Frontiers in digital health最新文献

{"title":"Trustworthy intelligent rooms: integrating blockchain, federated learning, and data-centric AI for healthcare 4.0.","authors":"Ramesh Kumar Veerapaneni, Radhakrishnan Delhibabu","doi":"10.3389/fdgth.2026.1758304","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1758304","url":null,"abstract":"<p><strong>Introduction: </strong>Intelligent room systems are experiencing a surge in demand within the Healthcare 4.0 ecosystem. The integration of Federated Learning (FL) and Data-Centric AI has led to substantial enhancements in the predictive capabilities of machine learning models while maintaining data privacy. However, centralized aggregation in FL remains a single point of failure and is vulnerable to poisoning attacks.</p><p><strong>Methods: </strong>This paper presents a novel, privacy-preserving architecture for Ambient Intelligence (AmI) that integrates Distributed Ledger Technology (DLT).</p><p><strong>Results: </strong>We explicitly note that while DLT does not preemptively prevent the generation of poisoned gradients, it provides an immutable, cryptographically secure audit trail. This ensures the trustworthiness and traceability of model updates for post-hoc detection, strict accountability, and targeted model rollbacks.</p><p><strong>Discussion: </strong>By fusing Data-Centric AI for quality assurance with a Blockchain-enabled FL framework, we propose a scalable, low-cost solution for real-time patient monitoring in diverse economic settings.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1758304"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13143973/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Older adults' acceptability and perceived barriers to digital health tools integrating nutrition and physical activity: a focus group study.","authors":"Gioi Spinello, Erica Gobbi, Antonio Paoli, Tatiana Moro","doi":"10.3389/fdgth.2026.1803847","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1803847","url":null,"abstract":"<p><strong>Introduction: </strong>The rapid ageing of the population presents significant public health challenges, particularly in countries with high life expectancy such as Italy. Although nutrition and physical activity are key determinants of healthy ageing, many older adults do not meet recommended guidelines. Mobile health (mHealth) technologies may support healthy behaviors; however, evidence on older adults' perspectives remains limited, especially in the Italian context. This study aimed to explore experiences, perceptions, and expectations regarding mHealth tools for nutrition and physical activity.</p><p><strong>Methods: </strong>A total of three in-person focus groups were conducted with older adults in Italy, recruited regardless of prior experience with mobile health technologies. Data were analyzed using reflexive thematic analysis.</p><p><strong>Results: </strong>Reflexive thematic analysis generated three themes: digital health as a \"robotic friend\", digital health as an emotional barrier and digital health to increase awareness. The findings demonstrated that participants had good mHealth literacy. Nevertheless, they described digital health technologies as low in engagement and external motivation, highlighting the emotional distance, a strong preference for in-person interactions, and a general mistrust toward digital health. While there were some concerns related to privacy and fear of injury, older adults expressed their interest in digital tools as sources of guidance, education, and supervision.</p><p><strong>Discussion: </strong>Fully automated digital interventions may not meet the needs and preferences of this population. Findings suggest that hybrid models combining both digital technologies and human interaction might be more acceptable and feasible for promoting physical activity and healthy nutrition in later life.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1803847"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144103/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147847027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Construction of patient trajectories to model clinical trial outcomes: application to myasthenia gravis.","authors":"Marc Garbey, Quentin Lesport, Henry J Kaminski","doi":"10.3389/fdgth.2026.1755031","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1755031","url":null,"abstract":"<p><strong>Introduction: </strong>Accurate prediction of patient outcomes in clinical trials is crucial for the timely assessment of treatment efficacy. This study proposes a novel approach to predict patient response using longitudinal clinical data.</p><p><strong>Methods: </strong>We construct temporal trajectories from longitudinal data and extrapolate these trajectories to forecast individual patient outcomes. Additionally, we assess when new patients align with established response patterns. The approach is evaluated using data from the MGTX trial involving patients with myasthenia gravis.</p><p><strong>Results: </strong>Our analysis demonstrates the predictability of patient trajectories and enables automatic clustering of patients based on treatment success. The clustering reveals potential associations with age and smoking status.</p><p><strong>Discussion: </strong>These findings highlight the potential of trajectory-based methods for early prediction of treatment response in clinical trials. We also discuss possible confounding factors that may influence the observed associations and predictive performance.