the experience wasn’t just a personal trauma; it was a data point. He realized the bottleneck wasn’t a lack of medical expertise, but ...;

Introduction: The Quiet Shift No One Is Talking About Artificial intelligence is already reshaping ophthalmology, but not in the way most clinicians think. Much of the current conversation focuses on diagnostics, imaging analysis and workflow optimisation. These developments are important, but they are not where the most immediate transformation is occurring. The real shift has already ...;

Artificial intelligence is entering healthcare faster than almost any technology before it. In 2024, 71 percent of hospitals reported using predictive AI integrated into their electronic health records, up from 66 percent the year prior. Automated monitoring tools are being deployed in schools, where more than 200 institutions now use AI to identify students experiencing ...;

Health AI is everywhere right now. Foundation models trained on patient records, agent-like systems that spit out care suggestions, deep learning pipelines that catch disease earlier than a human might, none of that is hypothetical anymore. The architecture exists. The money is there. Real deployments are already happening. And inside controlled environments, the results can ...;

We’re sharing an update on our mental health work, including some new changes to better connect people with the right information.

Here are Google’s latest AI updates from March 2026

The winners of the MedGemma Impact Challenge demonstrated the potential of Google’s open medical models for solving diverse healthcare challenges.

Google and DocMorris have announced a partnership to create a more intuitive and supportive digital health experience.

<p data-block-key="1tp3e">At Google’s annual health event, The Check Up, we shared how our products, research and partnerships are making the most of AI to help everyone live healthier lives.</p>

At The Check Up, Google announced a $10M investment in clinician AI training and how AI is upgrading Search and Fitbit for better health data.

An overview of how Google Earth AI is supporting the global health community’s work to predict outbreaks and deliver proactive care.

A new Google AI initiative aims to improve heart health outcomes for people living in remote Australian communities.

New research shows how Google AI helps radiologists detect breast cancer earlier and more accurately, while giving radiologists more time for patient care.

The expert panel of this week's HealthLeaders The Winning Edge webinar provide insights on how to determine the ROI of...

Beth Israel Lahey Health replaced costly legacy fax infrastructure with cloud-based digital solutions and AI automation, saving millions while accelerating...

The real payoff from automation is emerging through a few key areas.

As public policy squeezes budgets and payers delay payments, a new survey reveals that revenue cycle leaders are adjusting expectations...

The medical director of clinical informatics at AltaMed Health Services and chief medical information officer at Presbyterian Healthcare Services share...

By targeting after-hours documentation, Onvida Health demonstrates how rural health systems can leverage ambient AI to reclaim clinical capacity, accelerate...

Find out about best practices for adopting and implementing technology in clinical care from a pair of experts.

The real payoff from automation is emerging through lower administrative costs, workforce redesign and risk reduction?not dramatic increases in collections.

UT Dell Medical Center in Austin will be built with AI tools and other technology as foundational elements to improve...

As new regulations squeeze budgets and payers delay payments, a new McKinsey survey reveals that revenue cycle leaders are adjusting...

AI-driven, automated claim denials and excessive medical record requests are creating significant administrative waste for health systems, even in instances...

Objective;To assess the cost-effectiveness of using artificial intelligence (AI)–derived software to assist reading CT scans of the chest to identify and analyse lung nodules compared to unaided reading in symptomatic, incidental and screening populations.Methods;Decision tree structures were developed in TreeAge Pro 2021. Structures were informed by British Thoracic Society clinical guidelines and clinical opinion. Results were presented as incremental cost-effectiveness ratios (ICERs) expressed as cost per quality-adjusted life-year (QALY) over a lifetime from the UK National Health Service and Personal Social Services perspective.Results;For the symptomatic population, the unaided radiologist reading strategy dominated the AI-assisted reading strategy. In the incidental population, unaided radiologist reading was cost-effective with an ICER of approximately £1000 per QALY. Conversely, in the screening population, AI-assisted radiologist reading dominated unaided reading. The cause of AI assistance being cost-effective depended on the number of people who had undergone CT surveillance because of non-cancerous findings. Given the limitations in the quality and quantity of evidence to inform inputs, these results should be interpreted with caution.Conclusion;Current analyses based on limited evidence suggested that, in the symptomatic and incidental populations, unaided radiologist reading may be the more cost-effective strategy, while in the screening population, AI-assisted radiologist reading appeared to be the dominant strategy. Better quality evidence is required to have a definitive answer about their cost-effectiveness.Advances in knowledge;This paper shows whether adding AI-derived software to radiologists' reading of CT scans to identify lung nodules offers good value for money.;

