SEARCH · Search PubMed

Results for “regularized model”

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.

837 records · Page 2Linked to original sources

PubMed2026

Effect of urinary catheter securement band on meatal pressure injury in male intensive care patients: A randomized controlled trial.

OBJECTIVES: To investigate the effectiveness of a urinary catheter securement band in preventing meatal pressure injury (meatal-PI) in male ICU patients, and to identify associated risk factors and the timing of injury development. METHODS: A total of 248 adult male ICU patients were randomly allocated to an intervention group (n = 124) or a control group (n = 124) between December 2024 and April 2025. The intervention group received a catheter securement band in addition to standard care, while the control group received standard care alone. Meatal-PI was evaluated daily using a structured monitoring form and a validated staging system. RESULTS: The incidence of meatal-PI was significantly lower in the intervention group (6.5%) compared with the control group (16.1%) (p = 0.016). Multivariate analysis identified catheter securement, use of silicone catheters, higher Braden Scale scores, and shorter ICU length of stay as independent protective factors, while advanced age was associated with increased risk. Additional factors significantly associated with meatal-PI included comorbidities, higher device burden, latex catheter use, dry skin, lower Glasgow Coma Scale and Braden scores, sedation, and perineal oedema (p < 0.001). CONCLUSIONS: The use of a catheter securement band significantly reduces the incidence of meatal-PI in male ICU patients. Incorporating catheter securement devices into routine nursing care, prioritising silicone catheter use, and performing regular meatal assessments may enhance patient safety by reducing the risk of device-related pressure injuries. Further research comparing different catheter securement methods across diverse patient populations is warranted. IMPLICATIONS FOR CLINICAL PRACTICE: The use of catheter securement bands may reduce the incidence of meatal-PI in male ICU patients. Incorporating standardized catheter stabilization strategies into routine intensive care nursing practice may enhance patient safety and support pressure injury prevention efforts.

Humans↗

PubMed2026

Influence of Repeated-Sprint Bout Duration in Sprint Interval Training Intervention on Physical Performance Adaptations of Young Volleyball Players.

The objective of this study was to examine the effects of repeated-sprint training (RST) with varying bout durations on the physical fitness adaptations of young male volleyball players. Forty athletes were randomly allocated to one of three intervention groups performing RST with varying bout durations and similar repetition volumes, all executed at maximal effort. The 3-sec group (n = 10) completed two sets of 30 bouts, the 6-sec group (n = 10) performed two sets of 15 bouts, and the 9-sec group (n = 10) carried out two sets of 10 bouts, each adhering to a 1:3 work to rest ratio. An active control group (n = 10) engaged solely in regular volleyball training without the RST intervention. Physical fitness measures-including countermovement vertical jump (CMVJ), 10-m and 20-m linear sprints, T-test change-of-direction speed (T-CODS), reactive strength index (RSI), and the Wingate anaerobic power test-were assessed pre- and post-a 6-week training intervention (i.e., 18 sessions). All RST groups showed significant post-intervention improvements in physical fitness (main effect of time, p = 0.001), with greater adaptations compared with the control group and effect sizes ranging from small to very large. The 3-sec bout group demonstrated greater gains in CMVJ, 10-m and 20-m sprint performance, RSI, and peak power output compared with the 9-sec group (all, p < 0.05). Conversely, the 9-sec group exhibited superior adaptations in T-CODS and mean power output relative to the 3-sec group (all, p < 0.05). In conclusion, the 3-sec group experienced greater enhancements in explosive and sprint performances, while the 9-sec group showed superior gains in change of direction and mean power output. These findings indicate that manipulation of sprint-bout duration in RST can be used to optimize distinct performance adaptations in young volleyball players.

Humans↗

PubMed2026

Multimodal Therapy With Metformin, Inositol and Dietary Restriction Improves Insulin Resistance and Endocrine Outcomes in Women With Polyendocrine Metabolic Ovarian Syndrome: A Randomized Controlled Trial.

