SEARCH · Search PubMed

Results for “multi-sector services”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

294 records · Page 11Linked to original sources

PubMed2026

Comprehensive quality profiling and comparative metabolic characterization of seven dominant fresh-eating Chinese olive (Canarium album Lour.) cultivars in Southern China.

Fresh-eating Chinese olive (Canarium album Lour.) is a subtropical fruit endemic to southern China with considerable commercial value, yet systematic quality characterization of dominant cultivars remains scarce. This study established a multi-dimensional quality dataset for seven dominant cultivars from Fujian and Guangdong provinces, integrating nutritional components, soluble sugars, organic acids, mineral elements, volatile profiles, and non-targeted metabolomics. Significant cultivar-specific differences were observed across all evaluated dimensions: "Lingfeng" exhibited a sugar-dominant low-acid profile, whereas "Sanleng" showed elevated phenolic constituents accumulation. Volatile profiling identified terpenoid-based candidate discriminatory biomarkers, and metabolomic analysis revealed phenylpropanoid biosynthesis, tryptophan metabolism, and starch and sucrose metabolism as the most variable pathways. Correlations between untargeted profiling and targeted absolute quantification validated untargeted result reliability and revealed their complementarity in nutritional evaluation. These findings provide baseline data for FECO germplasm evaluation and targeted industrial utilization.

China↗

PubMed2026

Suicide in rural South Africa before and during COVID-19: evidence from forensic mortuary data in Limpopo Province.

Suicide is a growing public health concern in South Africa, with rural provinces such as Limpopo facing heightened vulnerability due to limited mental health services and socio-economic inequalities. Evidence on the impact of COVID-19 on suicide in rural contexts remains limited. This study examined suicide trends in Limpopo Province and changes associated with the COVID-19 period. A retrospective interrupted time series analysis was conducted using forensic mortality data from 1 January 2019 to 31 December 2021, allowing assessment of pre-existing trends and changes following COVID-19 lockdowns. Among 5770 unnatural deaths, 957 were suicides. The proportion of suicides increased from 29.5% in 2019 to 36.9% in 2021. Suicides predominantly occurred among males and young adults, with hanging accounting for over 90% of deaths throughout. Interrupted time series analysis revealed a significant downward trend in suicide cases during the strict national lockdown (Alert Level 5), with a 31% reduction in incidence (IRR = 0.69, 95% CI: 0.49-0.98). Less restrictive lockdown levels showed no significant effects. Suicide mortality increased prior to COVID-19, with a subsequent decline during the strictest lockdown period. Stable demographic patterns and methods highlight persistent vulnerabilities and the need for sustained suicide-prevention strategies beyond pandemic.

Humans↗

PubMed2026

Driven toward care, avoiding the end: A systematic review and meta-analysis of the relationship between death anxiety and healthcare utilisation.

Both overuse and underuse of the healthcare system have been recognised as significant problems. Relatedly, growing research has recognised the key role of death anxiety in driving various health-relevant behaviours. However, the relationship between death anxiety and healthcare utilisation has not yet been systematically explored. The current systematic review and meta-analysis addressed this gap. In total, 987 papers were screened for inclusion, of which 63 were included in the final review (Ntotal = 21,271). This included 33 quantitative studies, 27 qualitative studies and 3 mixed-methods designs. In total, 17 studies contained sufficient data to be meta-analysed. Overall, the included studies highlighted a significant relationship between death anxiety and healthcare utilisation; in particular, positive associations with desire for life-prolonging treatments and contact with hospitals and medical professionals. By contrast, a negative association was found with other aspects of healthcare utilisation, including hospice use and end-of-life communication. The sample type emerged as a significant moderator, suggesting that the relationship between death anxiety and healthcare usage was strongest in non-medical samples. The current findings suggest that death anxiety plays a key role in utilisation of the healthcare system. The fear of death may need to be targeted in psychological interventions, in order to ensure maximal effectiveness of health services, and improve outcomes for healthcare users.

