Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “machine learning models”

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.

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

885 records · Page 10Linked to original sources
PubMed2026

Risk prediction models for blood transfusion in patients undergoing total hip and knee arthroplasty: a systematic review and meta-analysis.

OBJECTIVE: To systematically review and evaluate published risk prediction models for perioperative blood transfusion in patients undergoing total hip or knee arthroplasty (THA/TKA). METHODS: We systematically searched PubMed, Web of Science, the Cochrane Library, and Embase from inception to May 31, 2025. Two researchers independently screened the literature, extracted data, and assessed the risk of bias and applicability using the Prediction model Risk Of Bias Assessment Tool (PROBAST). The area under the receiver operating characteristic curve (AUC) values were pooled via a meta-analysis using Stata 18.0. RESULTS: d Fourteen studies containing 36 prediction models were included. The incidence of blood transfusion among THA/TKA patients ranged from 3.2% to 30.8%. Preoperative hemoglobin (Hb) level, tranexamic acid (TXA) use, operative duration, intraoperative blood loss, and age were the most frequently incorporated predictors. Model sensitivity ranged from 58% to 94.5%, and specificity ranged from 71.3% to 94%. Meta-analysis showed that the pooled AUC value of the 13 validated models was 0.87 (95% CI: 0.85-0.90), suggesting good discriminatory performance. All models were rated as having a high risk of bias. The applicability of four studies was rated as unclear. CONCLUSION: Although the included studies demonstrated promising discriminative ability of prediction models for blood transfusion in THA/TKA, all were assessed as having a high risk of bias using the PROBAST tool. Therefore, future research should prioritize the development of models with larger sample sizes, rigorous study designs, and multicenter external validation.

Humans↗
PubMed2026

Development and Validation of a Predictive Model for Identification of Cognitive Impairment Risk in Older Adults with Subjective Cognitive Decline：A Longitudinal Study.

BACKGROUND: Subjective cognitive decline (SCD) is a transitional state between objective cognitive impairment and cognitively intact mental status, providing a critical window for implementing preventive interventions to delay objective cognitive decline. AIMS: We aimed to develop a predictive model for SCD progression in older adults with mild cognitive impairment (MCI). This model will facilitate the identification of risk factors and establishment of targeted interventions for community-based SCD management. METHODS: Data from the China Health and Retirement Longitudinal Study (CHARLS) was utilized in this study, extracting 18 indicators. Potential predictors selected through univariate Cox regression and LASSO regression analyses were sequentially incorporated into a multivariable Cox regression model. A nomogram was constructed to establish a predictive model. Model validation encompassed Area Under Curve (AUC) metrics for discriminative capacity, complemented by quantitative assessments using calibration curve analysis for precision verification and decision curve analysis (DCA) for clinical utility evaluation. RESULTS: A total of 1099 older adults with SCD were included in the final analysis, of whom 114 (10.3%) developed MCI. Multivariable Cox regression identified residence, marital status, educational level, social participation, gait speed, and baseline cognitive function. The model demonstrated time-dependent AUC values of 0.885, 0.830, 0.839, and 0.836 in the training set when evaluating discriminative capacity at 2-, 4-, 7-, and 9-year, respectively. The predictive model showed excellent predictive ability according to AUC, calibration curve, and DCA. CONCLUSIONS: A predictive model was created to estimate the risk of developing MCI in older individuals with SCD, offering clinician-actionable intervention benchmarks for preventive care.

Humans↗
PubMed2026

Validated UPLC-MS/MS quantification and intracellular PK-PD Modeling of periplocin-related cardiac glycosides in H/R-injured H9c2 cells.

