DIGITAL HEALTH最新文献

Background: In light of the accelerated growth of China's digital economy and the ongoing enlargement of the older adults population obsessed with chronic ailments, the ramifications of the digital divide on the quality of life of urban older adults with chronic diseases have emerged as a significant concern. This study seeks to elucidate the manner in which the digital divide influences the quality of life of urban older adults with chronic diseases.

Method: This study employs the Chinese Older Adults Social Tracking Survey (CLASS2020) dataset and a multivariate ordinary least squares (OLS) regression model for statistical analysis. In this study, quality of life is employed as the dependent variable, while digital access and digital use of the digital divide are utilised as core explanatory variables. The pertinent indicators are weighted using the CRITIC weighting method to construct a composite indicator of quality of life and digital use.

Result: Study findings show that digital access boosts urban older adults' quality of life with chronic conditions, but overuse could reduce it. Social support mediates this relationship, with formal support being most effective. Accessibility of community services and facilities have a negative moderating effect on this influence: they decrease the positive impact of digital access and reduce the negative effects of digital use. Notably, Accessibility of community services enhances quality of life by offering care and activities, showing a "replenishment" effect. It is suggested that the government promote the inclusion of smart older adults care through state subsidies and tax credits.

Conclusion: This study highlights the importance of social support and community environment in improving the quality of life for older adults with chronic diseases in the digital economy. It suggests that policymakers should focus on reducing the digital divide and enhancing accessibility of community services to support these individuals. Additionally, the potential mental health issues arising from digital use should be recognized, requiring appropriate digital education and psychological support for older adults.

Background: Artificial intelligence (AI) chatbots are increasingly used for health information dissemination. However, their effectiveness depends on the clarity, reliability, and quality of the content they deliver. This cross-sectional study aimed to evaluate the readability and reliability of kyphosis-related information provided by six major AI chatbots: ChatGPT, Gemini, Copilot, Perplexity, DeepSeek, and Grok.

Methods: We selected the top 10 kyphosis-related questions from Google's "People also ask" section and submitted them to each chatbot. Readability was assessed using FKGL, FKRS, GFOG, SMOG, CL, ARI, and LW indices. Quality and reliability were evaluated using the DISCERN tool, JAMA benchmark, Global Quality Score (GQS), Ensuring Quality Information for Patients (EQIP), and a kyphosis-specific content score (KSC). Statistical analyses were performed using the Kruskal-Wallis and Mann-Whitney U tests.

Results: No statistically significant difference was found among chatbots in FKGL, FKRS, SMOG, ARI, or GFOG scores. However, Perplexity had significantly higher DISCERN and EQIP scores, indicating superior content quality. All chatbots presented content at a readability level higher than the AMA-recommended sixth-grade level. While AI models provided more comprehensive and up-to-date information than traditional web sources, their outputs remained challenging for the average patient to comprehend.

Conclusions: AI chatbots offer promising tools for disseminating health information about kyphosis but require significant improvements in readability. Expert-reviewed and patient-centered refinements are necessary to ensure accessibility and safety in digital health communication.

Objective: Early and accurate detection of ear diseases is essential for preventing hearing impairment and improving population health. This study aimed to develop a lightweight, high-performance, and real-time deep learning model for otoscopic image classification and to deploy it in a cross-platform diagnostic system for clinical and community use.

Methods: We constructed a large-scale, multi-center otoscopy dataset covering eight common ear diseases and healthy cases. Based on this dataset, we developed Best-EarNet, an ultrafast lightweight architecture integrating a local-global spatial feature fusion module and a multi-scale supervision strategy to enhance feature representation. Transfer learning was applied to optimize performance. The model was evaluated on internal (22,581 images) and external (1,652 images) test sets, with subgroup analyses by age and gender. Grad-CAM visualizations were used to improve interpretability. A cross-platform intelligent diagnostic system, Ear-Keeper, was further developed for deployment on smartphones, tablets, and personal computers.

