DIGITAL HEALTH最新文献

{"title":"Artificial intelligence for healthcare in Nepal: A scoping review of clinical applications, implementation evidence, and health-system governance.","authors":"Prajjwol Luitel, Nischal Neupane","doi":"10.1177/20552076261483403","DOIUrl":"10.1177/20552076261483403","url":null,"abstract":"<p><strong>Background: </strong>Nepal offers a distinctive LMIC setting for evaluating AI-health adoption due to its difficult geography, specialist shortage, high case burden needing screening and triage, and emerging digital-health policy reforms.</p><p><strong>Objective: </strong>This scoping review aimed to synthesize existing literature on AI applications in medicine and healthcare in Nepal, including clinical use cases, education, implementation evidence, and health-system governance.</p><p><strong>Methods: </strong>A scoping review was conducted following the Arksey-O'Malley framework and PRISMA-ScR reporting standards. Searches of PubMed, Embase, Scopus, and grey literature sources were conducted through May 2026. Sources addressing AI applications, implementation, education, or governance in Nepalese healthcare settings were included. Evidence was narratively synthesized, and implementation barriers were interpreted using the NASSS framework.</p><p><strong>Results: </strong>Twenty-five sources were included. The literature showed transition from conceptual commentary toward empirical validation and early implementation studies. Imaging-based applications dominated, particularly in ophthalmology and chest radiography. Several AI systems demonstrated high diagnostic performance in selected datasets, including applications for retinopathy of prematurity, glaucoma, tuberculosis screening, and chest radiography. However, evidence was concentrated in urban or specialist settings, with no prospective implementation trials, limited rural validation, non-imaging clinical applications, health-economic analysis, and minimal governance evaluation. NASSS synthesis identified barriers related to local validation, accountability, workflow integration, regulatory uncertainty, and sustainability.</p><p><strong>Conclusion: </strong>Nepal's AI-health landscape has progressed from conceptual awareness toward early diagnostic validation and deployment, particularly in imaging-based screening and medical education. Yet, the current evidence remains insufficient for conclusions about scale-up, safety, equity, cost-effectiveness, or system-wide impact. Future studies should prioritize multi-site and rural validation, implementation studies, health-economic evaluation, non-imaging clinical applications, and phased AI-governance aligned with broader digital-health system development.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261483403"},"PeriodicalIF":3.9,"publicationDate":"2026-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13525273/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148857890","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Investigating the mobile messenger app-based intervention delivered by a therapeutic companion agent for young adults with depressive symptoms: A proof-of-concept and quasi-experimental study.","authors":"Youngjae Yoo, Minuk Kim, Soyoung Kim, Gayeon Lee, Jinwoo Kim","doi":"10.1177/20552076261464801","DOIUrl":"10.1177/20552076261464801","url":null,"abstract":"<p><strong>Background: </strong>Despite the clinical potential of digital interventions for young adults with depression, low engagement remains a persistent challenge. A conversational therapeutic companion agent (TCA) can foster a digital therapeutic alliance to address this barrier.</p><p><strong>Objective: </strong>This study aimed to develop a Wizard-of-Oz TCA prototype integrating behavioral activation, ecological momentary assessment/intervention (EMA/EMI), and gamification, and to evaluate its proof-of-concept feasibility, acceptability, and preliminary clinical efficacy for young adults with depressive symptoms.</p><p><strong>Methods: </strong>58 Korean young adults ages 20 to 39 (50.0% female) with subclinical to moderately severe depressive symptoms (Patient Health Questionnaire-9) were recruited. The intervention group (n=29) interacted with the TCA for six weeks, while a frequency-matched passive control group (n=29) continued usual routines. Primary outcomes were feasibility (adherence) and acceptability (satisfaction). Secondary outcomes for efficacy (Beck Depression Inventory-II, Generalized Anxiety Disorder-7, Quality-of-Life Enjoyment and Satisfaction Questionnaire-Short Form) were assessed biweekly.