INTRODUCTION: Polyendocrine metabolic ovarian syndrome (PMOS), formerly known as polycystic ovary syndrome (PCOS), is a common endocrine-metabolic disorder characterized by insulin resistance, hyperandrogenism and ovulatory dysfunction. Metformin, inositol supplementation and lifestyle modification are widely used treatments, but direct comparative evidence remains limited. Multimodal therapy combining metformin, inositol and dietary restriction produces greater metabolic and reproductive improvement than single-modality interventions. METHODS: We conducted a 12-week randomized controlled trial in 192 women aged 18-35 years diagnosed with PMOS according to Rotterdam criteria. Participants were allocated to metformin (1500-2000 mg/day), inositol (myo-inositol 2 g plus d-chiro-inositol 50 mg twice daily), calorie-restricted diet (1200-1500 kcal/day), or combination therapy. Primary outcomes included changes in body mass index (BMI) and insulin resistance assessed by HOMA-IR. Secondary outcomes included testosterone, LH/FSH ratio and menstrual regularity. Analysis was performed using analysis of covariance (ANCOVA), with post-intervention values as dependent variables and corresponding baseline values as covariates. Categorical outcomes were compared using the Chi-square test. RESULTS: All interventions improved metabolic and endocrine parameters. Combination therapy resulted in the greatest reduction in HOMA-IR (- 2.64, 95% CI - 2.82 to -2.46, p < 0.001) and BMI (- 2.8 kg/m2, 95% CI - 3.05 to -2.55, p < 0.001). Menstrual cyclicity improved across all groups, with the highest proportion of participants reporting cycle regularisation in the combination therapy group (85.4%), compared with dietary restriction (72.9%), inositol (64.6%), and metformin (39.6%) (p < 0.001). Given the short follow-up duration, these findings reflect early improvements rather than sustained normalisation. CONCLUSION: Multimodal therapy was associated with superior metabolic and reproductive outcomes compared with single-modality interventions in women with PMOS. CLINICAL TRIAL REGISTRATION: ClinicalTrials. gov (NCT07380841).

Humans↗

PubMed2026

A prospective crossover study comparing ICCS-recommended and Palmer-adjusted filling rates in children with spina bifida.

OBJECTIVE: This study aimed to investigate whether the maximal cystometric capacity (MCC) in children with spina bifida (SB) is indeed lower, as predicted by the Palmer formula, and to evaluate the impact of different bladder filling rates on urodynamic parameters. MATERIALS AND METHODS: This prospective, randomized, two-sequence crossover-controlled study included 70 children aged 3-18 years with spina bifida under regular follow-up. In Group 1, the first two bladder fillings were performed at the ICCS-recommended rate, and the third at 75% of that rate (Palmer formula). In Group 2, the sequence was reversed. Urodynamic parameters, including maximal cystometric capacity (MCC), bladder compliance, detrusor activity, filling pressures, and detrusor leak point pressure (DLPP), were analyzed across fillings. RESULTS: Cystometric bladder capacity was lower during fillings performed at the Palmer-adjusted rate compared with those at the ICCS-recommended rate. The proportion of reduced compliance significantly decreased in Group 1 (p = 0.046) but remained unchanged in Group 2. A significant positive correlation was observed between expected bladder capacity (EBC) and measured MCC in both groups (ρ ≈ 0.5-0.6). The highest correlation and agreement were found in Group 1 during the third filling at the Palmer rate (ICC = 0.606). No significant intra- or intergroup differences were observed in detrusor pressure, end-filling pressure, DLPP, or overactive bladder prevalence. CONCLUSION: Bladder filling rate was associated with differences in both cystometric capacity and bladder compliance in children with spina bifida. Fillings performed according to the Palmer formula (approximately 75% of the ICCS-recommended rate) were associated with capacities that more closely approximated age-expected values and with modest differences in bladder compliance. Conversely, faster filling rates did not produce similar benefits. These findings suggest that slower filling strategies may improve measurement consistency and agreement with expected bladder capacity estimates. However, the magnitude and direct clinical impact of these differences should be interpreted cautiously, particularly in light of the potential influence of sequence-related effects.