Results: Best-EarNet achieved accuracies of 95.23% on the internal test set and 92.14% on the external test set, with a model size of 2.94 MB. It processed images at 80 frames per second on a standard CPU. Subgroup analyses demonstrated consistently high performance across age and gender groups. Grad-CAM visualizations highlighted lesion-related regions, and Ear-Keeper enabled real-time video-based ear screening across multiple platforms.

Conclusion: Best-EarNet offers an accurate, efficient, and interpretable solution for ear disease classification. Its real-time performance and cross-platform deployment through Ear-Keeper support both clinical practice and community-level screening, with strong potential for early detection and intervention.

Background: Bloodstream infection (BSI) contributed significant mortality among patients in the intensive care unit (ICU). Traditional machine learning (ML) models often struggle to effectively capture complex temporal dependencies in high-dimensional data. The aim of this study was to develop a deep learning model transformer in the prediction of ICU-acquired BSI based on time series data.

Methods: Patients' electronic health records, whose all blood cultures (BC) collected 48 h after admission to the ICU, were extracted from Medical Information Mart for Intensive Care IV (MIMIC IV). The synthetic minority over-sampling technique (SMOTE) was applied to balance the dataset. We collected age, gender, vital signs and laboratory measures for consecutive 24 h with 1 hour interval. We also set three prediction windows (0, 12 and 24 h) to investigate the ability of early detection of the ML. The performances of the transformer and the CatBoost were evaluated by discrimination and calibration. Shapley Additive exPlanation (SHAP) was employed to identify key features.

Results: A total of 2408 patients were included in the study, of which 149 (6.2%) had an ICU-acquired BSI. The transformer model outperformed CatBoost at all prediction windows. At the 24-hour window, the Transformer achieved an AUROC of 0.918 and an AUPRC of 0.915, while CatBoost performance declined significantly with earlier prediction. SHAP values suggested that glucose, bicarbonate, mean blood pressure, temperature and blood urea nitrogen were top five early predictors.

Conclusion: The deep learning transformer using time series data demonstrates strong potential as a clinical decision support tool.

Objective: The rapid global aging trend has led to a substantial increase in the prevalence of frailty among older adults. We developed a mobile app-based multidomain intervention (MI) program as part of a South Korean study to prevent the progression of frailty and aging-related diseases using a digital MI (SUPERAGING). We aim to evaluate the feasibility of the SUPERAGING app-based intervention in prefrail or frail older adults.

Methods: We will recruit 40 community-dwelling older adults aged 60 to 90 years classified as frail or prefrail according to the Modified Fried frailty criteria. Participants will be randomly assigned to intervention and control groups at a 1:1 ratio. The intervention group will receive a personalized program through the SUPERAGING digital platform, comprising four components: disease management, cognitive training, physical exercise and nutritional intervention for 16 weeks. The control group will receive standard lifestyle education only. The primary outcomes are adherence, retention, and recruitment rates. The main secondary outcomes are frailty, disability, cognitive function, physical performance, nutritional assessment, mood, quality of life, vascular risk factors, and occurrence of aging-related diseases. There will be an exploratory evaluation of biological aging markers.

Results: The intervention program will be considered feasible if the following success criteria are met: (a) a retention rate of 70% or higher, (b) an adherence rate of 70% or higher, and (c) a recruitment rate of 50% or higher.

Conclusions: The results will provide information on the applicability of a MI using a mobile app targeting older adults with prefrailty or frailty.

Trial registration: ClinicalTrials.gov identifier: NCT06891573. Registered on February 25, 2025.

Introduction: Sexual and reproductive health (SRH) services remain inaccessible for many young adults in low-resource settings. Mobile health (mHealth) technologies present an innovative solution, yet adoption rates among university students in sub-Saharan Africa are poorly characterized. This study investigated the knowledge, utilization patterns, and key determinants of mHealth adoption for SRH services among Ghanaian university students using established behavioral frameworks.