</p><p><strong>Results: </strong>Of 58 participants, 57 completed the study. The intervention proved feasible, acceptable, and usable, achieving 69.5% strict on-time EMA adherence, 66.2% behavioral activation routine completion, a mean daily satisfaction of 3.49/5.0, and a mean System Usability Scale (SUS) score of 74.02, meeting <i>a priori</i> benchmarks. Exploratory analyses indicated the intervention group showed significantly greater reductions in depressive symptoms and improvements in quality of life compared to controls.</p><p><strong>Conclusions: </strong>The TCA-based intervention is feasible and acceptable, successfully cultivating a digital therapeutic alliance. While preliminary clinical signals are promising, future fully powered randomized controlled trials are required to definitively establish clinical efficacy.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261464801"},"PeriodicalIF":3.9,"publicationDate":"2026-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13519235/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148842032","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Learner experiences of a carer-focused massive open online course on psychosis: A qualitative study.","authors":"Kalya Win Aung, Angela Kibia, Juliana Onwumere","doi":"10.1177/20552076261470199","DOIUrl":"10.1177/20552076261470199","url":null,"abstract":"<p><strong>Background: </strong>People supporting individuals with psychotic disorders play a vital role in community-based mental health care but often face their own health and wellbeing challenges and limited access to support. Digital interventions, including online courses, may help address these unmet needs.</p><p><strong>Purpose: </strong>This study explored participants' experiences of the Caring for People with Psychosis and Schizophrenia MOOC, using the Kirkpatrick Evaluation Model, a widely applied framework for evaluating educational programmes.</p><p><strong>Methods: </strong>Individual semi-structured interviews were conducted online with participants who accessed the course. Interviews were audio-recorded, transcribed verbatim, and analysed using reflexive thematic analysis. Findings were then mapped onto the four levels of the Kirkpatrick model to explore learner reactions, knowledge gains, behavioural change, and wider impacts of the MOOC.</p><p><strong>Results: </strong>Nine participants (six carers and three non-carers) were interviewed, and five key themes were identified. Learners described enrolling during periods of emotional difficulty and uncertainty and found the course validating, informative, and connecting (reaction). They reported greater understanding of psychosis, caregiving strategies, and self-care approaches (learning), which increased their sense of confidence. Many carers applied this knowledge in their daily routines, communication, and relationships (behaviour), reportedly improving perceived wellbeing for themselves and their relatives (impact). The MOOC's flexible and multimodal design supported meaningful engagement alongside other responsibilities.</p><p><strong>Conclusions: </strong>Preliminary evidence from this model of online training suggests that a globally accessible MOOC can connect learners, enhance understanding, and build confidence in caregiving for psychosis. Subjective reports indicate the potential of MOOCs to address learners' informational and emotional needs through flexible, supportive online education.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261470199"},"PeriodicalIF":3.9,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13519204/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148841979","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Trends and research clusters in artificial intelligence-assisted ultrasound-guided regional anesthesia: A bibliometric analysis.","authors":"Xuewen Lu, Xiaodong Qiu, Jiuyu Ji","doi":"10.1177/20552076261484261","DOIUrl":"10.1177/20552076261484261","url":null,"abstract":"<p><strong>Background: </strong>This exploratory bibliometric study aimed to characterize publication trends, research clusters, and emerging topics in artificial intelligence (AI)-assisted ultrasound-guided regional anesthesia (UGRA).</p><p><strong>Methods: </strong>We searched the Web of Science Core Collection for English-language articles published or indexed from January 1, 2008, through December 10, 2025. Two independent reviewers performed eligibility assessment; discrepancies were reconciled via joint deliberation. Quantitative bibliometric processing was conducted using VOSviewer, CiteSpace, and the R package \"bibliometrix\".