Objectives;Deep learning neural network (DLNN)-based tools can automate body composition analysis for cancer cachexia research. We aimed to evaluate a DLNN tool trained on a European population of Chinese cancer patients.Methods;Computed tomography (CT) images at the 3rd lumbar vertebral (L3) level of Chinese gastric cancer patients were retrospectively collected. An externally validated DLNN tool (Mosamatic) was used to segment skeletal muscle, visceral adipose tissue (VAT), and subcutaneous adipose tissue (SAT). Manual segmentation was performed using SliceOmatic software (TomoVision, version 5.0). Geometric similarity between automated and manual segmentation, and the reliability was assessed.Results;The cohort comprised 203 patients with a median body mass index (BMI) of 22.2 kg/m², and 604 CT images at L3 were collected. The median Dice Similarity Coefficient (IQR) of skeletal muscle, VAT and SAT were 0.973 (0.961-0.980), 0.980 (0.964-0.989), and 0.967 (0.945-0.977), respectively. The median Lin’s Concordance Correlation Coefficient for skeletal muscle area (0.983), VAT area (1.000), SAT area (0.998), skeletal muscle radiation attenuation (0.995), VAT radiation attenuation (0.994), and SAT radiation attenuation (0.997) demonstrated excellent reliability. Low BMI (<18.5 kg/m²) and ascites impaired the agreement between the 2 methods. The automated method showed high diagnostic concordance with manual segmentation for sarcopenia (<span style="font-style:italic;">κ ;= 0.843, <span style="font-style:italic;">P ;< .001) and myosteatosis (<span style="font-style:italic;">κ ;= 0.946, <span style="font-style:italic;">P ;< .001).Conclusions;The Mosamatic tool displays excellent generalizability to analyse body compositions in Chinese gastric cancer patients and can facilitate cachexia research.Advances in knowledge;The Mosamatic tool displayed excellent generalizability without recalibration to analyse body composition on the 3rd lumbar vertebral CT images in Chinese gastric cancer patients.;

<span class="paragraphSection">Abstract;Recent advances in artificial intelligence (AI) offer significant potential to address the growing bottleneck in radiology caused by an increasing volume of imaging studies amidst a global shortage of radiology professionals. This study presents a comprehensive review of the latest developments in AI, particularly in vision-language models for radiology report generation, providing radiologists with a current reference. We conducted a focused literature search for studies published from 2020 to 2024 and included 14 studies in our review specifically on chest X-ray datasets with limited coverage of 3D modalities, reflecting the early stage of research and ongoing methodological advances in report generation for volumetric imaging. We analysed the model architectures, report generation capabilities, training datasets, evaluation metrics, and performance of these models. Our review highlights the evolution of AI in radiology report generation and underscores the critical need for diverse datasets and standardized evaluation metrics. Despite rapid progress, current AI models are not yet capable of consistently producing high-quality reports and require further improvements in data diversity, model training, and evaluation metrics to achieve a level comparable to human experts.;