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1755031"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13143922/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846923","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Effectiveness of the mobile application Holidaily in reducing work-related rumination when returning to work after vacation: a randomized controlled trial.","authors":"Alexandra Thomsen, Christine Syrek, Hanna A Brückner, Jessica de Bloom, Monique Janneck, Markus Domin, Jo Annika Reins, Dirk Lehr","doi":"10.3389/fdgth.2026.1698339","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1698339","url":null,"abstract":"<p><strong>Background: </strong>Vacations reliably improve indicators of mental health, largely by providing relief from work-related stress. Low levels of work-related rumination, a key transdiagnostic factor linked to burnout and depression, are considered prerequisites for successful recovery both during vacations and in daily working life. However, such benefits are typically short-lived, with a rapid \"fade-out\" upon return to work. To address this challenge, we developed Holidaily, a low-threshold, gamified mobile health intervention designed to translate recovery science into daily digital practice and sustain the mental health gains of vacations.</p><p><strong>Methods: </strong>In a randomized controlled trial (RCT), Holidaily was evaluated as a digital mental health intervention targeting work-related rumination, the primary outcome. Assessments were conducted two weeks prior to vacation and two weeks after the return to work, before waitlist controls were granted access. Given the novelty of the research, a wide range of exploratory outcomes was also assessed.</p><p><strong>Results: </strong>A total of 190 workers from the general population were randomized to either the intervention (<i>n</i> = 91) or waitlist control group (<i>n</i> = 99). ANCOVA, in accordance with the intention-to-treat principle, indicated that the intervention group reported significantly lower levels of work-related rumination at two weeks post-vacation compared with controls [<i>p</i> < 0.001, <i>d</i> = -0.67 (-1.0; -0.4)]. At this time, rumination levels were still reduced by 22.2% in the intervention group, compared with 6.9% in controls relative to baseline. Among app users, reductions persisted for up to four weeks (26.1%). Sensitivity analyses confirmed these results. These findings provide first evidence that a mobile health technology can extend vacation-related recovery benefits and reduce work-related rumination in workers.</p><p><strong>Conclusions: </strong>This is the first RCT to show that the rapid fade-out of vacation benefits is not inevitable. Holidaily appears to improve workers' ability to reduce levels of work-related rumination. These results highlight the potential of scalable digital interventions to foster sustainable mental health in working populations and support preventive public health efforts.</p><p><strong>Clinical trial registration: </strong>https://drks.de/search/de/trial/DRKS00013650, German WHO DRKS00013650.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1698339"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144003/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846850","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A randomized factorial experiment to optimize the design of a culturally tailored breast cancer screening outreach chatbot intervention.","authors":"Raina Langevin, Pranuti Kalidindi, Katie Arriaga, Ryan P Kyle, Shayla Akande, Gary Hsieh, Leah M Marcotte","doi":"10.3389/fdgth.2026.1720531","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1720531","url":null,"abstract":"<p><strong>Introduction: </strong>The main objective of this study is to assess the effects of chatbot persona and communication style on trust and intention to use for scheduling breast cancer screening (BCS).</p><p><strong>Methods: </strong>We conducted a mixed-methods analysis of a randomized factorial experiment to evaluate different chatbot designs for a BCS intervention. The study protocol is registered on ClinicalTrials.gov (NCT05472064). We tested different conditions in a 2 × 2 experimental design using a Black woman persona presented either as a primary care doctor or a breast cancer survivor and a communication style that was either direct or polite, compared with a control condition.</p><p><strong>Results: </strong>Among the experimental conditions, the doctor-polite condition was the most preferred in terms of both trust and intention to use, compared with the control. Qualitative feedback indicated that the doctor persona and polite communication style were perceived as professional and friendly, respectively. While some participants appreciated representation in the use of a Black woman persona and found it relatable, others perceived it as stereotyping, patronizing, or targeting.</p><p><strong>Discussion: </strong>Overall, both quantitative and qualitative findings indicate that a culturally tailored doctor persona with polite messaging may enhance trust and increase intention to use the chatbot for scheduling BCS through professional interactions that are perceived as warm and friendly. The development of culturally tailored personas should be done with caution to prevent the perpetuation of stereotypes in chatbot persona development.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1720531"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144116/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846893","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"AI-driven mental health decision support linked to clinician resilience and preparedness.","authors":"Margareta-Theodora Mircea, Jessica McFadyen, Ross Harper, Max Rollwage, Tobias U Hauser","doi":"10.3389/fdgth.2026.1755085","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1755085","url":null,"abstract":"<p><strong>Objectives: </strong>Mental health services are facing unprecedented demand, placing significant pressure on clinicians to conduct timely and effective patient assessments. Rising staff turnover and burnout threatens service quality across many countries. This study examined whether providing clinical information, collected via an artificial intelligence (AI)-enabled decision support tool for mental health assessments in the UK's National Health Service (NHS), was associated with differences in clinician wellbeing and patient assessment performance.