Reliable intracellular quantification is essential for characterizing the target-site disposition and exposure-response relationships of bioactive natural products. In this study, an ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) method was developed and validated for the simultaneous determination of periplocin and four related cardiac glycoside metabolites in H9c2 cell lysates. Acceptable linearity, precision, recovery, and stability were achieved for intracellular quantification. Cells were treated with each compound at 50 μM, and intracellular concentrations and cell viability were monitored over 48 h. In hypoxia/reoxygenation (H/R) -injured cells, the time to maximum intracellular concentration was shortened for all five compounds, indicating altered cellular disposition under injury conditions. Cell viability was improved by all compounds during the observation period. Pharmacokinetic-pharmacodynamic (PK-PD) integration was performed using a sigmoid Emax model, and acceptable model fits were obtained, with Akaike information criterion (AIC) values ranging from 79.22 to 130.46. Low apparent EC50 values were estimated under this single-dose design, whereas the estimated Ke0 values suggested delayed equilibration with the effect compartment. These findings indicate that sustained cytoprotective responses can be produced by periplocin and related metabolic markers in injured cardiomyocytes. This intracellular bioanalytical strategy provides a quantitative approach for linking cellular exposure to pharmacodynamic response and may support further evaluation of periplocin-related cardiac glycosides.

Tandem Mass Spectrometry↗
PubMed2026

Multiscale Modeling Primer: Focus on Chromatin and Epigenetics.

A central challenge in modern biology is to understand how molecular interactions produce cellular and organismal functions across vast spatiotemporal scales. Nowhere is this challenge more apparent than in the study of chromatin, where meters of DNA compact into a micron-sized nucleus. How this polymer folds is a dynamic process, regulated by epigenetic modifications-chemical changes to DNA and histones that involve only a handful of atoms. These small changes cooperate to produce emergent, higher-order structures that define cellular identity and function. To explain this system, we must integrate static, high-resolution snapshots from techniques like cryo-EM with dynamic, lower-resolution data from microscopy and genomics. Multiscale computational models are essential tools that bridge these experimental gaps and reveal the mechanisms of emergent behavior. However, the communication divide between experimental biologists and quantitative modelers often hampers progress. This primer addresses that gap. It first introduces the fundamental biology of chromatin and epigenetics at an introductory level for non-biologists audiences. We then survey the landscape of computational approaches, from atomistic to systems-level models, and connect them to the experimental data that inform and validate them at an introductory level for non-computationalists. We argue that the next frontier will require us to build integrative models that can predict how molecular perturbations mechanistically alter cellular phenotypes, which will open a new era of chromatin-targeted therapeutics.

Chromatin Dynamics↗
PubMed2026

A Pilot Study: Developing a Lactating Dairy Goat Model to Study Staphylococcus aureus Mastitis in Women.

INTRODUCTION: Lactational mastitis is common in lactating women, with Staphylococcus aureus as the most commonly isolated agent associated with infectious lactational mastitis. Currently, there are no evidence-based guidelines for antimicrobial treatment due to barriers in obtaining pharmacokinetic data from lactating women. To overcome this barrier, a suitable large animal model is needed. Goats are an ideal translational model for human mastitis due to their anatomical and physiological similarity to humans. The objective of this pilot study was to assess if goats would develop clinical mastitis following intramammary inoculation with a clinical human isolate of S. aureus with the goal of establishing an alternative in vivo model for future research. The hypothesis was that the infected mammary gland half would show similar clinical signs to women with mastitis and demonstrate a similar local immune response when compared to the control mammary gland half. METHODS: One half of the mammary gland of two healthy lactating does was inoculated with a clinical human isolate of S. aureus. The other half of the mammary gland was sham inoculated with sterile buffered saline. Physical examinations, mammary gland assessments, and sterile milk samples were collected every 12 hours post inoculation. At 96 hours post inoculation, the goats were euthanized, and the mammary glands were examined for pathological changes. RESULTS: Goats did not develop systemic signs of disease following inoculation. Focal infected mammary gland changes included warmth, swelling, redness, discoloration, and reduced milk production; the other mammary gland half remained normal throughout the study period. S. aureus was enumerated from only the infected mammary gland half. The microscopic findings of the infected half showed neutrophilic inflammation and cell necrosis consistent with acute mastitis. DISCUSSION: This pilot study demonstrated lactating does can develop clinical signs like those observed in women. Goats have the potential to be a promising animal model to study infectious lactational mastitis.

