DIGITAL HEALTH最新文献

Objective: Keloids are benign fibrous dermal tumors that typically result in excessive scar tissue formation, affecting appearance and potentially causing discomfort. As social media platforms like TikTok become important sources of health information, the number of keloid-related videos has increased. However, the quality and reliability of these videos remain unclear. This study aims to evaluate the content, quality, and reliability of keloid-related videos on TikTok.

Methods: A total of 85 keloid-related videos on TikTok were analyzed. Video characteristics, uploader types, and content themes were extracted. The Global Quality Score and modified DISCERN tool were used to assess video quality and reliability. Correlation analysis was conducted between video metrics and quality scores.

Results: Videos were generally short (median: 48 s) with high engagement (median likes: 166, saves: 44). Common topics included treatment (87.06%), clinical manifestations (55.29%), and diagnosis (51.76%), while prevention, precipitating factors, and recurrence were less frequently discussed. Videos uploaded by healthcare professionals had significantly higher quality than those from individual users. Positive correlations were found among engagement metrics (likes, comments, saves), but no correlation was observed between engagement and video quality.

Conclusions: While keloid-related TikTok videos show high engagement, their overall quality and reliability are low. Increasing healthcare professional involvement and improving platform content regulation are essential to enhance the educational value of health information.

Objective: The study aims to develop and externally validate interpretable machine learning (ML) models for predicting 28-day mortality in critically ill patients with coexisting atrial fibrillation (AF) and acute kidney injury (AKI).

Methods: We conducted a retrospective analysis using two large public databases, Medical Information Mart for Intensive Care IV (MIMIC-IV) and eICU Collaborative Research Database (eICU-CRD). Critically ill adults with both AF and AKI were included. In MIMIC-IV, patients were randomly divided into a training set and an internal test set for model development and evaluation. Nine ML algorithms were compared, and the best-performing model was further validated in the external eICU-CRD cohort. Model performance was primarily assessed using the area under the receiver operating characteristic curve (AUC). Interpretability was examined with the SHapley Additive exPlanations (SHAP) method, and an online risk calculator was developed to support clinical application.

Results: A total of 11,510 patients from MIMIC-IV and 2565 patients from eICU-CRD were included. The GBM model achieved the best predictive performance, with an AUC of 0.856 (95% CI: 0.839-0.873) in the internal test cohort and 0.761 (95% CI: 0.740-0.783) in external validation. SHAP analysis identified anion gap, heart rate, and age as the most influential predictors of 28-day mortality. The developed online application enables individualized risk stratification, supporting clinical decision-making.

Conclusions: We developed and externally validated interpretable ML models for 28-day mortality prediction in ICU patients with AF and AKI. These models may enhance prognostic accuracy, facilitate earlier intervention, and support clinical management in this high-risk population.

Objectives: This study evaluated the feasibility of using the open-source large language model (LLM) DeepSeek-R1 to generate standardized Subjective, Objective, Assessment, and Plan (SOAP)-format medication logs and its potential to support clinical pharmacists.

Materials and methods: Thirty complete oncology medication profiles were collected, from which 80 days of logs were extracted and converted into simulated pharmacist-patient dialogues. The experiment compared single-information-source inputs (dialogue only) with multi-information-source inputs (dialogue plus patient information, records, and test results), using five prompts of increasing complexity. Performance was measured using the Bidirectional Encoder Representations from Transformers (BERT) score and Recall-Oriented Understudy for Gisting Evaluation (ROUGE), alongside a blinded expert evaluation based on the Seven-Dimension Index (7DI) metric.

Results: DeepSeek-R1 effectively generated structured SOAP medication logs when integrated with multi-source information and complex prompts (especially Prompts 4 and 5). Both machine scores and manual 7DI evaluation confirmed the superiority of multi-source inputs over single-source dialogues. While Prompt 4 achieved the highest BERT-F1 and ROUGE scores, the model's output quality remained highly dependent on input data completeness and required pharmacist review to correct errors (e.g. incomplete analyses), after which scores improved significantly.