Humans↗

PubMed2026

Mechanisms of Hematopoietic Stem Cell Aging and Emerging Rejuvenation Strategies.

Hematopoietic stem cell (HSCs) aging is a complex biological process driven by both cell-intrinsic alterations and extrinsic cues from the bone marrow niche. Understanding these mechanisms is critical for developing therapies against aging-related hematopoietic disorders. This review synthesizes recent advances in the molecular mechanisms underlying HSCs aging, including microenvironmental aging, genomic instability, epigenetic dysregulation, mitochondrial dysfunction, and aberrant nuclear mechanotransduction. We summarize that the functional decline of HSCs during aging drives a compensatory expansion of the phenotypically defined stem cell pool, leading to an aberrant increase in cell number. We also highlight aging-associated HSCs heterogeneity, including CD150high and P-selectin-positive subsets that enrich for myeloid-biased or functionally compromised HSCs states while emphasizing that surface phenotype alone may not fully indicate functional rejuvenation. Finally, we discuss emerging rejuvenation strategies-including targeting myeloid-biased HSCs, modulating inflammatory pathways, and implementing epigenetic or metabolic interventions-supported by cutting-edge technologies such as single-cell multi-omics, gene editing, and computational modeling. These approaches hold promise for counteracting age-related hematopoietic decline and restoring immune competence.

Humans↗

PubMed2026

SGLF-Net:Staged Global-to-Local Cross-Scale Fusion Network for Colonoscopic Polyp Segmentation.

Polyp segmentation in colonoscopy images plays a pivotal role in computer-aided medical diagnosis and the early prevention of colorectal cancer. However, existing methods often suffer from performance degradation when confronted with extreme polyp scale variation and polyp boundary ambiguity. To address these challenges, we propose the Staged Global-to-Local Cross-Scale Fusion Network (SGLF-Net), which adopts a novel staged global-to-local learning paradigm to progressively refine segmentation from coarse global semantics to fine-grained local details. Specifically, the Global Semantic Perception Stage integrates a Swin Transformer Encoder and a Dynamic Attentive Decoder (DAD) to construct comprehensive multi-scale contextual representations. The Local Detail Refinement Stage employs an Edge-aware Dynamic Attentive Decoder (E-DAD) to enhance structural fidelity and boundary precision through explicit edge-guided supervision. Furthermore, we introduce the Cross Spatial-Scale Feature Aggregation and Reconstitution (CSSAR) module, equipped with hybrid attention mechanisms, to facilitate efficient semantic structural interaction between the two cascaded stages. Extensive experiments on five public benchmark datasets demonstrate that SGLF-Net consistently outperforms state-of-the-art methods in both segmentation accuracy and boundary preservation.

Journal Article↗

PubMed2026

Silent crises in the COVID-19 pandemic shadow: a six-year medico-legal review of non-lethal intimate partner violence in Morocco.

Intimate partner violence (IPV) is a major global public health concern and one of the most prevalent forms of gender-based violence, with significant consequences. This study analyzes non-lethal IPV cases reported to a medico-legal unit of Casablanca between 2018 and 2023. Data from 4,784 female victims were reviewed, focusing on socio-demographics, violence types, perpetrator relationships, and judicial involvement across the pre-COVID, peak COVID-19, and post-COVID periods. Married women represented 59.2% of victims. Physical violence predominated (83.1%), while sexual violence was underreported (14.7%). The year 2020 marked a sharp increase in reported non-lethal IPV cases (n = 1,005), correlating with pandemic confinement. Among married women, judicial requisition (legal request for forensic examination) and medico-legal certification (medical documentation for judicial use) rates were high (81.5% and 81.0%, respectively). Post-pandemic years declined moderately but remained above pre-COVID levels. International comparison confirmed a global rise in IPV during COVID-19, with Morocco displaying comparatively higher rates of medico-legal documentation and judicial involvement. IPV in Morocco is entrenched and intensified by COVID-19. While the medico-legal system is robust in documentation, the system must evolve by integrating systematic psychological assessment into medico-legal evaluations, standardizing certification and referral protocols, and strengthening coordination between medico-legal, health, and social support services to improve victim protection.