Animals↗
PubMed2026

How the Social Context and Peer Helping Contribute to Better Alcohol Outcomes Among Sober Living House Residents: Mediation Analyses.

BACKGROUND: Sober Living Houses (SLHs) adopt a social model approach, which emphasizes peer helping. Although SLHs appear to be effective, little is known regarding why. This longitudinal study examined whether higher SLH social model adherence produces better resident outcomes by increasing resident helping. METHODS: Baselines were conducted with 205 residents entering 28 SLHs, with follow-ups through 6 months. Measures included 1-month perceived SLH social model adherence; 2-month help given to and received from SLH residents; and 6-month alcohol use and severity. Analyses were lagged, multivariate mediation models accounting for clustering within SLH. Separate models examined help given and received for each outcome, yielding four model tests. RESULTS: The hypothesized model was unsupported, with all four tests showing nonsignificant indirect effects. However, exploratory post-hoc tests showed significant indirect effects between higher 1-month resident helping and better 6-month alcohol outcomes via higher 1-month SLH social model adherence. Effects held across three of four model tests. CONCLUSIONS: Results suggest that residences adhering to social model principles do not achieve better outcomes by stimulating helping, but rather that more resident helping may foster a supportive SLH social environment, which itself drives better outcomes. Thus, residences might emphasize both resident helping and social model principles.

Sober living↗
PubMed2026

Review: The African turquoise killifish as a model for the integrative physiology of vertebrate aging.

With increasing emphasis on extending healthy lifespan, aging research requires vertebrate models that permit efficient mechanistic investigation and intervention testing within practical time and cost constraints. The African turquoise killifish (Nothobranchius furzeri) has attracted growing attention because it combines an exceptionally short life cycle with an intact vertebrate physiological context and an expanding genetic toolkit, enabling relatively rapid evaluation of candidate aging interventions and mechanistic analysis across molecular, tissue, and organismal levels. This review assesses N. furzeri from an integrative-physiology perspective, focusing on germline-soma interactions, gut microbiota-host crosstalk, nutrient sensing and metabolic remodeling, temperature responsiveness, and AMPK-mTOR-linked programs. It also examines expanding genome-engineering and reporter approaches that support mechanistic and tissue-resolved investigation of these physiological processes. Building on recent reviews of killifish biology, disease modeling, regeneration, and the hallmarks of aging, we synthesize evidence across major intervention domains, distinguish established phenotypic effects from incompletely resolved mechanisms, and highlight functional endpoints, methodological standardization, and the appropriate interpretation of the model's translational relevance. Together, these features position N. furzeri as a strategically useful vertebrate platform for rapid mechanistic testing, intervention evaluation, and prioritization of aging-related pathways. Future progress will require improved methodological standardization, tissue-resolved causal studies, and question-driven cross-species validation where appropriate.

Animals↗
PubMed2026

Efficacy of pharmacological and microbiota-based therapies in preclinical models of autism spectrum disorder: a systematic review.