Objectives;There is a growing need to develop user-friendly, bladder-specific image analysis tools that can produce reliable artificial intelligence (AI)-quantitative imaging biomarkers (QIBs) derived from multiparametric (mp)MRI data for clinical applications. To address it, we developed an AI-powered BLADdEr multiparametric MRI Analysis for Clinical Application (AI-BLADE, current release v1.0) toolbox designed for extracting mpMRI-derived quantitative metrics.Methods;AI-BLADE is an advanced tool for bladder-specific mpMRI data analysis with 2 core functionalities: (1) Deep Feature Analysis (MRI-DFA toolkit) and (2) Data-Driven Model-Based Analysis (MRI-MBA toolkit). AI-BLADE offers customizable options and serves as a one-stop shop solution for bladder cancer (BCa) clinical applications. The models within DFA and MBA were tested separately on 2 patient cohorts. DFA was used to classify BCa histology subtypes (<span style="font-style:italic;">n; = 104) with T2-weighted images, while MBA was used to interrogate tumour physiology by deriving mpMRI QIBs, including apparent diffusion coefficient (ADC), and volume transfer constant (K^trans) obtained from 34 BCa patients.Results;Out of the 17 AI models tested, the VGG19 model with a decision tree classifier and no feature selection for the fully connected layer 7 achieved the highest area under the curve of the receiver operating characteristic of 0.79 in classifying BCa histology subtypes, demonstrating the strongest performance. The mean ADC and K^trans values were 1.22 × 10⁻³ (mm²/s) and 0.27 (min⁻¹), respectively, reflecting underlying tumour physiology.Conclusion;The AI-BLADE (v1.0), a flexible and user-friendly software toolbox for analysing mpMRI data, shows strong potential for application in BCa oncology, offering capabilities that can enhance diagnostic accuracy and support improved patient outcomes.Advances in knowledge;This is the first study to design, develop, and implement a novel bladder-specific AI toolbox for analysing mpMRI data. AI-BLADE enables an advanced image analysis workflow, facilitating AI-QIB-based clinical decision-making for patients with BCa.;

Objectives;Deep learning models developed for the classification of radiological reports have lacked explainability. We aimed to validate and explain a pretrained classification model by applying it to the removal of confounding data from a radiological dataset.Methods;Two radiologists categorized 2038 anonymized MRI head free-text radiology reports for abnormality and for small vessel disease presence. Of these reports, 80% (<span style="font-style:italic;">n; = 1630) were used to fine-tune pretrained transformer models to classify scans. Five-fold cross-validation was used in model development. The models were tested on the remaining 20% of the reports (<span style="font-style:italic;">n; = 408). SHapley Additive exPlanations (SHAP) were used to explain the results.Results;The models exhibited excellent classification performance, with a mean receiver operating characteristic (ROC) area under the curve (AUC) of 0.98 for abnormality classification and 0.99 for small vessel disease classification. SHAP highlighted relevant words in both cases.Conclusions;This application validated the use of a pretrained transformer in detecting confounding data in research cohorts, and exhibited explainable results that allow the models’ decisions to be understood. By highlighting the specific report terms that drive each prediction, the explainable model output can be reviewed and critiqued by subject matter experts, supporting trust, error analysis, and iterative refinement of AI tools within clinical workflows.Advances in knowledge;This application demonstrates the feasibility of explainable report classification, and the fine-tuned model could be used in future for automatic removal of confounding data from radiology datasets, while providing transparent, case-level justifications that support audit, governance, and clinician acceptance.;

Objectives;Dynamic MRI of the upper vocal tract is increasingly used to study speech. Image segmentation is often required to analyse the organs of speech; however, manual segmentation is labour intensive and time consuming and automatic methods are being developed. In this paper, a new hybrid transformer network is proposed for such task.Methods;We introduce a deep learning-based decoder model termed “Progressively Refinement Decoding (PRORED).” This model incorporates a directional field (DF) module designed to capture the contour details of features. The acquired contour information is leveraged to refine the boundaries both between and within classes. By integrating the DF module at different stages of the decoder, features are enhanced progressively, ensuring a more detailed and accurate segmentation.Results;Our model is evaluated using a publicly accessible speech MRI dataset and a cardiac dataset. The metrics employed are the Dice coefficient and the Hausdorff distance. Results indicate that our model attains an average Dice coefficient of 97.78% and a Hausdorff distance of 6.84 mm. Additionally, our network was able to identify closure patterns more efficiently than the baseline network and previously published work. In addition, the model was also evaluated on a cardiac dataset, and achieved 91.90% dice score.Conclusions;The proposed model leads to a more accurate segmentation of speech MRI data and in particular allows for a better velopharyngeal closure study. The proposed model was also evaluated on a cardiac dataset and achieved competitive performance, showing its strong generalizability.Advances in knowledge;First model that utilizes vision transformer and progressive refinement decoder to segment dynamic speech MRI.;