Digital Health

Digital health encompasses the application of digital technologies across the healthcare continuum, transforming prevention, diagnosis, treatment and management of disease. Core components include electronic health records (EHRs) for longitudinal data capture, telemedicine and mobile health (mHealth) tools that remove geographic barriers, and wearable sensors that continuously monitor physiological signals. Artificial intelligence (AI) and big data analytics enable predictive modelling and personalised care pathways by identifying patterns in complex datasets. Large language models (LLMs) are emerging as conversational interfaces for patient education, clinical decision support and information retrieval. Citizen-generated data—from smartphones, fitness trackers and social media—augments traditional clinical records, fostering patient engagement and self-management. Point-of-care diagnostics and cloud-based platforms further decentralise care delivery, facilitating real-time interventions outside hospital settings. Collectively, these innovations drive a shift from reactive to proactive, preventive healthcare, emphasising individualised treatment while improving efficiency, equity and global accessibility.

Research from Nature Portfolio

Vision–language foundation model for echocardiogram interpretation (Nature Medicine, 2024) presents EchoCLIP, trained on over one million ultrasound videos and expert reports. Without task-specific training, EchoCLIP predicts ejection fraction with mean absolute error of 7.1%, identifies intracardiac devices with AUC up to 0.97, and supports long-context retrieval of patient transitions such as transplants, illustrating how foundation architectures can unify diverse cardiac imaging tasks.

Blinded, randomized trial of sonographer versus AI cardiac function assessment (Nature, 2023) reports a non-inferiority trial of AI-guided versus sonographer-guided left ventricular ejection fraction (LVEF) estimation. Among 3,495 echocardiograms, AI initial assessments led to fewer substantial cardiologist adjustments (16.8% vs 27.2%), reduced interpretation times and achieved superior agreement with final reads, demonstrating AI’s capacity to streamline workflows without sacrificing accuracy.

A formal validation of a deep learning-based automated workflow for the interpretation of the echocardiogram (Nature Communications, 2022) compares AI-driven measurements of 23 echocardiographic parameters—volumes, Doppler indices and ejection fraction—with core-lab sonographers. Using the individual equivalence coefficient, AI disagreement was lower than human inter-reader variability, validating automated quantification as at least as reproducible as expert measurements, while offering operational efficiencies.

Digital Health publication trend

The graph below shows the total number of articles in digital health across all publications each year (not limited to Nature Index journals).

Technical terms

Electronic Health Record (EHR): A digital version of a patient’s medical chart that consolidates health information over time, enabling data sharing among providers.

Telemedicine: Healthcare delivery and consultation conducted remotely via telecommunications technology, including video conferencing and remote monitoring.

Wearable technology: Devices worn on the body—such as smartwatches and fitness trackers—that continuously collect and transmit health-related data.

Point-of-care technology: Diagnostic or therapeutic tools used at or near the site of patient care to facilitate immediate clinical decision-making outside traditional laboratories.

Big data analytics: Techniques for processing and interpreting very large, complex datasets to uncover patterns, predict outcomes and support evidence-based care.

Large language model (LLM): An AI system trained on extensive text corpora to understand and generate human-like language, used for clinical documentation, patient interaction and information retrieval.

Citizen-generated data: Health information produced by individuals through personal devices, apps or social media, enhancing patient engagement and complementing clinical records.

References

1. Vision–language foundation model for echocardiogram interpretation. Nature Medicine (2024).
2. Blinded, randomized trial of sonographer versus AI cardiac function assessment. Nature (2023).
3. A formal validation of a deep learning-based automated workflow for the interpretation of the echocardiogram. Nature Communications (2022).
4. A remote digital memory composite to detect cognitive impairment in memory clinic samples in unsupervised settings using mobile devices. npj Digital Medicine (2024).
5. Efficacy of a Mobile Phone–Based Intervention on Health Behaviors and HIV/AIDS Treatment Management: Randomized Controlled Trial. Journal of Medical Internet Research (2023).
6. Virtual fitness buddy ecosystem: a mixed reality precision health physical activity intervention for children. npj Digital Medicine (2024).