BACKGROUND: Autism spectrum disorder (ASD) is a multifactorial neurodevelopmental condition in which pharmacological and microbiota-targeted interventions are emerging as promising therapeutic avenues. Animal models are the main tool to investigate etiology, molecular mechanisms and screening for pharmacological therapies. Methodological differences, outcome measure variability, incomplete reporting, biological confounders, and overgeneralization of the results made evaluating innovative pharmacological agents challenging. These limitations in the field highlight a need for systematic and standardized research to reliably assess and translate pharmacological interventions from ASD animal models to human clinical relevance. SUBJECTS: This systematic review synthesized efficacy evidence for pharmacological and microbiota-based therapies across established ASD animal models. RESULTS: We identified 52 recent (2010-2025) studies that reported key ASD behavioral outcomes after pharmacological or microbiota-focused treatments. Interventions were grouped into therapeutic classes - including oxytocinergic agents, E/I balance therapeutic targets, metabolic drugs, cannabinoids, purine-based interventions and emerging targets - alongside microbiota-directed strategies such as probiotics, prebiotics, and fecal microbiota transplantation. By integrating effect directions and robustness across models, we identified most potential drug candidates, evaluated the efficacy of novel strategies, and recognized critical translational gaps. The reviewed studies demonstrate that ASD-like behavioral deficits in preclinical models can be modulated through interventions targeting diverse biological systems, including neurotransmission, neuroinflammation, metabolism, and the gut-brain axis. CONCLUSIONS: These findings support the multifactorial nature of ASD pathophysiology which arises from a network of interacting systemic processes rather than a single molecular defect. It could explain the limited success of traditionally narrowly targeted interventions and suggest a paradigm shift into a more systemic approach.

Animals↗
PubMed2026

Nimodipine in animal models of demyelination relevant to multiple sclerosis: a systematic review.

BACKGROUND: Multiple sclerosis (MS) is the most common inflammatory neurodegenerative disease in which axonal injury, neuronal death, and demyelination occur. Treatment for MS relapses remains limited, which alleviates acute loss of function but has no impact on long-term disability. This study aimed to perform a systematic review of the effects of nimodipine on experimental demyelination models, including experimental autoimmune encephalomyelitis (EAE) and Cuprizone models in rodents. METHODS: This study was conducted following the PRISMA statement. A systematic search was performed in PubMed, Scopus, the Cochrane Library, and Google Scholar. The primary outcome was EAE clinical disease severity (peak clinical score and/or cumulative disease burden). Secondary outcomes included relapse activity (when reported), histological myelin outcomes, oligodendrocyte lineage markers, neuroaxonal injury markers, and inflammatory readouts. Risk of bias was assessed using the SYRCLE tool. RESULTS: Out of 4660 results, 5 studies were included in the systematic review (four EAE studies and one cuprizone model). Nimodipine was administered using heterogeneous regimens (oral, intravenous, intraperitoneal, subcutaneous, or osmotic pump delivery; 1-30 mg/kg/day). The included studies reported the variable effects of nimodipine on relapse-related outcomes, myelination, inflammatory processes, and neuroprotection in the EAE model of MS. Across EAE studies, nimodipine generally reduced clinical disease severity or cumulative burden, although relapse-related outcomes were inconsistent. CONCLUSIONS: Preclinical evidence suggests that nimodipine may attenuate disease severity and demyelination and may promote repair-related processes in rodent models relevant to MS. However, to evaluate the clinical applicability of nimodipine in MS patients, well-powered, transparently reported preclinical replication and early-phase clinical studies are required before clinical translation.

Animals↗
PubMed2026

Modelling peak microbial pollution events caused by combined sewer overflows in a source-to-sea system.

Predicting peak microbial pollution events in downstream coastal bathing waters caused by combined sewer overflows (CSOs) is essential for protecting public health. In urban areas, wastewater effluents, CSOs, and surface runoff can contribute to elevated microorganism loads to downstream waters. These pressures are likely to be intensified by growing population density and more frequent heavy rainfalls due to climate change. This study developed a process-based model to simulate Escherichia coli (E. coli) emissions, transport, and fate from the initial sources to coastal beaches. A three-year retrospective simulation (2017-2019) shows that E. coli concentrations in CSO discharges varied widely across the catchment (4.6 - 7.3 (log10 CFU 100 ml-1)). 99th percentile E. coli concentrations (4.0 (log10 CFU 100 ml-1)) at the inland water outlet were dominated by local CSO emissions, whereas 90th percentile E. coli concentrations (3.6 (log10 CFU 100 ml-1)) reflected cumulative upstream contributions from both CSO and effluent emissions. With the simulation accuracy of 89%, the model reliably reproduced the E. coli dynamics on the downstream beach and showed strong performance in representing peak concentrations based on Complementary Cumulative Distribution Function (CCDF) analysis. The process-based model enables quantitative tracking of source contributions and identification of pollution hotspots, providing support for mitigation measures. The study lays down a source-to-sea modelling framework for representing pollution transport across the aquatic continuum and provides a transferable tool for microbial pollution forecasting and climate adaptation planning.