Methods: We conducted a cross-sectional study of 390 undergraduates at Kwame Nkrumah University of Science and Technology, selected through multistage sampling across six residence halls. A validated questionnaire, incorporating constructs from the Unified Theory of Acceptance and Use of Technology and Health Belief Model, assessed: (1) mHealth knowledge, (2) SRH service utilization, and (3) psychosocial determinants of adoption. Confirmatory factor analysis validated measurement scales, while hierarchical regression analyzed predictors across three models (demographic, behavioral, and interaction terms).

Results: The study revealed three key findings: First, while 73.8% of respondents demonstrated mHealth awareness, actual usage for SRH remained low (mean = 3.40/7, SD = 1.83). Second, performance expectancy (β = 0.345, p < 0.001) and life quality improvements (β = 0.197, p < 0.05) significantly predicted adoption. Third, substantial barriers included social stigma (β = -0.210, p = 0.016), resistance to technology adoption (β = -0.221, p = 0.001), and privacy concerns (β = -0.208, p = 0.004). Notably, perceived health threats did not moderate the relationship between performance expectations and usage (p = 0.488).

Conclusions: This study highlights a critical gap between awareness and actual use of mHealth for SRH among Ghanaian university students. It identifies key barriers-such as perceived usefulness, privacy concerns, and behavioral resistance-and reveals how sociocultural norms influence youth engagement with digital health tools. The findings underscore the need for culturally sensitive, trust-based mHealth initiatives that strengthen digital literacy and promote behavioral change. Such approaches could enhance youth-focused mHealth strategies in Ghana and offer valuable lessons for similar contexts across sub-Saharan Africa.

Objective: This study aims to systematically review the current literature on the application of machine learning to predict return-to-sport (RTS) decisions after athletic injuries. The review focuses on identifying the types of machine learning models used, the commonly used predictive variables, and the methodological characteristics and limitations between studies in terms of design, model development, evaluation, and reporting.

Method: A comprehensive literature search was conducted on 1 May 2025 in three electronic databases: Web of Science, PubMed, and SPORTDiscus (EBSCO). Two independent reviewers selected the retrieved studies based on predefined inclusion and exclusion criteria. The Prediction Model Risk of Bias Assessment Tool (PROBAST) was used to assess the risk of bias in the included prognostic modeling studies.

Results: Of the 56 studies initially identified, 11 met the inclusion and exclusion criteria. Knee injuries were the most frequently modeled injury type for RTS decision-making (n = 4). The area under the receiver operating characteristic curve (ROC AUC) was the most commonly reported performance metric, presented in 82% of the included studies. Random Forest (RF) was the most widely used machine learning algorithm, applied in six studies (55%), and demonstrated the best predictive performance in four of them, with two studies reporting an AUC greater than 0.9. Some studies employed feature importance analysis or interpretability methods (e.g. SHAP) to identify key predictive variables. However, challenges remain in translating these models into clinical practice.

Conclusions: Machine learning techniques demonstrate promising potential for predicting RTS in athletes. Nevertheless, substantial heterogeneity across studies-particularly in RTS definitions, feature selection, and model development which limits the generalizability and clinical applicability of current models.

Objective: Investigating progression risk insights of high-grade gliomas through habitat radiomics analysis.

Methods: A cohort of 89 patients with high-grade gliomas was enrolled, with 63 patients in the train cohort and 26 patients in the test cohort. The methodology involved delineating the region of interest (ROI) within the tumor area on magnetic resonance imaging images, followed by multisequence registration. The ROI was further divided into subregions using K-means clustering. Radiomics features were extracted from the subregions, and feature selection was performed using least absolute shrinkage and selection operator regression. Four separate models were created: radiomics model, clinical model, habitat model, and combined model that merged the habitat signature with clinical factors. The accuracy of the models was evaluated using concordance index (C-index) and receiver operating characteristic analysis.

