Frontiers in digital health最新文献

Introduction: Adherence to weekly iron/folic acid (IFA) supplementation, a vital intervention to combat anaemia among adolescent girls, remains a global challenge, including in Maluku Province, Indonesia. This study assessed the effect of "Jaga Diri" application, in enhancing knowledge and adherence to IFA supplementation among adolescent girls from Salahutu Sub-District of Maluku Province, Indonesia.

Methods: In mid-2024, a quasi-experimental study was conducted among 82 adolescent girls from two senior high schools in Salahutu Sub-District, Maluku Province, Indonesia. The intervention group used the "Jaga Diri" Android-based application for four weeks, which delivered weekly reminders and brief educational messages on anaemia and iron-folic acid (IFA) supplementation, while the control group received routine school-based services. Knowledge was measured using a validated 15-item questionnaire. Adherence was defined as consumption of ≥75% of the provided weekly IFA tablets over the previous four weeks, assessed by self-report, and supported by haemoglobin measurement. Group differences were analyzed using non-parametric and chi-square tests, and multivariable binary logistic regression was used to assess factors associated with high knowledge and adherence.

Results: After four weeks of using the "Jaga Diri" application, adolescent girls from the intervention school showed a significantly higher level of knowledge about anaemia (p = 0.011) and adherence to weekly IFA supplementation (p < 0.001) than those from the control school. The improved adherence was shown by the reduction of anaemia prevalence in the intervention school, from 35% to 17.5%. In the control school, the prevalence increased from 19% to 28.6%.

Conclusions: The "Jaga Diri" application effectively improves knowledge about anaemia and adherence to IFA supplementation among adolescent girls. Further investigation with larger and more varied groups are required to confirm its effectiveness before it can be widely implemented in larger areas of Maluku and Indonesia.

Background: Point-of-care lung ultrasound (LUS) has been described for the evaluation of lung pathologies such as pneumothorax, pneumonia, and COVID-19 infections. It is rapidly deployed, portable, and accurate for LUS diagnoses. However, a learning curve limits its use, and teleguidance has been proposed as a solution. In this study, we primarily seek to measure the effect of tele-guided lung ultrasound (T-LUS) on chest X-ray (CXR) utilization in patients presenting with COVID-19 symptoms. Secondarily, we measure the effect of T-LUS on clinical decision-making, length of stay, and clinical outcomes.

Results: We performed a retrospective observational study using a before-after design in an adult urgent care (AUC) setting. A total of 303 patients with symptoms suggestive of COVID-19 were included. AUC providers used T-LUS on 31% of patients with COVID-19 symptoms (n = 34). Abnormal LUS findings were found in 41% of patients (n = 14), with B-lines (86%) and pleural irregularities (79%) being the most common findings. Among all patients in the study period, those who received a T-LUS did not show a statistically significant difference in CXR utilization [-12% difference; 95% confidence interval (CI) -25% to 5%] as compared to patients who did not receive a T-LUS, and a similarly non-significant difference was observed in the intervention period (-5% difference; 95% CI: -21% to 14%). Length of stay was longer for patients in whom T-LUS was used (median difference 26 min, 95% CI 11-41). However, a comparison of patients in the intervention period revealed no significant difference in length of stay between patients who received T-LUS and those that did not (median difference 16 min, 95% CI -5 to 37).

Conclusion: T-LUS is feasible and alters clinical decision-making for novice ultrasound users in the care of patients with suspected COVID-19 infection. Our results indicated that there was a no statistically significant difference trend in CXR utilization and no improvement in length of stay by the end of the 2-week trial.