</p><p><strong>Results: </strong>Ninety publications from 33 countries or territories and 223 institutions were included. Annual output increased overall but fluctuated, peaking in 2021 with 17 publications. China contributed the highest volume (28 works), trailed by the United States (14) and the United Kingdom (12). The University of London ranked first in output volume (18 contributions). <i>The British Journal of Anaesthesia</i> had the highest publication count (TP = 7) and h-index (h = 6), whereas <i>Regional Anesthesia and Pain Medicine</i> had the highest total citation count (TC = 225). Bowness JS stood out both in terms of productivity and citation influence. Keyword co-occurrence analysis identified eight algorithmically derived clusters, which were subsequently consolidated into three dominant thematic research hotspots: accurate identification of anatomical structures, risk classification and prediction of perioperative complications, and guidelines and education in UGRA. Temporal keyword burst detection revealed \"nerve block\", \"brachial plexus\", and \"American Society\" as newly intensified focal points within the field.</p><p><strong>Conclusion: </strong>This exploratory bibliometric analysis summarizes the principal research themes in AI-assisted UGRA. Future studies should evaluate technical performance together with clinician-rated usability and patient-relevant outcomes and should validate AI tools across institutions, devices, and patient populations.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261484261"},"PeriodicalIF":3.9,"publicationDate":"2026-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13519185/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148842009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrating wearable biosensing with clinical interviews for suicide risk detection in adolescents.","authors":"Xin Huang, Zhenyu Zou, Zihang Su, Shiyao Zhu, Chang Lei, Yinan Duan, Zhenxing Zhang, Zhijun Wu, Jingyan Yan, Jingyi Wang, Qian Wang, He Ding, Mengling Feng, Hao Jin, Runsen Chen","doi":"10.1177/20552076261483391","DOIUrl":"10.1177/20552076261483391","url":null,"abstract":"<p><strong>Background: </strong>Adolescent suicide is a critical public health challenge. Traditional risk screening relies on self-report measures limited by various biases, while passive EDA monitoring during daily activities, often yields limited predictive validity due to environmental confounding such as motion artifacts, thermoregulatory sweating, and ambient temperature fluctuations. This study proposes an innovative \"interview-embedded\" framework to capture physiological signatures of suicide ideation (SI) during a standardized clinical probe.</p><p><strong>Methods: </strong>A total of 151 adolescents (102 with active suicide ideation, 49 matched controls) were enrolled. Their electrodermal activity signals were continuously recorded throughout clinical interview (Mini International Neuropsychiatric Interview for Children and Adolescents) and analyzed through advanced machine learning approaches.</p><p><strong>Results: </strong>XGBoost model achieved superior classification performance (AUC=0.802, sensitivity=0.857, specificity=0.8) compared to CatBoost, and Balanced Random Forest, significantly outperforming resting-state models and clinical symptom baselines (Resting EDA Model: AUC = 0.598, sensitivity = 0.571, specificity = 0.4; Full Clinical Interview Baseline: AUC = 0.926, sensitivity = 0.476, specificity = 1; Symptom-only Clinical Baseline Model: AUC = 0.712, sensitivity = 0.619, specificity = 0.8). Feature importance analysis revealed that dynamic features reflecting physiological reactivity were the most discriminative markers, providing predictive information potentially beyond self-report. EDA features did not significantly correlate with continuous BSS severity scores (all ρ < 0.15, all p > 0.05; regression R² < 0), supporting a threshold rather than dose-dependent physiological response to suicidal ideation. The results of robustness checks and subgroup analyses showed modest but clinically relevant utility of the model, even when accounting for highly comorbid factors such as depression and non-suicidal self-injury.</p><p><strong>Conclusions: </strong>This study demonstrates that a task-embedded biosensing framework, integrating wearable biosensing into standardized clinical interviews is a feasible and effective approach for adolescent suicide risk detection. Embedding physiological data collection within established clinical workflows offers a scalable solution to improve early suicide risk screening in real-world settings, such as schools and primary care.