Humans↗

PubMed2026

Transcriptomic insights into the molecular mechanism of antifouling agent-induced settlement inhibition in the Pacific oyster Crassostrea gigas.

Marine biofouling remains a persistent challenge to maritime industries and marine ecosystems worldwide. In this study, we systematically evaluated the acute toxicity, settlement inhibitory efficacy, and underlying molecular mechanisms of an N-oleyl-1,3-propanediamine-based antifouling agent using pediveliger larvae of the Pacific oyster Crassostrea gigas. The 96 h-LC50 of the agent was determined to be 0.81 mg/L, and exposure to 1.68 mg/L achieved complete larval settlement inhibition without inducing significant acute toxicity. Transcriptomic analysis identified 791 differentially expressed genes, dominated by downregulated genes associated with ribosomal function, translation, cell adhesion, and cytoskeletal organization. The agent exerts its inhibitory effect primarily through the global suppression of protein synthesis, disruption of cell-substrate adhesion and cytoskeletal integrity, and induction of proteotoxic stress responses. These findings reveal a multi-pathway molecular mechanism underlying antifouling agent-induced settlement inhibition in oyster larvae and provide key molecular biomarkers to support the development of eco-friendly antifouling technologies.

Animals↗

PubMed2026

Antibiotic Self-Medication and Public Awareness During the 2023 Gaza War: A Cross-Sectional Investigation.

BACKGROUND: Antimicrobial resistance is driven by inappropriate antibiotic use, particularly self-medication. The 2023 Gaza War disrupted healthcare services, increasing resistant infections and reliance on self-treatment practices. OBJECTIVES: To determine the prevalence of antibiotic self-medication and assess public awareness during the 2023 Gaza War. METHODS: A cross-sectional survey was conducted between May and October 2025 among adults residing in the Gaza Strip during the 2023 Gaza War. A multistage non-probability recruitment strategy, organized across predefined governorate and residential-setting strata, was used to recruit 422 participants from primary healthcare centers, community pharmacies, and displacement shelters. The questionnaire demonstrated good internal consistency (Cronbach's α = 0.857). Ethical approval was obtained from the relevant institutional review board, and informed consent was secured from all participants. RESULTS: The overall KAP score was moderate (67.46%), with attitudes scoring highest (74.62%), followed by knowledge (66.23%) and practices (61.52%). Reported antibiotic self-medication increased from 60.4% before the war to 80.6% during the war (p < 0.001), with significant changes in reasons for self-medication and antibiotic procurement sources. The mean antimicrobial-resistance perception score was 82.19 ± 11.10%, and knowledge was strongly correlated with the total KAP score (r = 0.836, p < 0.001). CONCLUSIONS: Reported antibiotic self-medication was markedly higher during the 2023 Gaza War despite moderate public awareness, highlighting the urgent need for education, antibiotic stewardship, and improved healthcare accessibility.

Gaza War 2023↗

PubMed2026

Organizational bullying among nursing faculty: A systematic review of consequences, contributing factors, and interventions.

