DIGITAL HEALTH最新文献

Background: Nonalcoholic fatty liver disease (NAFLD) is recognized as one of the most common chronic liver diseases worldwide. This study aims to assess the efficacy of automated machine learning (AutoML) in the identification of NAFLD using a population-based cross-sectional database.

Methods: All data, including laboratory examinations, anthropometric measurements, and demographic variables, were obtained from the National Health and Nutrition Examination Survey (NHANES). NAFLD was defined by controlled attenuation parameter (CAP) in liver transient ultrasound elastography. The least absolute shrinkage and selection operator (LASSO) regression analysis was employed for feature selection. Six algorithms were utilized on the H2O-automated machine learning platform: Gradient Boosting Machine (GBM), Distributed Random Forest (DRF), Extremely Randomized Trees (XRT), Generalized Linear Model (GLM), eXtreme Gradient Boosting (XGBoost), and Deep Learning (DL). These algorithms were selected for their diverse strengths, including their ability to handle complex, non-linear relationships, provide high predictive accuracy, and ensure interpretability. The models were evaluated by area under receiver operating characteristic curves (AUC) and interpreted by the calibration curve, the decision curve analysis, variable importance plot, SHapley Additive exPlanation plot, partial dependence plots, and local interpretable model agnostic explanation plot.

Results: A total of 4177 participants (non-NAFLD 3167 vs NAFLD 1010) were included to develop and validate the AutoML models. The model developed by XGBoost performed better than other models in AutoML, achieving an AUC of 0.859, an accuracy of 0.795, a sensitivity of 0.773, and a specificity of 0.802 on the validation set.

Conclusions: We developed an XGBoost model to better evaluate the presence of NAFLD. Based on the XGBoost model, we created an R Shiny web-based application named Shiny NAFLD (http://39.101.122.171:3838/App2/). This application demonstrates the potential of AutoML in clinical research and practice, offering a promising tool for the real-world identification of NAFLD.

Objective: Prior research has not assessed the value of remote patient monitoring (RPM) systems for patients undergoing anticoagulation therapy after cardiac surgery. This study aims to assess whether the clinical follow-up through RPM yields comparable outcomes with the standard protocol.

Methods: A crossover trial assigned participants to SOC-RPM or RPM-SOC, starting with the standard of care (SOC) for the first 6 months after surgery and using RPM for the following 6 months, or vice-versa, respectively. During RPM, patients used the Coaguchek^© to accurately measure International Normalized Ratio values and a mobile text-based chatbot to report PROs and adjust the therapeutic dosage. The study assessed patients' and clinicians' experience with RPM and compared direct costs.

Results: Twenty-seven patients participated. The median time in therapeutic range (TTR) levels during RPM were 72.2% and 50.6% for the SOC-RPM and RPM-SOC arms, respectively, and during SOC, they were 49.4% and 58.4% for SOC-RPM and RPM-SOC arms, respectively. Patients and the clinical team reported high trust and satisfaction with the proposed digital service. Statistically significant differences were only found in the cost of RPM in the RPM-SOC, which was higher than SOC in the SOC-RPM arm.

Conclusions: Portable coagulometers and chatbots can enhance the remote management of patients undergoing anticoagulation therapy, improving patient experience. This presents a promising alternative to the current standard procedure. The results of this study seem to suggest that RPM may have a higher value when initiated after a SOC period rather than starting RPM immediately after surgery.Trial registration: ClinicalTrials.gov NCT06423521.

Background: The deep integration of digital technology and healthcare services has propelled the healthcare system into the era of digital health. However, vulnerable populations in the field of information technology, they face challenges in benefiting from the digital dividends brought by digital health, leading to the emerging phenomenon of the "health digital divide."

Methods: This study utilized the sample of 3547 urban from the 2021 Chinese Social Survey data for analysis. Models were constructed with digital access divide, digital usage divide, and digital outcome divide for urban residents, and structural equation modeling was implemented for analysis.

Results: The impact β coefficients (95% CI) of urban residents' digital access on the frequency of digital use, internet healthcare utilization, and patient experience were (β = 0.737, P < 0.001), (β = 0.047, P < 0.05), and (β = 0.079, P < 0.001), respectively. Urban elderly groups were at a disadvantage in digital access and usage (β = -0.007, β = -0.024, and β = -0.004), as well as those with lower educational levels (β = 0.109, β = 0.162, and β = 0.045). However, these two factors did not have a significant direct impact on the patient experience in urban areas.

Conclusions: The health digital divide of urban residents exhibits a cascading effect, primarily manifested in the digital access and usage divide. To bridge health digital divide among urban residents, efforts must be made to improve digital access and usage among the elderly and those with lower educational levels.

Objective: Current studies lack a comprehensive understanding of the environmental factors influencing type 2 diabetes, hindering an in-depth grasp of the overall etiology. To address this gap, we utilized network science tools to highlight research trends, knowledge structures, and intricate relationships among factors, offering a new perspective for a profound understanding of the etiology.