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261483391"},"PeriodicalIF":3.9,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13504123/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148820197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Playing for better health: Gamification in digital workplace health promotion - A structured review of design elements, outcomes, and mechanisms from a self-determination theory perspective.","authors":"Kristina Hall, Laura Watkowski, Berit Helmus","doi":"10.1177/20552076261481633","DOIUrl":"10.1177/20552076261481633","url":null,"abstract":"<p><strong>Background: </strong>Digital technologies increasingly shape the work environment of knowledge workers and can negatively affect employee well-being. While organizations have introduced digital workplace health promotion programs to address these challenges, many initiatives struggle with low engagement. Gamification is considered a promising approach to enhance participation, yet it remains unclear how different game design elements influence various dimensions of employee well-being, leaving organizations with limited guidance on how to design effective gamified health interventions.</p><p><strong>Objective: </strong>This study examines how gamification in digital workplace health promotion programs contributes to employee well-being by (1) identifying the targeted well-being dimensions, (2) analyzing game design elements and their theoretical foundations, and (3) extracting Self-Determination Theory-related success factors.</p><p><strong>Method: </strong>We conducted a structured literature review across AIS eLibrary, PubMed, JSTOR, ScienceDirect, and Web of Science, last searched in 2023. The search yielded a total of 823 papers for screening. Studies were included if they empirically examined gamification technologies in workplace contexts and their effects on well-being or health-related behavior among knowledge workers; non-empirical studies, studies outside the workplace context, and studies without gamification or well-being-related outcomes were excluded. Eligible studies were analyzed using a concept matrix to systematically map game design elements, well-being dimensions, and Self-Determination Theory-related success factors.</p><p><strong>Results: </strong>Eighteen studies were included in the final synthesis. Most studies focused on physical well-being, while mental well-being and job-related outcomes were examined less frequently. Four design strategies emerged: progress-, social-, financial-, and immersion-oriented elements. Self-Determination Theory appears useful for understanding gamification effects, although few studies explicitly apply theoretical frameworks.</p><p><strong>Conclusion: </strong>Gamified digital workplace health promotion programs show promise in enhancing employee well-being, yet existing research remains conceptually narrow and insufficiently theory driven. Therefore, we propose a research agenda calling for greater conceptual clarity, stronger theoretical grounding, and the evaluation of long-term, personalized, and adverse effects.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261481633"},"PeriodicalIF":3.9,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13504128/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148820145","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and evaluation of a digitally based functional fitness self-assessment and monitoring tool for older adults.","authors":"Jason Tallis, Darren Richardson, Sharn Shelley, Neil Clarke, David Broom, Michael Duncan","doi":"10.1177/20552076261481297","DOIUrl":"10.1177/20552076261481297","url":null,"abstract":"<p><strong>Background: </strong>Maintaining physical function is central to healthy ageing and is emphasised in health policy aimed at shifting care from treatment to prevention. However, accessible tools that support routine monitoring of functional fitness remain limited despite evidence demonstrating links to adverse health outcomes. A Functional Fitness Monitoring Tool (FFMT) was developed as a digital, self-administered platform enabling older adults and healthcare professionals to assess and track physical function. The study aimed to; 1) evaluate perceived importance and usability; 2) compare between-session reliability of self-administered (SL) versus instructor-administered (IL) assessments.</p><p><strong>Method: </strong>Thirty-three community-dwelling older adults (68 ± 6 years) completed three SL assessments followed by two IL assessments, or vice versa. The FFMT incorporated nine validated functional tests with digital instructions, demonstration videos, and a results-monitoring interface. After all assessments, participants completed an online usability survey.</p><p><strong>Results: </strong>Between-session reliability across SL and IL assessments was good to excellent (ICC > 0. 816). Reliability remained across groups, though wider confidence intervals appeared for initial SL trials in some tests. Thresholds for measurable change were small. Participants found the FFMT easy to use, interpret, and valuable for long-term monitoring.