BACKGROUND: Organizational bullying among nursing faculty is a systemic and pervasive issue with profound psychological, professional, and institutional consequences. Despite increasing awareness, the literature remains fragmented, and a comprehensive synthesis is lacking. METHODS: This systematic review followed PRISMA guidelines and examined empirical studies on organizational bullying among nursing faculty, with no date restrictions applied. A structured search across five databases (PubMed, Scopus, Web of Science, CINAHL, and Embase) identified studies that met inclusion criteria related to the consequences, contributing factors, and interventions. The Mixed Methods Appraisal Tool (MMAT) was used to assess study quality. RESULTS: Fifteen studies meeting the quality threshold were included. Organizational bullying was consistently associated with psychological distress (e.g., anxiety, depression, burnout, suicidal ideation), professional disengagement, and intent to leave. Contributing factors were categorized into structural and organizational (e.g., hierarchical power imbalances, toxic workplace culture), managerial and HR-related (e.g., lack of leadership support, poor reporting mechanisms), and individual/social (e.g., gender or ethnic discrimination, early-career vulnerability). Intervention strategies identified included policy reforms, leadership training, supportive reporting systems, and psychological support services, though evidence on their effectiveness remains limited. CONCLUSIONS: Organizational bullying in nursing academia is a serious and multifaceted challenge with detrimental effects on individuals and institutions. Addressing it requires a comprehensive, evidence-based approach that includes structural reforms, leadership development, and psychosocial support systems. Academic institutions must prioritize the creation of safe, inclusive, and respectful environments to retain faculty and sustain the quality of nursing education.

Faculty, Nursing↗

PubMed2026

Nurse-led titration models of care for heart failure reduced ejection fraction: a systematic narrative review of characteristics, patient outcomes, and healthcare resource utilization.

AIMS: Nurse-led titration (NLT) models of care assist with delivery of guideline directed medical therapy for patients with heart failure with reduced ejection fraction (HFrEF). Effectiveness of NLT is established but there is limited information of characteristics of models, patient outcomes and healthcare resource utilization. To build upon the existing evidence by providing a systematic narrative review of the literature of NLT of medications for patients with HFrEF. This review syntheses characteristics of NLT models of care, patient outcomes and healthcare resource utilization. METHODS AND RESULTS: A systematic narrative literature review with systematic search strategy, identification of results, thematic analysis and narrative synthesis. A search was conducted from 2012 to 2025 in Medline, Cinahl complete, Embase and Cochrane. Sixteen studies of NLT models of care were identified from 1944 screened records. Characteristics of models of care were participation of nurses, multidisciplinary teams, follow-up and common features of service delivery. Patient outcomes of mortality were favourable for those that received NLT. There is some evidence of changes in healthcare resource utilization; studies in which the NLT groups received more HF nurse visits and greater HF medication use also reported reduced rehospitalizations. CONCLUSION: Findings reinforce the published benefits of NLT. Additional studies examining adverse events and quality-of-life outcomes are needed to strengthen the evidence base. Several studies suggest a shift in resource use with NLT, highlighting the need for an economic evaluation to inform a cost-effective model of care.

Humans↗

PubMed2026

Health and Physical Activity Outcomes in Age-Friendly Cities and Communities: A Systematic Review of Emerging Evidence and a Future Research Agenda.

OBJECTIVES: The World Health Organization's (WHO) Global Network of Age-Friendly Cities and Communities (AFCCs) promotes the development of urban environments, policies and services that support the health and participation of older adults. This systematic review examined contemporary evidence concerning associations between WHO AFCC conditions and directly measured health and physical activity outcomes among older residents. METHODS: The registered review adhered to the PRISMA protocol for systematic reviews and meta-analyses and applied the Downs and Black quality criteria for randomised and non-randomised research. RESULTS: Structured Boolean searches of five research repositories identified 17 peer-reviewed studies published between 2017 and 2025 based upon original research conducted in WHO AFCC signatory cities. Although most studies reported positive associations between age-friendly features and domains, such as accessible transport, walkable environments, outdoor infrastructure and self-rated health or physical activity, the strength of evidence was limited by methodological inconsistency, variable study quality and reliance on self-reports. Barriers to evaluation included limited use of longitudinal or quasi-experimental designs, heterogeneous outcome measures, subjective response data and the challenge of establishing appropriate comparison conditions in complex municipal settings. CONCLUSIONS: Strengthening evaluation frameworks for AFCC initiatives is essential for evidence-based urban health policy and governance in rapidly ageing societies. A research agenda is proposed to strengthen AFCC evaluation through standardised measurement, community-based and mixed-methods research, and a greater commitment to co-designed assessment frameworks.