Discussion: This study confirms DeepSeek-R1's utility in generating SOAP medication logs using multi-source data and structured prompts, potentially enhancing pharmacists' efficiency. Limitations such as oncology-specific scope and artificial intelligence (AI) hallucinations necessitate pharmacist review and future validation across specialties, alongside comparisons with closed-source LLMs and explainable AI integration.

Conclusion: This study demonstrates that DeepSeek-R1 can generate structured SOAP-format medication logs when guided by prompt-engineered multi-source clinical information, while highlighting that output quality depends on input completeness and that pharmacist review remains essential for clinical reliability.

Objective: Chronic Obstructive Pulmonary Disease (COPD) remains widely underdiagnosed in low- and middle-income countries due to reliance on costly, infrastructure-heavy diagnostic methods such as spirometry and radiographic imaging. This study aimed to design and validate a low-cost, artificial intelligence (AI)-powered Breath-Based Diagnostic (BBD) system for real-time COPD detection using exhaled volatile organic compounds (VOCs).

Methods: The BBD system integrates an array of metal oxide semiconductor sensors with a lightweight one-dimensional convolutional neural network deployed on a Raspberry Pi 5 for edge-based inference. Diagnostic performance was evaluated using a publicly available VOC dataset and a custom dataset collected under controlled conditions with the prototype device. Data augmentation strategies simulated sensor drift and environmental variability to improve model robustness. System performance was assessed in terms of accuracy, precision, latency, power efficiency, cost trade-offs, and usability. A multilingual mobile interface and Retrieval-Augmented Generation chatbot were developed to support patient engagement, while adherence to HIPAA and FHIR standards ensured regulatory compliance.

Results: The proposed system achieved 96.68% accuracy and 100% precision for COPD detection, with inference latency of 0.02 ms and power consumption below 2.5 W. A five-sensor configuration preserved 98% of diagnostic performance while reducing hardware cost by 30%. Usability testing with 31 participants yielded a System Usability Score of 86.3/100 and a chatbot trust rating of 4.4/5, confirming strong user acceptance.

Conclusion: The study demonstrates the feasibility of deploying an explainable, low-cost, and energy-efficient BBD system for early COPD detection in resource-limited settings. By combining edge AI, affordable sensor arrays, and multilingual patient engagement, the BBD system offers a scalable and ethically grounded pathway for integration into national healthcare infrastructures and global respiratory health strategies.

Objective: To evaluate the informational quality and user engagement of acne-related videos on Bilibili and TikTok, and examine associations with uploader characteristics, disease-related topics, presentation formats, and factors linked to high-quality content.

Methods: A cross-sectional analysis was conducted on 272 videos (122 from Bilibili, 150 from TikTok) retrieved in May 2025. Video characteristics, uploader types, disease-related topics, and presentation formats were recorded. Quality was assessed using the Journal of the American Medical Association benchmark criteria (JAMA), modified DISCERN instrument (mDISCERN), Global Quality Scale (GQS), and Video Information and Quality Index (VIQI). Engagement metrics (likes, comments, collections, shares) were analyzed. Correlation and predictive modeling were applied to examine associations between quality and engagement.

Results: Bilibili videos were longer (median 409 s vs 51 s; P < 0.001) and scored higher on VIQI (12.05 ± 2.94 vs 10.90 ± 1.97; P = 0.002). TikTok videos were more often uploaded by verified (68.10% vs 33.33%) and professional accounts (65.52% vs 25.64%), achieved higher JAMA (1.28 ± 0.45 vs 0.92 ± 0.98; P < 0.001) and mDISCERN scores (2.10 ± 0.52 vs 2.02 ± 0.93; P = 0.009), and demonstrated higher daily engagement. High-quality content was primarily produced by verified and professional uploaders, particularly in anatomy/physiology topics and doctor monologues. Official media, epidemiology, and television programs/documentaries achieved the greatest engagement. VIQI and GQS were strongly correlated (ρ = 0.755). VIQI (area under the curve [AUC] 0.922) and collections (AUC 0.901) were the strongest discriminators of high-quality content.