</p><p><strong>Conclusion: </strong>The FFMT demonstrates high reliability and usability, with self-administered assessments performing comparably to instructor-led administration. These findings support the FFMT as an accessible, scalable tool capable of enabling accurate self-monitoring of functional fitness across clinical, community, and population settings, with potential to facilitate early detection of functional decline and promote prevention-focused healthy ageing.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261481297"},"PeriodicalIF":3.9,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13504135/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148820151","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence chatbots as information sources for postherpetic neuralgia: An evaluation of reliability and readability.","authors":"Chenxin Sun, Jingxuan Sun, Shihao Jing, Yong Fei, Ling Ma, Jianjun Zhu","doi":"10.1177/20552076261475785","DOIUrl":"10.1177/20552076261475785","url":null,"abstract":"<p><strong>Background: </strong>Artificial intelligence (AI) chatbots are increasingly used for online health information seeking and have become a common source of information on postherpetic neuralgia (PHN). The quality and readability of the information they provide have not been systematically evaluated.</p><p><strong>Objective: </strong>To assess the quality, transparency, and readability of postherpetic neuralgia information generated by mainstream artificial intelligence chatbots and to examine whether this content meets the recommended sixth-grade reading level for patient education materials.</p><p><strong>Methods: </strong>Using the MeSH term Neuralgia, Postherpetic, we identified 11 highly relevant postherpetic neuralgia-related search queries worldwide from 2020 to 2025 through Google Trends. These queries were entered verbatim into ChatGPT-4o, Gemini-1.5, Perplexity Pro, and Copilot using a standardised input procedure. Responses were assessed with DISCERN, EQIP, JAMA, and GQS for information quality, completeness, transparency, and overall educational value. Readability was evaluated with ARI, FRES, GFI, FKGL, CL, and SMOG. The Kruskal-Wallis test was used for between-model comparisons, and the Wilcoxon signed-rank test was used to compare readability indices against sixth-grade reading benchmarks.</p><p><strong>Results: </strong>DISCERN, EQIP, and GQS scores differed significantly across chatbots in DISCERN (P < 0.001), EQIP (P < 0.001), and GQS (<i>P</i> < 0.001) scores among the four chatbots; JAMA scores showed no significant difference (<i>P</i> = 0.668). No chatbot reached an excellent overall rating, although Perplexity Pro and Copilot performed better than ChatGPT-4o and Gemini-1.5 on most quality-related measures. For readability, none of the chatbots met the sixth-grade reading standard on any metric: all FRES scores were <80, and all other indices (ARI, GFI, CL, FKGL, SMOG) exceeded the grade-6 thresholds. Readability did not differ significantly among chatbots (all <i>P</i> > 0.2).</p><p><strong>Conclusion: </strong>Mainstream artificial intelligence chatbots can provide postherpetic neuralgia information with moderate quality, but the readability of this content remains consistently above recommended patient education levels. The findings should be interpreted within the scope of the selected chatbots, default operational settings, search queries, and the time-sensitive nature of chatbot outputs. Further improvement in source reporting and plain-language presentation remains warranted.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261475785"},"PeriodicalIF":3.9,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13504130/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148820116","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Laparoscopic image segmentation of the hepatobiliary system: Current status and advances from a deep learning perspective.","authors":"Qunzhou Sun, Feiyang Yang, Ping Zhang, Tianyang Zhang, Jun Qi, Lu Sun","doi":"10.1177/20552076261483318","DOIUrl":"10.1177/20552076261483318","url":null,"abstract":"<p><p>Hepatobiliary laparoscopy provides a minimally invasive approach for treating hepatobiliary diseases, complex anatomical variations, and temporary intraoperative tissue occlusion, rendering these procedures highly challenging. To evaluate how computer vision can overcome these visual limitations, this review systematically analyses deep learning-based laparoscopic image segmentation for the hepatobiliary system. Based on searches across Google Scholar, Web of Science, and PubMed, 16 eligible studies focusing on pixel-level laparoscopic