Humans↗

PubMed2026

Status of dementia care among healthcare practitioners in Nigerian tertiary hospitals: a cross-sectional study.

BACKGROUND/OBJECTIVES: Dementia is an escalating public health concern globally. This study evaluated the knowledge, attitudes, practices, and perceived barriers to dementia care among healthcare practitioners in Nigerian tertiary hospitals, aiming to identify practitioner-related sociodemographic predictors and systemic barriers affecting dementia care delivery. METHODS: We collected data from May 2024 to May 2025 for this cross-sectional study in 12 purposively selected tertiary hospitals across Nigeria's six geopolitical zones. Participants included physicians, nurses, pharmacists, and other professionals involved in geriatric psychiatric care. Using multistage and convenience sampling, 394 respondents were recruited (response rate: 99.5%). Data were collected via a validated Dementia Care Practice Questionnaire (Cronbach's α = 0.84) and analyzed with SPSS v22. Descriptive statistics, Chi-square tests, and odds ratios (ORs) identified associations (significance: p ≤ 0.05). RESULTS: Of 394 respondents, 51.5% were aged ≥40 years, and 54.8% were female. While 62.9% demonstrated adequate knowledge, negative perceptions (51.3%) and attitudes (56.9%) were common. Despite this, 71.3% reported engagement in dementia care, and 75.6% demonstrated appropriate professional help-seeking behaviour when confronted with dementia care challenges. Practitioner-reported barriers included limited training opportunities, geographical barriers affecting patient access to dementia services, and inadequate staffing. Predictors of desirable care practices among healthcare practitioners included age ≥40 years, female gender, Christian affiliation, and ≥5 years of professional experience. CONCLUSION: Although many healthcare practitioners are involved in dementia care, gaps in perceptions, attitudes, and structural support persist. Interventions should focus on targeted training, system strengthening, and policy reform to improve dementia care outcomes.

Barriers to care↗

PubMed2026

ORBIT: Oncogenic Representation Learning via Bi-Prototype Contrastive Learning in Hyperbolic Space for cancer driver gene identification.

Accurate identification of cancer driver genes is crucial for precision oncology but remains challenging due to the complexity of integrating heterogeneous data and modeling dynamic biological systems. To address these limitations, we propose ORBIT (Oncogenic Representation Learning via Bi-Prototype Contrastive Learning in Hyperbolic Space). Our framework synergistically fuses multi-omics profiles with functional network data using a context-adaptive graph reweighting mechanism to capture cancer-specific dynamics. The model employs a bi-prototype contrastive learning strategy within hyperbolic space, which aligns gene representations around distinct driver and non-driver semantic anchors while preserving the intrinsic hierarchy of biological networks. Comprehensive evaluations demonstrate that ORBIT achieves highly competitive stability in pan-cancer analysis while consistently outperforming state-of-the-art methods in cancer-specific predictions. Furthermore, functional enrichment analysis confirms that the model effectively segregates core cancer pathways, and drug sensitivity profiling validates the clinical relevance of the identified drivers. By integrating hyperbolic geometry with context-adaptive learning, ORBIT offers a robust and interpretable paradigm for precision medicine. The source codes and datasets are publicly accessible at https://github.com/spcho-dev/ORBIT.

Humans↗

PubMed2026

Applications and outcomes of virtual reality in inpatient psychiatry: A systematic review.