Advances in artificial intelligence and multi-omic analysis are transforming fetal medicine from a diagnostic discipline into a predictive one. Yet the legal, deontological, and ethical frameworks that govern prenatal and fetal data have not evolved accordingly. Current regulations protect the mother as a patient but do not recognize the fetus-or the future child-as a legal data subject. As a result, information generated before birth remains confined within maternal medical records, creating uncertainty about who may later access or reuse it. This paper examines the emerging ethical and legal challenges of predictive fetal medicine, focusing on the transition from maternal consent to the child's future right to their own prenatal data. Through the lens of professional deontology and comparative law, we analyze the tensions between confidentiality, autonomy, and beneficence. We propose a framework of prenatal data stewardship, shifting from static notions of data ownership to shared responsibility across time. Establishing national or international repositories under transparent governance could enable ethical reuse of fetal data while safeguarding maternal privacy and ensuring future individuals' rights. Ultimately, aligning predictive fetal medicine with ethical and legal coherence requires collective action among clinicians, ethicists, jurists, policymakers, and industry. Only through such stewardship can information generated before birth become a trusted tool for care rather than control.

Background: The rapid evolution of general large language models (LLMs) provides a promising framework for integrating artificial intelligence into medical practice. While these models are capable of generating medically relevant language, their application in evidence inference in clinical scenarios may pose potential challenges. This study employs empirical experiments to analyze the capability boundaries of current general-purpose LLMs within evidence-based medicine (EBM) tasks, and provides a philosophical reflection on their limitations.

Methods: This study evaluates the performance of three general-purpose LLMs, including ChatGPT, DeepSeek, and Gemini, when directly applied to core tasks of EBM. The models were tested in a baseline, unassisted setting, without task-specific fine-tuning, external evidence retrieval, or embedded prompting frameworks. Two clinical scenarios, namely SGLT2 inhibitors for heart failure and PD-1/PD-L1 inhibitors for advanced NSCLC were used to assess performance in evidence generation, evidence synthesis, and clinical judgment. Model outputs were evaluated using a multidimensional rubric. The empirical results were analyzed from an epistemological perspective.

Results: Experiments show that the evaluated general-purpose LLMs can produce syntactically coherent and medically plausible outputs in core evidence-related tasks. However, under current architectures and baseline deployment conditions, several limitations remain, including imperfect accuracy in numerical extraction and processing, limited verifiability of cited sources, inconsistent methodological rigor in synthesis, and weak attribution of clinical responsibility in recommendations. Building on these empirical patterns, the philosophical analysis reveals three potential risks in this testing setting, including disembodiment, deinstitutionalization, and depragmatization.

Conclusions: This study suggests that directly applying general-purpose LLMs to clinical evidence tasks entails some limitations. Under current architectures, these systems lack embodied engagement with clinical phenomena, do not participate in institutional evaluative norms, and cannot assume responsibility for reasoning. These findings provide a directional compass for future medical AI, including ground outputs in real-world data, integrate deployment into clinical workflows with oversight, and design human-AI collaboration with clear responsibility.

Trauma-focused mobile health (mHealth) applications, such as PTSD-Coach, hold significant potential to address acute psychological needs following large-scale emergencies, yet adoption remains inconsistent. This study examined associations between psychosocial resources and intention to adopt the Hebrew version of PTSD-Coach in Israel after the October 7, 2023, terror attack, which triggered widespread collective trauma and ongoing war. Survey data from Israeli adults (n = 86) measured trauma literacy, self-efficacy, citizenship (willingness to share/recommend), and adoption intention. Quantitative analyses using multivariable regression identified a sequential pathway: trauma literacy enabled users to recognize symptom relevance, self-efficacy converted knowledge into capability, and citizenship extended adoption intentions into social spaces. Trauma literacy was the only significant predictor of intention in the full model, while demographic and clinical variables including trauma exposure were non-significant. Self-efficacy strongly predicted willingness to recommend the app, and once self-efficacy was included, the direct effect of knowledge diminished, supporting a sequential process: Knowledge → Self-efficacy → Citizenship → Intention. Rooted in social psychiatry and trauma-informed public mental health perspectives, this study theoretically interprets how individual psychological resources and social dynamics may shape early digital help-seeking in crisis conditions. Findings suggest that trauma literacy and perceived capability are central correlates of adoption readiness, challenging assumptions that lived trauma experience automatically increases help-seeking. This pattern may reflect how acute stress impairs information uptake and perceived self-efficacy. From a mental health systems perspective, these findings point to the potential importance of proactive psychoeducation, stigma-reduction strategies, and community-based outreach to support digital intervention uptake during collective trauma. Strengthening trauma literacy and self-efficacy may support timely self-management, help-seeking, and community resilience where formal psychiatric services are strained or inaccessible.