Methods: The Web of Science database was employed to retrieve documents relevant to environmental risk factors in type 2 diabetes from 2012 to 2024. Bibliometric analysis using Microsoft Excel and OriginPro provided a detailed scientific production profile, including articles, journals, countries, and authors. Co-occurrence analysis was employed to determine the collaboration state and knowledge structures, utilizing social network tools such as Gephi, Tableau, and R Studio. Additionally, theme evolutionary analysis was conducted using SciMAT to offer insights into research trends.

Results: The publications and themes related to environmental factors in type 2 diabetes have consistently risen, shaping a well-established research domain. Lifestyle environmental factors, particularly diet and nutrition, stand out as the most represented and rapidly growing topics. Key focal hotspots include sedentary and digital behavior, PM_2.5, ethnicity and socioeconomic status, traffic and greenspace, and depression. The theme evolutionary analysis revealed three distinct paths: (1) oxidative stress-air pollutants-PM_2.5-air pollutants; (2) calcium-metabolic syndrome-cardiovascular disease; and (3) polychlorinated biphenyls (PCBs)-persistent organic pollutants (POPs)-obesity.

Conclusions: Digital behavior signifies a novel approach for preventing and managing type 2 diabetes. The influence of PM_2.5 and calcium on oxidative stress and abnormal vascular contraction is intricately linked to microvascular diabetes complications. The transition from PCBs and POPs to obesity underscores the disruption of endocrine function by chemicals, elevating the risk of diabetes. Future studies should explore the connections between environmental factors, microvascular complications, and long-term outcomes in diabetes.

Introduction: Worldwide, healthcare systems aim to achieve the best possible quality of care at an affordable cost while ensuring broad access for all populations. The use of artificial intelligence (AI) in healthcare holds promise to address these challenges through the integration of real-world data-driven insights into patient care processes. This study aims to assess nurses' awareness and attitudes toward AI-integrated tools used in clinical practice.

Methods: A descriptive cross-sectional design captured nurses' responses at three governmental hospitals in Saudi Arabia by using an online questionnaire administered over 4 months. The study involved 220 registered nurses with a minimum of one year of clinical experience, selected through a convenience sampling method. The online survey consisted of three sections: demographic information, an assessment of nurses' AI knowledge, and the general attitudes toward the AI scale.

Results: Nurses displayed "moderate" levels of awareness toward AI technology, with 70.9% having basic information about AI and only 58.2% (128 nurses) were considered "aware" of AI as they dealt with one of its healthcare applications. Nurses expressed openness to AI integration (M = 3.51) on one side, but also had some concerns about AI. Nurses expressed conservative attitudes toward AI, with significant differences observed based on gender (χ² = 4.67, p < 0.05). Female nurses exhibited a higher proportion of negative attitudes compared to male nurses. Significant differences were also found based on age (χ² = 9.31, p < 0.05), with younger nurses demonstrating more positive attitudes toward AI compared to their older counterparts. Educational background yields significant differences (χ² = 6.70, p < 0.05), with nurses holding undergraduate degrees exhibiting the highest positive attitudes. However, years of nursing experience did not reveal significant variations in attitudes.

Conclusion: Healthcare and nursing administrators need to work on increasing the nurses' awareness of AI applications and emphasize the importance of integrating such technology into the systems in use. Moreover, addressing nurses' concerns about AI's control and discomfort is crucial, especially considering generational differences, with younger nurses often having more positive attitudes toward technology. Change management strategies may help overcome any hindrances.

Background: Although the prevalence of childhood illnesses has significantly decreased, acute respiratory infections continue to be the leading cause of death and disease among children in low- and middle-income countries. Seven percent of children under five experienced symptoms in the two weeks preceding the Ethiopian demographic and health survey. Hence, this study aimed to identify interpretable predicting factors of acute respiratory infection disease among under-five children in Ethiopia using machine learning analysis techniques.

Methods: Secondary data analysis was performed using 2016 Ethiopian demographic and health survey data. Data were extracted using STATA and imported into Jupyter Notebook for further analysis. The presence of acute respiratory infection in a child under the age of 5 was the outcome variable, categorized as yes and no. Five ensemble boosting machine learning algorithms such as adaptive boosting (AdaBoost), extreme gradient boosting (XGBoost), Gradient Boost, CatBoost, and light gradient-boosting machine (LightGBM) were employed on a total sample of 10,641 children under the age of 5. The Shapley additive explanations technique was used to identify the important features and effects of each feature driving the prediction.

Results: The XGBoost model achieved an accuracy of 79.3%, an F1 score of 78.4%, a recall of 78.3%, a precision of 81.7%, and a receiver operating curve area under the curve of 86.1% after model optimization. Child age (month), history of diarrhea, number of living children, duration of breastfeeding, and mother's occupation were the top predicting factors of acute respiratory infection among children under the age of 5 in Ethiopia.

Conclusion: The XGBoost classifier was the best predictive model with improved performance, and predicting factors of acute respiratory infection were identified with the help of the Shapely additive explanation. The findings of this study can help policymakers and stakeholders understand the decision-making process for acute respiratory infection prevention among under-five children in Ethiopia.

Background: During the hospital-to-home transition, stroke survivors and their caregivers face a significant lack of support and information which impacts their psychosocial recovery. We aimed to co-design a program combining individual support by a trained case-manager (dedicated professional providing individual support) and an online information platform to address needs of stroke survivors and caregivers.