Climate projection↗
PubMed2026

Genomic characterization and pathogenicity of ruminant Listeria monocytogenes isolates in a murine oral infection model.

Listeria monocytogenes is a major foodborne pathogen; its ruminant isolates display zoonotic characteristics, causing similar clinical signs in humans, including abortion and encephalitis. However, data on whole genome sequencing and pathogenicity of ruminant L. monocytogenes isolates remain sparse. This study aimed to analyze the genotypic characteristics of L. monocytogenes isolates from ruminants with listeriosis. Furthermore, we assessed the in vivo pathogenicity of four ruminant L. monocytogenes isolates, characterized via whole-genome sequencing-based genetic clustering, in orogastrically inoculated mice. The isolate LM18 (serotype 1/2b, ST224, SL6178) had the lowest lethal dose compared to the other three isolates including previous hypervirulence type (serotype 4b, ST1, SL1) and caused secondary bacteremia in lungs, with sustained bacterial loads in the spleen and liver. Genomic (listeria pathogenicity island -1 and -3) and virulence gene (actA and llsX) mutation analyses associated with virulence suggested from well-recognized studies could not elucidate the virulence of the isolates. SSI-1, which only exists in the isolate LM18 (serotype 1/2b, ST224, SL6178), may help L. monocytogenes survive in the gastrointestinal environment, thereby affecting its virulence. Further research should investigate the role of SSI-1 in the pathogenicity of L. monocytogenes. Moreover, additional studies utilizing larger datasets of ruminant isolates are required to validate our genotypic characterization and to obtain a comprehensive picture of further genotypic differences crucial for L. monocytogenes pathogenicity.

Animals↗
PubMed2026

Predictive Models for Hypoglycemia Risk in Haemodialysis Patients With Diabetic Kidney Disease: Systematic Review and Meta-Analysis.

AIM: To provide evidence for selecting and developing reliable clinical assessment tools for hypoglycemia in diabetic kidney disease patients during haemodialysis. DESIGN: Review. METHODS: Systematic searches were performed in 9 Chinese and English databases to collect literature regarding the development of hypoglycemia risk prediction models in haemodialysis patients with diabetic kidney disease. Two reviewers independently performed literature screening, data extraction, risk-of-bias assessment, and applicability evaluation. The Prediction Model Risk of Bias Assessment Tool was used to assess the risk of bias and applicability of the included studies. Meta-analysis was conducted using R software. DATA SOURCES: CNKI, Wanfang, VIP, CBM, PubMed, Cochrane Library, EMbase, Web of Science, and CINAHL. The search period covered from the establishment date of each database to December 2025. RESULTS: Six studies, comprising six prediction models, were included. Two studies performed internal validation, and three conducted external validation. All models reported the area under the curve, ranging from 0.813 to 0.866, and calibration measures. Four studies were rated as having a high risk of bias, while all six demonstrated good overall applicability. The meta-analysis showed that the pooled AUC value of the six studies was 0.846 (95% CI: 0.823-0.867). CONCLUSION: Research on hypoglycemia risk prediction models in haemodialysis patients with diabetic kidney disease remains in the developmental stage. Although the included prediction models exhibited satisfactory apparent discriminatory ability and clinical applicability, most of the original studies suffered from a high risk of bias and lacked adequate validation. The true predictive performance and clinical application value of these models remain to be further verified. Accordingly, routine and unconditional clinical application is not recommended at this stage. Future studies should include more high-quality, multicenter external validation and develop models with high generalizability, favourable clinical applicability, and robust predictive performance to facilitate early identification of hypoglycemia risk in this population. IMPACT: This study systematically evaluated the hypoglycemia risk prediction models for diabetic kidney disease patients during haemodialysis, and the research on hypoglycemia risk prediction models for maintenance haemodialysis patients during dialysis is still in the development stage. This study provides a reference for clinical medical staff to select or develop hypoglycemia risk prediction and assessment tools for diabetic kidney disease patients during haemodialysis. REPORTING METHOD: This study was conducted in accordance with the relevant guidelines of the EQUATOR Network and followed the TRIPOD-SRMA Checklist. PATIENT OR PUBLIC CONTRIBUTION: No patient or public contribution. TRIAL REGISTRATION: PROSPERO: CRD420251243352.