Results: The ROI was divided into three subregions, from which 36 features were extracted and selected. The habitat model, radiomics model, clinical model, and combined model were constructed by combining the extracted features with clinical data. The habitat model showed excellent predictive performance with the C-index values of 0.879 in the train cohort and 0.781 in the test cohort. Using this model, patients were classified into high-risk and low-risk groups, resulting in significantly different median progression-free survival (mPFS) times of 7 and 31 months, respectively (P < 0.001). Stratifying the patient cohort based on isocitrate dehydrogenase status also revealed distinct mPFS outcomes for high-risk and low-risk patients.

Conclusion: The habitat model demonstrated outstanding predictive performance for forecasting the progression risk of patients with high-grade gliomas.

Objectives: Social media use has expanded rapidly across healthcare, creating opportunities for professional networking, patient engagement, and education. However, difficulties have arisen in workplace support of social media use as opinions on professionalism differ among generations. There is no consensus on how social media professionalism is defined and enforced. Our objective is to examine healthcare professionals' perceptions of professional versus unprofessional social media behaviors, explore generational differences, and identify gaps in training and institutional guidelines.

Methods: We conducted a cross-sectional, quantitative survey at a large academic medical center. The survey, developed and pilot-tested for clarity and content validity, was distributed via institutional listservs to medical students, nurses, advanced practice providers, and physicians across multiple specialties. Respondents reported demographics, social media usage patterns, account types, and views on appropriateness of specific content for private versus public accounts. Bivariate analyses using chi-square and Fisher's exact tests were carried out to compare responses across age groups.

Results: Of 389 consenting participants, 260 (67%) completed the survey. Most respondents were female (75.4%) and White (75.2%). Facebook and Instagram were the most common platforms (51.5% each). Generational differences emerged: younger participants (18-40 years) were more permissive of personal and lifestyle posts, while older groups (41-60 and 60+) were more restrictive. Most participants (84%) reported never receiving formal training on social media professionalism, though 69.8% endorsed its value, particularly at the medical school level. Awareness of standardized institutional guidelines was limited (43%).

Conclusions: Generational differences influence perceptions of social media professionalism in healthcare. Despite widespread social media use, formal training and guideline awareness remain limited. These findings support the need for structured education on digital professionalism and clearer institutional policies to balance personal expression with professional standards.

Objective: Herpes zoster (HZ) is a common viral skin disease. As short video platforms have become an important source of health information, evaluating the quality of HZ-related content is increasingly important.

Methods: This cross-sectional study analyzed 186 HZ-related videos collected from the Chinese short video platforms TikTok (n = 96) and Bilibili (n = 90) from 12 to 16 July 2025. Video characteristics (duration, likes, shares, comments, and collections), content domains (etiology, clinical manifestations, diagnosis, treatment, and prognosis), and uploader identity were recorded. Quality was evaluated using the Global Quality Score (GQS) and modified DISCERN (mDISCERN).

Results: TikTok videos had shorter durations (47.5 vs.196.0 seconds) but significantly higher engagement metrics, including likes (481.5 vs. 29.5), comments (20.5 vs. 3.5), and shares (152.5 vs. 19.0). Most videos were uploaded by pain medicine specialists (31.72%), followed by traditional Chinese medicine physicians (27.96%) and dermatologists (19.35%). Treatment was the most frequently addressed topic (67.74%), whereas prognosis was least discussed (18.28%). The median GQS was 2.00 (interquartile range (IQR): 1.00-3.00) and the mDISCERN score was 3.00 (IQR 1.00-3.00). Dermatologists achieved higher GQS and mDISCERN scores than patients and traditional Chinese medicine physicians (p < 0.05). No significant correlations were observed between engagement metrics and GQS scores (p > 0.05).

Conclusions: HZ-related videos on short video platforms are of moderate quality and skewed toward treatment. Dermatologists involvement and better content regulation are needed. More emphasis should be placed on prognosis education to enhance public understanding.