BACKGROUND: Virtual reality (VR) has been widely used in outpatient psychiatric services and has demonstrated benefits across several clinical diagnoses, but its use and effects in inpatient settings remain to be explored. This systematic review aimed to examine the use of VR during psychiatric hospitalization, including types of VR applications, barriers and facilitators of implementation, and effects on various outcomes. METHODS: The review was registered in PROSPERO (#CRD42023446524). Following PRISMA guidelines, databases (Ovid, SciVerse, Web of Science, Cochrane Library, ProQuest, and WorldCat) were searched from 1983 to 2025 using keywords related to VR and psychiatric disorders. Studies involving the use of VR with psychiatric inpatients (≥85%) were included. Descriptive statistics and narrative syntheses were used to summarize findings. Study quality was assessed with the Mixed Methods Appraisal Tool. RESULTS: After full-text screening, 37 studies (N = 1,004) met inclusion criteria. VR was used for both assessment and intervention, with cognitive-behavioral therapy/exposure (35%) and assessment (24%) being the most frequently used. VR use in inpatient units appeared feasible, acceptable, and safe for inpatients and clinicians, though findings remain preliminary. Several facilitators (e.g. adequate staff training and supervision) and common barriers (e.g. technical difficulties and limited resources) were identified. The most consistent improvements were observed in clinical symptoms (e.g. anxiety) compared with psychosocial, cognitive, and physiological outcomes. CONCLUSIONS: These findings suggest that inpatient settings represent a promising, yet understudied context for VR-based assessments and interventions. High-quality trials and systematic reporting of implementation are needed in future studies to inform research and clinical practice.

Humans↗

PubMed2026

Surveilled subjectivation: narratives of drug policing among people who use prohibited drugs in Sweden.

In Sweden, possession and personal use of drugs are criminalized since 1988, resulting in police work being directed towards minor drug offenses. Despite this, police and other authorities are encouraged to protect the health and wellbeing of people who use prohibited drugs (PWUPD). Knowledge is scarce on how this drug policy plays out in practice. This study therefore analyzes interviews with 20 PWUPD who visited harm reduction services and interacted with policing agents in Stockholm, Sweden. The analysis is based on the participants' narratives of drug policing, and it concerns how they produced themselves as subjects through relations between materiality and discourse. We utilize the concept of surveilled subjectivation to elucidate what the participants could do, what they knew and who they could be or become under drug policing. Four themes were identified illustrating the link between discourse and materiality in PWUPD's surveilled subjectivation: "Material aspects of surveillance"; "Resisting the 'drug abuser' identity"; "Fighting power with power"; and "Crossing boundaries and becoming-other". The participants described nonstop efforts to prevent their bodies, activities, belongings and environments from being enfolded by drug law enforcement, which otherwise would fuel even more surveillance. They therefore disassociated themselves from the "drug abuser" identity, and managed encounters with policing agents by keeping a low profile or acting compliantly. While the study highlights the skills and knowledges the participants deployed to navigate omnipresent drug policing, we conclude that their production of autonomous and empowered subjectivities would be facilitated if possession and use of drugs were no longer criminalized.

Humans↗

PubMed2026

Improving insurance deduction identification: a hybrid artificial intelligence model using machine learning and expert systems.

PURPOSE: Financial challenges in healthcare systems worldwide, especially in low- and middle-income countries like Iran, have increased hospitals' reliance on insurance reimbursements. Unrecognized insurance deductions often cause severe financial shortages, making efficient deduction management crucial. This study aimed to design a hybrid intelligent system for identifying and predicting insurance deductions by combining machine learning and expert system frameworks. DESIGN/METHODOLOGY/APPROACH: A mixed-methods design was applied in four stages. First, a scoping review identified the causes and patterns of insurance deductions. Second, interviews with 15 insurance experts produced a validated checklist and a dataset from inpatient billing records. Third, using the CRISP-DM methodology, machine learning algorithms were developed and tested in SPSS Modeler alongside a fuzzy expert system developed in MATLAB. Finally, the model was validated using the holdout method. FINDINGS: Four categories of deduction drivers were identified: service provision, registration errors, document submission issues, and revenue conversion processes. The CHAID decision tree outperformed other algorithms with a 99% precision rate and the lowest Mean Absolute Error (9.43). A brief assessment of potential overfitting was conducted to ensure that the CHAID model's high accuracy was interpreted cautiously and supported by the validation results. The fuzzy expert system with validated rules was adaptable for deduction classification, especially for cases unsuitable for quantitative modeling. ORIGINALITY/VALUE: The hybrid model improves detection and prevention of deductions, offering actionable insights for hospital administrators, insurers, and policymakers. Its implementation can enhance hospital information systems, streamline claims processing, and optimize revenue management amid financial constraints.