Humans↗
PubMed2026

Robust optimisation for photon radiotherapy: A scoping review of models, paradigms, and reporting.

BACKGROUND AND PURPOSE: Robust optimisation offers an alternative to conventional margin-based photon radiotherapy planning by explicitly modelling uncertainty, but practice is variable and not standardised. MATERIALS AND METHODS: A scoping review was conducted to map robust optimisation for photon external beam radiotherapy. Electronic searches of Scopus, PubMed and Google Scholar (2000-2025, English language) identified planning studies that incorporated modelled uncertainties into the optimisation process and reported at least one robustness-related outcome. Data were charted on clinical context, uncertainty models, optimisation paradigms, robustness metrics and evidence for clinical implementation. RESULTS: Seventy-one studies were included. Most investigated prostate, breast or lung cancer and used intensity-modulated radiotherapy or volumetric-modulated arc therapy in commercial or research treatment planning systems. Scenario-based worst-case (minimax) optimisation was the dominant paradigm in clinically oriented work, while chance-constrained, conditional value at-risk, distributionally robust and adaptive formulations were confined to small methodological series. Uncertainty modelling focused mainly on rigid set-up error; fewer studies incorporated respiratory motion, inter-fraction anatomical change, dose-calculation uncertainty or biological variation. Robustness was evaluated with diverse scenario-based dose-volume metrics, probabilistic coverage measures, composite robustness indices and, less often, biological endpoints. Direct clinical implementation reports were scarce. CONCLUSION: Robust photon planning is technically feasible and generally maintains or improves target coverage and organ sparing compared with margin-based planning. However, heterogeneity in uncertainty models, optimisation configuration and robustness reporting limits comparison and synthesis. Pragmatic minimum standards are proposed to support future consensus and wider clinical adoption.

Humans↗
PubMed2026

Metagenome-scale modeling to assess microbiome metabolic complementarity for precision microbiota transplantation therapies.

Fecal microbiota transplantation (FMT) holds therapeutic promise beyond recurrent Clostridioides difficile infection, but clinical outcomes remain unpredictable and donor-selection strategies remain limited, in part because the role of donor‒recipient metabolic interactions in shaping the post-FMT community remains poorly understood. Here, we leverage metagenome-scale metabolic modeling to quantify metabolic niche complementarity between donor and recipient microbiomes and predict post-FMT community composition. Using MICOM-derived metabolic models, we show that donor genomes whose metabolic flux profiles are more dissimilar from the recipient community colonize at significantly higher rates in a murine FMT model. In a human IBS trial, the same metric predicted post-FMT community composition via leave-one-out cross-validation and captured known disease-associated alterations in short-chain fatty acid, sulfur, and gas metabolism. We then performed 2,548 in silico FMT simulations between IBS-D/M patients and donors from the OpenBiome biobank to evaluate personalized donor screening, identifying super-donors characterized by high taxonomic diversity, broad metabolic niche coverage, and community interaction networks dominated by cross-feeding rather than competition. Together, these results support metabolic niche complementarity as a potential determinant of post-FMT community composition and provide a mechanistic basis for evaluating donor-recipient metabolic compatibility. This framework offers a scalable approach for generating testable hypotheses for personalized donor selection.