hepatobiliary image segmentation published between 2018 and early 2026 were selected and synthesised. Findings reveal that current research predominantly focuses on gallbladder segmentation, followed by the liver and bile ducts. Methodologically, reviewed network architectures-ranging from foundational convolutional networks to hybrid vision models-are categorised into seven goal-driven optimisation schemes, primarily addressing spatiotemporal modelling, class imbalance mitigation, and model lightweighting. Quantitative synthesis highlights a critical trade-off between algorithmic performance and engineering overhead; notably, while conventional three-dimensional (3D) models enhance spatiotemporal continuity, they demand substantial computational and memory resources. Furthermore, we delineate key bottlenecks hindering clinical translation, including multicentre data variance, alarm fatigue, and compliance with DECIDE-AI reporting guidelines. Ultimately, by mapping this translational landscape, this review provides a roadmap to accelerate the deployment of deep learning paradigms in laparoscopic hepatobiliary surgery.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261483318"},"PeriodicalIF":3.9,"publicationDate":"2026-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13500991/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148814814","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence-based prediction of metabolic dysfunction-associated steatotic liver disease (MASLD) and liver fibrosis in patients with obstructive sleep apnea: An innovative approach.","authors":"Songül Özyurt, Neslihan Özçelik, Ali Erdem Özçelik, Emre Bendeş, Çağlayan Keklikkıran, Eray Halisdemir, Aziz Gümüş","doi":"10.1177/20552076261480861","DOIUrl":"10.1177/20552076261480861","url":null,"abstract":"<p><strong>Background: </strong>Obstructive sleep apnea (OSA) is strongly associated with metabolic dysfunction and may contribute to the development of hepatic disorders such as metabolic dysfunction-associated steatotic liver disease (MASLD) and liver fibrosis. Early identification of hepatic involvement in OSA is important for preventing long-term complications. This study aimed to develop and compare artificial intelligence (AI)-based models using non-invasive clinical parameters to predict steatosis, MASLD, and liver fibrosis in patients with OSA.</p><p><strong>Methods: </strong>This prospective study included patients diagnosed with OSA based on polysomnography (PSG). Hepatic steatosis and fibrosis were assessed non-invasively using the controlled attenuation parameter (CAP) and liver stiffness measurement (LSM) obtained by FibroScan. A total of 700 patients were screened, and 517 patients who met the inclusion criteria were analyzed, including 42 demographic, anthropometric, laboratory, and sleep-related variables. A total of 42 variables were recorded, comprising 37 candidate predictors and 4 target outcomes. Following feature selection, 11 of the 37 candidates were retained for model training. Four supervised machine learning algorithms-logistic regression, support vector machine, random forest, and artificial neural network were trained to predict steatosis, MASLD, and liver fibrosis. Model performance was evaluated by accuracy, F1-score, and area under the ROC curve. Feature importance analysis identified the most influential predictors.</p><p><strong>Results: </strong>Among the four models, the Random Forest algorithm achieved the best performance, with an accuracy of 96% for binary classification; in an exploratory five-class analysis, overall accuracy reached 92%, although predominantly driven by the majority F0 class. The most influential predictors were apnea-hypopnea index (AHI), minimum oxygen saturation (minSpO₂), and oxygen desaturation index. A sensitivity analysis excluding elastography-derived features (CAP and LSM) was performed using 9 predictor variables (Model 2). Model 2 achieved best AUC values of 0.700, 0.706, and 0.634 for steatosis, MASLD, and fibrosis, respectively, compared with 0.967, 0.969, and 0.985 in the original 11-feature model (Model 1).</p><p><strong>Conclusion: </strong>AI-based models, particularly Random Forest, demonstrated high predictive performance in predicting MASLD and liver fibrosis in OSA patients. These models may support detection and risk stratification of hepatic involvement in sleep-disordered breathing.</p>","PeriodicalId":51333,"journal":{"name":"DIGITAL HEALTH","volume":"12 ","pages":"20552076261480861"},"PeriodicalIF":3.9,"publicationDate":"2026-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13504122/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148820182","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}