Conclusions: Acne-related videos on Bilibili and TikTok were generally of suboptimal quality. Bilibili favored coherence and accuracy, while TikTok favored transparency and engagement. Quality assessments outperformed engagement metrics in identifying high-quality content. These findings highlight the need to improve credentialing and promote engaging, evidence-based formats to enhance the reliability and impact of dermatologic information on short-video platforms.

Objective: We sought to evaluate the differential effects of wearable device-based interventions on weight-related and metabolic health outcomes among adults and youth with overweight or obesity.

Methods: We conducted a systematic review and meta-analysis of randomized controlled trials (RCTs) published up to November 2025, identified through PubMed, Embase, the Cochrane Library, and ClinicalTrials.gov. Eligible studies compared wearable device interventions to control conditions and reported outcomes such as weight, body mass index (BMI), waist circumference, blood pressure, lipid levels, and hemoglobin A1c (HbA1c). Data were pooled using random-effects models, and subgroup analysis was performed by age group.

Results: Eighteen RCTs were included. Wearable device interventions significantly reduced BMI overall (mean difference [MD], -0.63 kg/m²; 95% confidence interval [CI], -1.01 to -0.25; P = 0.001), with a greater effect in children and adolescents (MD, -0.91 kg/m²; 95% CI, -1.26 to -0.56; P < 0.00001). In adults, waist circumference decreased significantly (MD, -5.22 cm; 95% CI, -9.03 to -1.40; P = 0.007), and HbA1c also improved (MD, -0.24%; 95% CI, -0.30 to -0.18%; P < 0.00001). No significant differences were observed for overall weight change, blood pressure, or lipid profiles in adults. Pediatric participants showed more consistent improvements across multiple indicators.

Conclusion: Wearable device-based interventions led to modest but significant improvements in metabolic health, particularly in children and adolescents. These findings underscore the potential of wearable technologies as supportive tools for obesity management and highlight the importance of age-specific strategies in intervention design.

Background: Chimeric antigen receptor T-cell (CAR-T) therapy represents a transformative advancement in cancer treatment. While public interest in CAR-T has surged, particularly through short-video platforms like TikTok and Bilibili in China, concerns remain regarding the reliability and quality of health information disseminated through such media.

Objective: This study aimed to systematically evaluate the content quality, scientific integrity, and user engagement of CAR-T-related videos on Bilibili and TikTok, and to assess whether high traffic equates to high information quality.

Methods: A total of 200 Chinese-language videos (100 per platform) were identified using the keyword "CAR-T." Videos were evaluated using three scoring tools: the DISCERN instrument for reliability, the Global Quality Score (GQS), and a novel CAR-T-specific checklist assessing 12 core domains. Content characteristics, source types, and engagement metrics (likes, comments, shares, and saves) were also extracted and compared across platforms and content types.

Results: TikTok videos demonstrated significantly higher user engagement but poorer structure and lower DISCERN scores than Bilibili (P < 0.001). Videos posted by medical professionals were more common on TikTok (56%) and had higher engagement, but not necessarily higher quality. Bilibili, dominated by academic sources, produced longer videos with more complete and structured information. Correlation analysis revealed strong consistency among quality scoring tools but weak associations between quality and engagement metrics, suggesting a "high popularity-low quality" paradox.

Conclusion: CAR-T-related content on Chinese short-video platforms is characterized by a disconnect between popularity and information quality. Effective science communication strategies and platform-level interventions are needed to mitigate misinformation risks and improve the dissemination of high-quality medical content.

Objective: With the increasing global burden of cancer, there is a growing need for innovative strategies to improve oncology care. Health-related quality of life (HRQoL) is an outcome measure for assessing the overall wellbeing of patients with cancer. We used machine learning to predict HRQoL and to identify key factors that can inform patient-centered cancer care.

Methods: We conducted a cross-sectional study enrolling patients diagnosed with lung, breast, or colorectal cancer across two provinces in China. We collected data on demographics, clinical characteristics, and patient-centered features. HRQoL was assessed using the widely accepted EQ-5D-5L instrument in cancer care. We trained and evaluated seven machine learning models. SHapley Additive exPlanations (SHAP) analysis was employed to assess feature importance.