</p><p><strong>Method: </strong>In this observational study, we surveyed mental health clinicians (<i>N</i> = 131) from nine NHS Mental Health Talking Therapies services on how the information provided by an AI-based decision-support tool related to their experience with conducting clinical assessments. Clinicians reported on assessments where information from the AI tool was available, as well as when it was not (e.g., general practitioner referrals or telephone intakes). Outcomes included clinician wellbeing, task performance, and cognitive load during assessments, with additional analyses assessing the influence of moderating factors, such as clinician experience, workload, and exposure to the tool.</p><p><strong>Results: </strong>Relative to traditional methods, assessments supported by information provided by the AI tool were associated with significantly higher clinician wellbeing and task performance, and significantly lower cognitive load, irrespective of the clinician's experience. These associations were magnified by workload.</p><p><strong>Conclusion: </strong>These findings provide preliminary evidence that AI-powered pre-assessment tools may be associated with differences in clinician experience including higher wellbeing, higher task performance, and lower cognitive burden. By targeting systemic drivers of burnout, such tools may represent a potentially scalable approach to support workforce sustainability and service quality in mental health care.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1755085"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13143997/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence in fundus photography for type 2 diabetes: a scoping review of systemic biomarkers and multi-organ risk prediction.","authors":"Liting Huang, Hongyun Lu, Mingqi Yang, Yanyan Liu, Mini Han Wang, Kang Zhang","doi":"10.3389/fdgth.2026.1768780","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1768780","url":null,"abstract":"<p><p>Type 2 diabetes mellitus (T2DM) is associated with multi-organ complications, including cardiovascular and renal disease. Fundus photography provides a non-invasive window into systemic microvascular health, and artificial intelligence (AI) has enabled extraction of retinal biomarkers for systemic risk prediction beyond diabetic retinopathy detection. We conducted a methodologically structured scoping review following PRISMA-ScR guidance to map AI applications in retinal imaging for multi-organ risk stratification in T2DM. Studies using machine learning or deep learning models to predict cardiovascular, renal, or cerebrovascular outcomes were identified and characterized. Rather than quantitative pooling, we examined modeling strategies, validation approaches, performance reporting, and translational readiness across heterogeneous study designs. AI models frequently demonstrated promising discrimination; however, substantial heterogeneity was observed in cohort size, outcome definitions, imaging modalities, and validation strategies. External validation was limited, calibration was inconsistently assessed, and subgroup analyses addressing fairness and device-related domain shift were rarely reported. Most studies emphasized discrimination metrics without comprehensive evaluation of clinical utility.Retinal AI shows potential for scalable systemic risk surveillance in T2DM, but rigorous external validation, standardized reporting, and prospective implementation studies are required to enable safe and equitable clinical translation.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1768780"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144078/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846931","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Locally-deployed vs. cloud-based AI in healthcare: evaluating DeepSeek-R1:8b, DeepSeek-R1, and ChatGPT o3-mini-high for complex medical diagnostics.","authors":"Ning He, Lin Yang, Xinhong Hu, Yuanfang He","doi":"10.3389/fdgth.2026.1785443","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1785443","url":null,"abstract":"<p><p>Reasoning large language models are increasingly considered for healthcare-related artificial intelligence applications, but their practical value depends not only on diagnostic accuracy, but also on responsiveness and operational reliability. In this study, we benchmarked six model settings on 1,000 questions from the MedQA dataset: DeepSeek-R1, its distilled 8-billion-parameter local variant DeepSeek-R1:8b, ChatGPT o3-mini-high, and their knowledge-base-augmented counterparts. We evaluated performance across three dimensions: diagnostic accuracy, response latency, and first-attempt connection reliability. DeepSeek-R1 achieved the highest accuracy (89.5%, 95% CI: 87.4-91.2) but showed substantially longer response times (median 26.54 s) and higher connection failure rates (4.6%). ChatGPT o3-mini-high responded faster (median 10.05 s) and showed the most favorable tail-latency profile, but its accuracy (78.2%, 95% CI: 75.5-80.7) was lower than that of DeepSeek-R1. The locally deployed DeepSeek-R1:8b demonstrated markedly stronger connection reliability (failure rate 0.2%, 95% CI: 0.0%-0.5%) but substantially reduced accuracy (55.0%, 95% CI: 51.9%-58.5%). Knowledge-base augmentation did not consistently improve performance; for DeepSeek-R1, it significantly reduced accuracy by 4.36% ( <math><mi>p</mi> <mo>=</mo> <mn>0.0002</mn></math> ), while no significant benefit was observed for the other models. These findings show that reasoning model performance in medical question answering is best understood as a trade-off among accuracy, latency, connection reliability, and deployment mode, and that retrieval augmentation is not universally beneficial. More broadly, this study provides deployment-relevant benchmarking evidence for evaluating reasoning models in healthcare-related settings, while also indicating the need for richer knowledge resources and more realistic task environments before such systems can be meaningfully assessed for real-world clinical use.