Machine Learning↗

PubMed2026

Understanding psychosocial adjustment in military-to-civilian transition: A latent profile analysis of ex-serving Australian Defence Force members.

Military-to-civilian transition is a critical life stage that can expose veterans to elevated risks of psychological distress, social difficulties, and reduced wellbeing. Although psychosocial factors are central to successful reintegration, little is known about distinct patterns of needs among ex-serving Australian Defence Force (ADF) members. This study used latent profile analysis (LPA) to identify psychosocial needs profiles across five domains of the Military-Civilian Adjustment and Reintegration Measure (M-CARM) in a sample of 725 ex-serving ADF members. The optimal three-class solution identified: a Low Adjustment Need (LAN) group (20.7%) reporting minimal reintegration challenges; a Cultural Adjustment Need (CAN) group (42.9%) characterized by cultural adaptation difficulties, particularly beliefs about civilians and regimentation; and a Cultural and Psychological Need (CPN) group (36.4%) showing broader challenges across beliefs about civilians, purpose and connection, regimentation, and resentment and regret. The CAN group was more likely to be male, have lower educational attainment, and have no combat deployment history. The CPN group was similarly male dominated with lower education, and was additionally characterized by Navy service, unemployment or not being in the labor force, and medical discharge. Compared with the LAN group, both CAN and CPN groups reported higher levels of depression, anxiety, posttraumatic stress, and nightmare distress, as well as poorer quality of life and greater functional impairment. These findings highlight persistent reintegration challenges among veterans and support the need for stratified support models, ranging from psychoeducation to intensive multidisciplinary care, to better address diverse psychosocial needs of ex-serving ADF members.

Australian defense force↗

PubMed2026

Automated CEAP Classification of Venous Duplex Reports Using Multimodal Artificial Intelligence.

OBJECTIVE: To develop and internally validate a prototype multimodal artificial intelligence system for automated CEAP (Clinical, Etiological, Anatomical and Pathophysiological) classification of venous duplex ultrasound (VDUS) reports, integrating natural language processing of free-text components with computer vision analysis of hand-drawn anatomical diagrams. METHODS: Single centre retrospective observational study using routinely collected clinical data. One thousand consecutive venous duplex ultrasound reports from Cambridge University Hospitals NHS Foundation Trust, UK (July 2024 - May 2025) were labelled according to the CEAP classification, excluding the Etiological component, which could not be reliably determined from duplex reports alone. Transfer learning was applied using ClinicalBERT for text and MobileNetV3 for diagrammatic data. Clinical classes were predicted from request line text. Text- and image-based pathophysiological models were developed for four anatomical territories (Great Saphenous Vein, Small Saphenous Vein, Deep system, Perforators), combined using late fusion with probability averaging. RESULTS: The clinical CEAP model achieved accuracy of 0.91, macro-F1 of 0.82, and macro-AUC of 0.98. Pathophysiological prediction varied, with text models broadly outperforming image models. Fusion yielded heterogeneous benefits, improving SSV performance but reducing Deep system accuracy. The performance of the final pathophysiological CEAP fusion models varied across anatomical territories: accuracy ranged from 0.70-0.92 and macro-AUC from 0.80-0.92. CONCLUSION: This study demonstrates the feasibility of automated CEAP classification from VDUS reports. Despite class imbalance affecting minority class predictions, the strong discriminatory performance validates this multimodal ML model for extracting clinically meaningful information from real-world data. This approach offers potential, pending external validation, to streamline vascular services through automated triage and guideline-compliant decision making.

Artificial intelligence↗