Introduction: This study describes the user-centred design and evaluation of AreAreaAlzheimer, a web-based digital platform developed to support family caregivers of individuals with dementia, especially Alzheimer disease. The initiative sought to ensure that technological solutions effectively address caregivers' actual needs through active user participation at every stage of development.

Methods: Following an iterative, participatory design approach, 419 individuals contributed to the project. The first phase combined a survey of 210 caregivers and focus groups with 22 participants to identify priority support dimensions. Thematic analysis highlighted four main areas: informational guidance, logistical assistance, emotional and communication strategies, and peer social connection. Based on these insights, 147 additional participants provided feedback that refined platform features and content. Finally, platform evaluation included standardized usability measures including the Single Ease Question (SEQ) for task difficulty, the System Usability Scale (SUS) for overall usability perception, the Perceived Usefulness Scale (PUS) completed by 40 caregivers, and scenario-based testing with 19 users who discussed experiences and improvement opportunities.

Results: Quantitative findings showed high ratings in accessibility (average score: 4.5/5), usability (scored 74.3/100), and perceived usefulness was rated lower (average score: 3.4/5). Qualitative feedback supported these results, emphasizing the platform's practical value in everyday caregiving. However, participants with lower digital literacy reported persistent challenges, indicating the need for simplified navigation and adaptive interface features.

Discussion: AreAlzheimer demonstrates the potential of participatory design to create inclusive, effective digital health tools for dementia care. Involving caregivers and people living with dementia enriched the design, promoting autonomy and cognitive sensitivity. Future research will integrate these insights into formal scientific protocols to expand participatory digital health innovations in dementia support.

Introduction: Colorectal cancer frequently leads to liver metastases (CRLM), posing a major challenge to long-term survival. Prognosis remains heterogeneous, and traditional clinical risk scores often lack biological depth and spatial information. Advances in radiomics and machine learning (ML) offer the potential for improved, explainable outcome prediction; however, robust and interpretable prognostic models for CRLM remain an unmet need. This study aimed to develop and validate explainable ML models based on radiomic features extracted from both metastatic lesions and background liver tissue, enhancing the prediction of recurrence and overall survival (OS) status in patients with CRLM.

Materials and methods: Patient data and contrast-enhanced CT images from two independent cohorts were analysed: a publicly available TCIA-CRLM series, employed as the discovery set, and a real-life clinical cohort, used as an external validation set. Segmentation focused on the largest liver metastasis (L-MAX) and surrounding healthy liver tissue (L-BKG), extracting radiomic features from both areas and their ratios (L-MAX/L-BKG). An end-to-end pipeline for data preprocessing and classification was designed. Multiple ML and Deep Learning (DL) classifiers were trained and validated. Model interpretability was assessed using SHapley Additive exPlanations (SHAP) analysis to identify key predictive radiomic determinants. Performances were compared to recognized clinical models.

Results: For recurrence prediction, the best-performing classifier was a soft-voting ensemble of a multilayer perceptron (MLP) optimized via a Genetic Algorithm (GA); for OS status classification, the best performance was obtained by a hard-voting ensemble of a GA-optimized MLP. Both classifiers demonstrated robust discrimination capabilities in external validation, with AUCs of 0.78 and 0.68, respectively. The explainability analysis performed with SHAP revealed the most relevant radiomic determinants in the classification. These features retained prognostic significance in the independent cohort, supporting their use for clinical risk stratification.

Discussion: Explainable ML models leveraging both lesion-centric and contextual liver radiomics offer clinically transparent prediction of recurrence and survival in CRLM. SHAP highlighted clinically plausible, reproducible imaging determinants, enabling risk stratification. The validation of specific radiomic determinants suggests the potential practical utility of this approach, laying out the groundwork for integrating with DL and multi-omic data in future oncology strategies.

Introduction: Interstitial/alveolar syndrome (IS) is a condition detectable on lung ultrasound (LUS) that indicates underlying pulmonary or cardiac diseases associated with significant morbidity and increased mortality rates. However, diagnosing IS using LUS can be challenging and time-consuming, and it requires clinical expertise.