Fecal Microbiota Transplantation↗
PubMed2026

Meta-analysis of source identification and apportionment in soil: A systematic review of analytical procedures, receptor modeling, and environmental applications.

Soil pollution poses significant risks to ecosystems and human health, necessitating accurate source identification and apportionment to guide mitigation strategies. This systematic review evaluates the application of Positive Matrix Factorization (PMF) and other receptor models in soil pollution studies, focusing on analytical procedures, tracer indicators, and environmental applications. This review aims to provide a comprehensive framework for conducting soil source apportionment studies, aiding policymakers in designing effective, region-specific environmental management strategies by compiling global trends and methodological insights. The study addresses sampling protocols, emphasizing representativeness and quality control. Data from 500 peer-reviewed publications highlight the dominance of research in China, Eastern Europe, and South Asia, with agricultural soils being the most frequently studied. Key findings reveal that traffic emissions (20.8 %) and industrial activities (19.4 %) are the primary global contributors to soil contamination, with regional variations such as coal combustion in cold climates and agricultural inputs in developing regions. Policy recommendations include stricter industrial regulations, sustainable agricultural practices, and targeted remediation efforts based on source-specific risks.

Soil Pollutants↗
PubMed2026

Externally validated risk prediction models for gestational diabetes mellitus: A systematic review and meta-analysis.

INTRODUCTION: Risk prediction models for gestational diabetes mellitus (GDM) offer potential for early identification and targeted prevention. External validation is crucial to assess model performance across diverse populations. Despite the availability of numerous GDM prediction models, limited evidence exists on their external validation frequency, methodological quality, and clinical applicability. This systematic review evaluated externally validated GDM prediction models, focusing on methodological rigor, reporting standards, and clinical relevance to inform future research and implementation. MATERIAL AND METHODS: Databases including Ovid MEDLINE, Embase, Scopus, Emcare, and CINAHL were searched up to May 1, 2025. Studies reporting external validation of GDM risk prediction models were included. Two reviewers independently screened studies. Data were extracted using the CHARMS framework, and risk of bias and applicability were assessed using PROBAST+AI. The study protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO; CRD420251125758). RESULTS: Twenty-six studies validated 33 models, with validation sample sizes ranging from 50 to 75 161. Over half used the IADPSG criteria to define GDM. Discrimination metrics were commonly reported, but calibration, overall performance, and clinical utility were often lacking. Meta-analysis was feasible for only four models: Teede et al., Nanda et al., Naylor et al., and Van Leeuwen et al., each showing fair discrimination. The Teede et al. model was the most widely validated, with 11 external validations across six continents and a pooled AUC of 0.72 (95% CI: 0.67-0.76). Despite fewer validations, the Nanda et al. model achieved the highest pooled discrimination (5 validations; pooled AUC 0.77, 95% CI: 0.74-0.80). The Naylor et al. and van Leeuwen et al. models also underwent meta-analysis, as sufficient external validation studies were available to support comparative performance assessment. Notably, 69.23% of studies had a high risk of bias. CONCLUSIONS: While many models showed acceptable predictive performance, most validations were methodologically weak. Future studies should follow best-practice guidelines and promote scalable validation strategies, such as algorithm sharing, to enhance clinical utility.

Humans↗
PubMed2026

Risk Factors and Predictive Model for Postoperative High Myopia in Children Undergoing Congenital Cataract Surgery With Intraocular Lens Implantation.

