Generative AI voice agents—conversational systems powered by large language models that can understand and produce natural speech in real time—are poised to transform how health systems engage with patients. While technical and implementation challenges…
Continue Reading
Distinct metabolic susceptibilities (beta-cell dysfunction, insulin resistance (IR), and impaired incretin response) underlie type 2 diabetes (T2D). However, their relationships with habitual lifestyle behaviors are underexplored. This study integrated high-resolution lifestyle data from wearable…
Continue Reading
Sepsis accounts for a substantial portion of global deaths and healthcare costs. The objective of this reproducibility study is to validate Duke Health’s Sepsis Watch ML model, in a new community healthcare setting and…
Continue Reading
Embodied artificial intelligence (EAI) integrates perception, memory, reasoning and action through physical interaction, enabling multimodal dynamic learning and real-time feedback. In ophthalmology, EAI supports data-driven decision-making, improving the precision and personalization of diagnosis, surgery,…
Continue Reading
Findings from a previous study (ClinicalTrials.gov: NCT05118035) demonstrated that an AI-enabled electrocardiogram (AI-ECG), combining AI reports and physician alerts, effectively identified hospitalized patients at high risk of mortality and reduced all-cause mortality. This study…
Continue Reading
Oral pre-exposure prophylaxis (PrEP) effectively prevents HIV among men who have sex with men, but its adherence faces significant hindrances. We evaluated the effectiveness of real-time digital intervention in promoting oral PrEP adherence through…
Continue Reading
Thalassemia is an inherited blood disorder. Current diagnostic methods mainly rely on sophisticated equipment and specifically trained technicians. This study aims to identify and genotype thalassemia by applying machine learning (ML) algorithms to routine…
Continue Reading
Retinal age has emerged as a promising biomarker of aging, offering a non-invasive and accessible assessment tool. We developed a deep learning model to estimate retinal age with enhanced accuracy, leveraging retinal images from…
Continue Reading
Phenotypic information for cancer research is embedded in unstructured electronic health records (EHR), requiring effort to extract. Deep learning models can automate this but face scalability issues due to privacy concerns. We evaluated techniques…
Continue Reading
Large language models (LLMs) are used to seek health information. Guidelines for evidence-based health communication require the presentation of the best available evidence to support informed decision-making. We investigate the prompt-dependent guideline compliance of…
Continue Reading