Methods: In this study, multiple convolutional neural network (CNN) models were trained as binary classifiers to accurately screen for IS in LUS frames by distinguishing between IS-present and healthy cases. The CNN models were initially pre-trained using a generic image dataset to learn general visual features (ImageNet) and then fine-tuned on our specific dataset of 108 LUS clips from 54 patients (27 healthy and 27 with IS, two clips per patient) to perform a binary classification task. Each clip in the dataset was assessed by a clinical sonographer to determine the presence of IS features or confirm healthy lung status. The dataset was split into training (70%), validation (15%), and testing (15%) sets.

Results: Following the process of fine-tuning, we successfully extracted features from pre-trained DL models. These extracted features were then utilised to train multiple machine learning (ML) classifiers, resulting in significantly improved accuracy in IS classification compared with the individual CNN models. Advanced visual interpretation techniques such as heatmaps based on gradient-weighted class activation mapping (Grad-CAM) and local interpretable model-agnostic explanations (LIME) were implemented to further analyse the outcomes. The best-trained ML model achieved a test accuracy rate of 98.2%, with specificity, recall, precision, and F1 score values above 97.9%.

Conclusion: Our study demonstrates the feasibility of using a pre-trained CNN as a diagnostic tool for IS screening on LUS frames, integrating targeted data filtering, feature extraction, and fusion techniques. The data-filtering technique refines the training dataset by excluding LUS frames that lack IS-related features (e.g., absence of B-lines). Feature fusion combines features learnt from different models or "fused" to enhance overall predictive performance. This study confirms the practicality of using pre-trained CNN models with feature extraction and fusion techniques for screening IS using LUS frames. This represents a noteworthy advancement in improving the efficiency of diagnosis. In the next steps, validation on larger datasets will assess the applicability and robustness of these CNN models in more complex clinical settings.

Background: Prediction models for Type 2 Diabetes Mellitus (T2DM) often rely on biochemical markers such as glycated hemoglobin, fasting glucose, or lipid profiles. While clinically informative, these indicators typically reflect established dysglycemia, limiting their value for early prevention. In contrast, psychosocial stress, sleep disturbance, tobacco use, and dietary quality represent modifiable, non-clinical factors that can be observed long before metabolic abnormalities are clinically detectable. Yet most studies examine these factors in isolation or as additive lifestyle scores, overlooking how their interdependencies reorganize in the preclinical phase. A systems-level approach is therefore needed to capture how disruptions in behavioral coherence signal emerging vulnerability.

Methods: This study develops a dual-analytic framework that integrates Cox proportional hazards models with artificial neural network (ANN) coherence analysis. Using longitudinal data from the UK Biobank (n=15,774; follow-up up to 17 years), we identified non-clinical predictors of incident T2DM and examined how behavioral networks reorganize across health states. Predictors were screened through multivariate survival analysis and mapped into ANN-derived influence matrices to quantify stability, direction, and systemic coherence of relationships among diet, sleep, psychosocial states, and demographics.

Results: Eighteen significant predictors of T2DM onset were identified. Elevated risk was linked to loneliness, psychiatric consultation, emotional distress, insomnia, irregular sleep, tobacco use, and high intake of processed meat, beef, and refined grains. Protective effects were observed for 7-8 h of sleep, oat and muesli consumption, and fermented dairy. ANN analyses revealed a pronounced breakdown of behavioral coherence in T2DM: foods that stabilized mood in healthy individuals became associated with distress, age and BMI lost their anchoring roles, and emotional states emerged as dominant but erratic drivers of diet. These reversals and destabilizations were consistent across model iterations, suggesting robust signatures of preclinical vulnerability.

Conclusion: T2DM risk is better conceptualized as systemic reorganization within behavioral networks rather than the additive effects of isolated factors. By combining survival models with ANN-derived coherence mapping, this study demonstrates that early prediction is possible from modifiable, everyday behaviors without laboratory measures. The framework highlights leverage points for psychologically informed, personalized prevention strategies.