PURPOSE: To identify risk factors associated with the development of high myopia following congenital cataract surgery and to establish a robust predictive model. DESIGN: Retrospective clinical cohort study. SUBJECTS: This retrospective study included 106 pediatric patients who underwent congenital cataract surgery with primary IOL implantation (mean follow-up 8.19 years). The model was externally validated in an independent cohort of 72 patients with a mean follow-up of 7.83 years. METHODS: Preoperative and postoperative ocular biometric parameters were collected. Risk factors for postoperative high myopia were analyzed using Cox proportional hazards regression, which served as the basis for model construction. The predictive performance of the model was rigorously evaluated for discrimination and calibration. Discriminative ability was quantified using Harrell's C-index and the area under the receiver operating characteristic curve (AUC). Model calibration was assessed via calibration plots by comparing predicted probabilities with actual observed outcomes. Internal validation was performed using a bootstrapping method (500 iterations) to ensure model stability and adjust for potential overfitting. RESULTS: An initial postoperative refraction of <+0.75D, and a higher IOL Power to Axial length Ratio (IOL/AL ratio) were identified as significant risk factors for the development of postoperative high myopia. Shorter preoperative axial length was associated with a greater magnitude of postoperative myopic shift. The predictive model demonstrated robust performance, achieving a C-index of 0.711 (internal validation C-index: 0.713). The area under the receiver operating characteristic curve (AUC) values for predicting high myopia at 5 and 10 years were 0.858 and 0.745, respectively. Furthermore, calibration curves demonstrated excellent agreement between the predicted and observed outcomes throughout the follow-up period. In external validation, the model achieved a C-index of 0.825, 5-year AUC of 0.833, and 10-year AUC of 0.713. CONCLUSIONS: Our analysis established that initial postoperative refraction <+0.75D, and an elevated IOL/AL ratio are key determinants of high myopia risk following surgery. Shorter preoperative axial length was associated with a greater magnitude of postoperative myopic shift. This predictive framework provides clinicians with a practical tool to optimize preoperative IOL selection and identify high-risk infants who require vigilant myopia prevention and balanced amblyopia management.

Humans↗
PubMed2026

A bimodal large language model reduces misalignment in patient education: A double-blinded randomized trial.

BACKGROUND: Effective patient education requires accurate communication aligned with patients' emotional and semantical needs. Text-based large language models (LLMs) lack access to non-verbal cues, which may contribute to misaligned responses. METHODS: We evaluated emotional and semantic misalignment in a text-based LLM using 64,200 utterances from 16,583 patient education cases across six departments and three centers. Dolphin was developed integrating text and audio cues and evaluated through emotion recognition, semantic consistency assessment, branch-level ablations, and a double-blinded randomized trial against a matched text-based LLM comparator (Chinese Clinical Trial Registry: (ChiCTR2500095933). FINDINGS: The text-based LLM showed emotional misalignment in 36.7% of responses and semantic misalignment in 28.3% of cases, with higher misalignment under greater burden. Dolphin outperformed the text-based LLM in emotion recognition accuracy (0.886 vs. 0.713) and semantic consistency (84.9% vs. 82.1%; both adjusted p < 0.001). Ablations supported contribution of audio branches. Dolphin received higher expert ratings than the text-based LLM and human educators (all p < 0.001). In 555 patients, Dolphin was associated with greater patient satisfaction (98.6% vs. 93.8%), suggestion acceptance (76.1% vs. 58.9%; p < 0.001), proactive disclosure (44.6% vs. 26.5%; p < 0.001), and fewer 7-day unplanned recontact (12.9% vs. 22.9%; p = 0.002). No unsafe recommendations or safety events were identified. CONCLUSIONS: Compared with text-based LLM, Dolphin improved emotional-semantic alignment and patient-education outcomes, supporting bimodal alignment as a strategy for reducing misalignment-driven communication failures. FUNDING: National Natural Science Foundation of China, State Key Laboratory Special Fund, and Chinese Academy of Medical Sciences Innovation Fund.

Humans↗