Results: Data from 924 patients with cancer were available. The random forest and extreme gradient boosting models had superior predictive performance. Positive SHAP values were primarily observed in patients with early-stage cancer and those enrolled in Urban Employees Basic Medical Insurance. Negative SHAP values were mainly associated with longer duration of chronic comorbidities, colorectal cancer, and ongoing chemotherapy. Age and time since cancer diagnosis exhibited bidirectional impacts.

Conclusions: Our study demonstrates the potential of machine learning models to predict HRQoL in patients with cancer. We identified key predictors of patient HRQoL, like duration of chronic comorbidities, early-stage cancer diagnosis, age, and health insurance coverage. Our findings would facilitate early identification of patients with lower HRQoL and promote the provision of patient-centered oncology care.

Objective: To assess the feasibility, acceptability, and preliminary effects of circadian rhythm for sleep (CRS), a mobile digital therapeutic that delivers closed-loop, wearable- and light-sensor-driven circadian coaching for insomnia.

Methods: Six-week, single-arm feasibility study in adults with short-term or chronic insomnia. CRS provided daily recommendations emphasizing stable wake-up time, morning light exposure, and daytime activity. Feasibility and acceptability outcomes were assessed (completion, passive-sensor data capture/adherence, satisfaction), and the primary clinical outcome (exploratory) was change in Insomnia Severity Index (ISI) from baseline to Week 6; the key secondary clinical outcome (exploratory) was Pittsburgh Sleep Quality Index (PSQI). Objective sleep-wake metrics from wearable device were explored.

Results: Twenty-three participants were enrolled; 20 completed the program (87.0%), and 16 comprised the prespecified analysis set based on data-fidelity criteria. Among these 16 participants, valid passive sensor data from the wearable and light sensors were captured on 88.6% of study days. ISI significantly improved from baseline to Week 6 (median 21.0 → 14.0; p < .001 by within-subject analysis), and PSQI improved (mean 10.9 → 7.7; p < .001; partial η² ≈ 0.50). Objective wearable metrics (total sleep time, time-in-bed, sleep onset, wake time) did not change significantly over time in this short pilot. Satisfaction was favorable (mean 37.9/45). No adverse events occurred.

Conclusions: CRS was feasible and acceptable, and was associated with within-subject improvements in subjective insomnia symptoms in this single-arm feasibility study; however, because there was no control group, these findings are preliminary and hypothesis-generating, supporting further evaluation in larger randomized controlled trials.

Trial registry name: Clinical Research Information Service.

Url: https://cris.nih.go.kr.

Trial registration number: KCT0010801.

Background: Mobile health (mHealth) applications are increasingly seen as essential components of healthcare digitalization, yet many national initiatives struggle to progress beyond pilot phases. In Czechia, the EZKarta mobile application was launched by the Ministry of Health as a secure digital gateway to vaccination records and preventive check-ups, marking a first step toward a national eHealth platform.

Objective: This study provides an exploratory, analytically grounded insight into early user perceptions and stakeholder views on EZKarta during its pilot phase, focusing on institutional, governance, and user-level factors influencing sustainability and integration.

Methods: A mixed-methods design was applied. Quantitative data from a national online survey (n = 209) were analyzed using nonparametric tests. The qualitative component included semistructured interviews with key stakeholders. Findings were integrated through joint interpretation and thematic triangulation.

Results: The mean usability score (UMUX = 33.2 ± 6.5) was significantly below the international benchmark (p < 2.2 × 10^-16). Only 38% of users reported satisfaction, while 72% indicated willingness to use the application if integrated with their provider's clinical system. Triangulation of survey and interview data suggests that low engagement was driven primarily by limited functionality, lack of clinical system integration, and unclear perceived added value. Stakeholders highlighted fragmented governance as key barriers, while recognizing EZKarta's potential role in national digital health coordination.

Conclusions: EZKarta exemplifies both the opportunities and constraints of mHealth adoption in transitional health systems. Stronger institutional coordination and transparent communication are essential for long-term relevance. The findings may inform policymakers in Central and Eastern Europe.