Methods: A two-step methodology was used. The first step followed a "user-centered design" approach during four workshops with stroke survivors, caregivers, and healthcare professionals to develop the platform and define the case-manager profile. The second step was a usability test of the platform following a Think Aloud method with patients and caregivers. The workshops and interviews were analyzed following a qualitative thematic analysis. The analysis of Think Aloud interviews was based on User Experience Honeycomb framework by Morville.

Results: Eight participants attended the workshops: two patients, two caregivers, three nurses, and a general practitioner. Activities, training, and skills of the case-manager were defined according to stroke survivors and caregivers needs. Name, graphics, navigation, and content of the platform were developed with the participants, a developer and a graphic designer. The usability of the platform was tested with 5 patients and 5 caregivers. The Think Aloud confirmed satisfaction with graphics and content but a need for improvement regarding the navigability. An update of the platform was conducted in order to answer the needs expressed by participants.

Conclusion: We developed, with a participatory approach, a patient-centered transition program, which will be evaluated in a randomized controlled trial.

Objective: Clinical observations suggest that individuals with a diagnosis of bipolar face difficulties regulating emotions and impairments to their cognitive processing, which can contribute to high-risk behaviours. However, there are few studies which explore the types of risk-taking behaviour that manifest in reality and evidence suggests that there is currently not enough support for the management of these behaviours. This study examined the types of risk-taking behaviours described by people who live with bipolar and their access to support for these behaviours.

Methods: Semi-structured interviews were conducted with n = 18 participants with a lived experience of bipolar and n = 5 healthcare professionals. The interviews comprised open-ended questions and a Likert-item questionnaire. The responses to the interview questions were analysed using content analysis and corpus linguistic methods to develop a classification system of risk-taking behaviours. The Likert-item questionnaire was analysed statistically and insights from the questionnaire were incorporated into the classification system.

Results: Our classification system includes 39 reported risk-taking behaviours which we manually inferred into six domains of risk-taking. Corpus linguistic and qualitative analysis of the interview data demonstrate that people need more support for risk-taking behaviours and that aside from suicide, self-harm and excessive spending, many behaviours are not routinely monitored.

Conclusion: This study shows that people living with bipolar report the need for improved access to psychologically informed care, and that a standardised classification system or risk-taking questionnaire could act as a useful elicitation tool for guiding conversations around risk-taking to ensure that opportunities for intervention are not missed. We have also presented a novel methodological framework which demonstrates the utility of computational linguistic methods for the analysis of health research data.

Objective: Poor conditions in the intraoral environment often lead to low-quality photos and videos, hindering further clinical diagnosis. To restore these digital records, this study proposes a real-time interactive restoration system using segment anything model.

Methods: Intraoral digital videos, obtained from the vident-lab dataset through an intraoral camera, serve as the input for interactive restoration system. The initial phase employs an interactive segmentation module leveraging segment anything model. Subsequently, a real-time intraframe restoration module and a video enhancement module were designed. A series of ablation studies were systematically conducted to illustrate the superior design of interactive restoration system. Our quantitative evaluation criteria contain restoration quality, segmentation accuracy, and processing speed. Furthermore, the clinical applicability of the processed videos was evaluated by experts.

Results: Extensive experiments demonstrated its performance on segmentation with a mean intersection-over-union of 0.977. On video restoration, it leads to reliable performances with peak signal-to-noise ratio of 37.09 and structural similarity index measure of 0.961, respectively. More visualization results are shown on the https://yogurtsam.github.io/iveproject page.

Conclusion: Interactive restoration system demonstrates its potential to serve patients and dentists with reliable and controllable intraoral video restoration.

Introduction: Telemedicine has been shown to be an effective approach for people with substance-related disorders. Analyzing patient satisfaction with telemedicine is necessary for improving treatment outcomes. This study aims to assess patient satisfaction with telemedicine for substance-related disorders at the Centro Asistencial Córdoba in Argentina.

Methods: A cross-sectional, descriptive, and correlational design was carried out. A patient satisfaction survey was created, consisting of eight questions and a quality-of-life question, which was administered to N = 115 patients.

Results: The results showed that more than 90% agreed with the ease of use of virtual consultations, 82% felt they received the same level of care as if the consultation had been in person, 86% agreed with the adequacy of time utilized during the virtual session, and over 85% agreed to repeat their telemedicine treatment. Regarding the composite variable "users' assessment of telemedicine," we found an average of 17.41 ± 2.80. Concerning satisfaction with virtual care and the previous use of telemedicine, 95.7% were satisfied, and nearly 61.7% reported not having used virtual care previously. In terms of money and time saved, 93.9% saved money with virtual consultations, 66.1% saved more than two hours per week, 23.5% saved more than one hour per week, and 10.4% saved less than one hour per week.

Conclusions: Overall, there is significant approval of telemedicine among users of substance-related disorders services. In particular, they were satisfied with the time employed, the benefits of saving time and money, and the ease of use of telemedicine; furthermore, they were positive about undergoing telemedicine treatment in the future.