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1785443"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144068/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846952","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An embodied cognition based model of medical experts' tacit knowledge: structure, hierarchies, and transformation.","authors":"Hailing Zhou, Xinyue Chang, Xiaoyang Zhou, Jin Shi, Zuojian Zhou, Sheng Zhong","doi":"10.3389/fdgth.2026.1779044","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1779044","url":null,"abstract":"<p><strong>Background: </strong>Tacit knowledge plays a crucial role in clinical decision-making and medical innovation, particularly through experience-based and practice-oriented expertise. However, existing research has not yet provided a sufficiently integrated framework to explain how such knowledge is structured and transformed within medical practice.</p><p><strong>Methods: </strong>Grounded in embodied cognition theory, this study constructs a medical experts' tacit knowledge model encompassing four key elements of expert agent, context, thinking, and action. Building upon the layered perspective of the onion model, the study organizes tacit knowledge across three levels and explains its dynamic and bidirectional transformation.</p><p><strong>Results: </strong>The resulting framework integrates embodied experience, cognitive processes, and clinical practice into a coherent system. A case analysis of the acupuncture expert Wang Leting and his \"Lao Shi Zhen\" prescription is used to illustrate how the model operates in practice.</p><p><strong>Conclusion: </strong>The study provides a systematic perspective for understanding medical experts' tacit knowledge and offers theoretical insights for medical education, knowledge transmission, and clinical decision support.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1779044"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144038/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846871","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Big data integration for enhanced epidemiological research: insights and directions from NHLBI's workshop.","authors":"Md Mobashir Hasan Shandhi, Joseph Coresh, Jessilyn Dunn, Eric J Shiroma, Kenneth J Wilkins, Dana L Wolff-Hughes, Ruzhang Zhao, Yuling Hong, Gabriel Anaya","doi":"10.3389/fdgth.2026.1770258","DOIUrl":"https://doi.org/10.3389/fdgth.2026.1770258","url":null,"abstract":"<p><p>The landscape of epidemiological research is experiencing a technological transformation, driven by the rapid expansion of big data and advancements in artificial intelligence (AI) and machine learning (ML). This workshop explored the opportunities and challenges associated with integrating diverse data sources into population-based research at different levels, including electronic health records (EHRs), genomic and omics data, imaging, wearable device data, and social determinants of health measures, among others. AI/ML tools present powerful capabilities for analyzing these vast datasets, offering advancements in health risk prediction, disease pattern identification, and the development of personalized interventions. However, the integration of big data introduces technical barriers related to data heterogeneity, privacy and security concerns, and the potential to exacerbate health disparities through algorithmic biases. In September 2023, the National Institutes of Health's (NIH) National Heart, Lung, and Blood Institute (NHLBI), in collaboration with the National Cancer Institute (NCI) and the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), hosted a workshop to address these challenges and discuss the integration of big data into epidemiology and population-based studies. Key themes from the workshop emphasized interdisciplinary collaboration, data standardization, and the development of robust ethical frameworks, as well as the importance of advancing data governance, implementing transparent consent processes, and employing privacy-preserving techniques to maintain public trust. Additionally, the workshop highlighted the transformative potential of digital health technologies, such as wearable devices, which, when integrated with EHRs, enhance data granularity, facilitate early disease detection, and strengthen public health surveillance. Ethical, legal, and social issues (ELSI) are central to responsibly leveraging big data and AI in research, unbiased algorithms, the use of diverse datasets in AI training, and continuous human oversight to mitigate risk and ensure validity. The workshop also emphasized the need for workforce training and education in data science and bioinformatics to prepare researchers for utilizing these technologies effectively. The workshop concluded by recognizing the need for a balanced approach that addresses data integration challenges while harnessing AI/ML to improve healthcare outcomes. By fostering interdisciplinary collaboration, prioritizing privacy, and embracing data-driven methodologies, epidemiological research can unlock the full potential of big data to transform public health and clinical practice.</p>","PeriodicalId":73078,"journal":{"name":"Frontiers in digital health","volume":"8 ","pages":"1770258"},"PeriodicalIF":3.2,"publicationDate":"2026-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13144158/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147846942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}