Humans↗

PubMed2026

A New Highly Concentrated Insulin Aspart AT278 (500 U/mL) Demonstrates Ultra-Rapid Pharmacokinetic and Pharmacodynamic Properties in Type 2 Diabetes Regardless of BMI.

AIMS: To evaluate the pharmacokinetics, pharmacodynamics, and safety of a novel U500 insulin aspart formulation (AT278 [500 U/mL]; AT278-U500) compared with standard concentration insulin aspart (InsAsp [100 U/mL]; InsAsp-U100) and U500 human regular insulin (HumIns [500 IU/mL]; HumIns-U500). MATERIALS AND METHODS: This single-centre, randomised, double-blind crossover 12-h euglycaemic clamp study was conducted in 41 overweight and obese people with type 2 diabetes (BMI 25.0-38.7 kg/m2) receiving a single subcutaneous dose (0.5 U/kg) of AT278-U500 and InsAsp-U100. HumIns-U500 was consecutively studied open label in a 24-h clamp. RESULTS: AT278-U500 exhibited a significantly faster insulin absorption than InsAsp-U100 and HumIns-U500 (t Early50%Cmax: 9 min vs. 35 min vs. 55 min), leading to a significantly higher glucose-lowering effect within the first hour (AUCGIR,0-60min) compared with both InsAsp-U100 (treatment ratio 2.02 [95% CI 1.64; 2.50]) and HumIns-U500 (3.91 [2.89; 5.27]). When divided by median BMI (29.7 kg/m2), AUCGIR,0-60min was significantly higher with AT278-U500 in both the low-BMI and high-BMI subgroup compared to InsAsp-U100. Linear regression showed a significant inverse relationship between BMI and AUCGIR,0-60min for InsAsp-U100 (slope -0.142, p < 0.0001), whereas AT278-U500 showed no such relationship. Overall insulin exposure was similar for AT278-U500 and InsAsp-U100, while overall glucose-lowering effect was comparable across all three treatments. CONCLUSIONS: AT278-U500 maintains its ultra-rapid onset characteristics independent of BMI, representing the first ultra-rapid U500 option for prandial dosing in insulin-resistant people with type 2 diabetes requiring high-dose therapy. TRIAL REGISTRATION: ClinicalTrials.gov identifier: NCT05754424.

Humans↗

PubMed2026

Care Navigation for Methamphetamine Use Disorder: A Randomized Clinical Trial.

IMPORTANCE: Stimulant-involved deaths continue to increase in the US, and methamphetamine use remains a weighty public health concern. Treating methamphetamine use disorders is complicated. Contingency management has demonstrated the best effectiveness but is not widely implemented. OBJECTIVE: To examine the effectiveness of dedicated care navigation in linking patients to treatment. DESIGN, SETTING, AND PARTICIPANTS: This prospective randomized clinical trial was conducted at an integrated safety-net health system in Denver, Colorado, between April 10, 2023, and December 31, 2024. Eligible participants were 18 years or older who had a methamphetamine-related encounter in an acute care setting; those with involuntary treatment holds, substance treatment in past 90 days or actively seeking treatment, and inability to provide consent were excluded. Participants completed baseline, 30-day, and 90-day study visits. INTERVENTION: Dedicated care navigation, incorporating contingency management principles, with a focus on addressing health-related social needs. MAIN OUTCOMES AND MEASURES: Linkage to treatment within 30 and 90 days of enrollment defined as a composite measure of at least 1 of the following: electronic health record data indicating a visit at the health system's substance treatment clinic, a behavioral health encounter at an outpatient clinic, temporary residential treatment, or self-reported treatment on the 30- and/or 90-day follow-up survey. RESULTS: Of 192 participants enrolled in the Beginning Early and Assertive Treatment for Methamphetamine Use trial, 156 (81.3%) were male, and the median age was 39 (IQR, 31-47) years. Most participants were unstably housed (163 [84.9%]), not currently employed (158 [82.3%]), and without regular access to a working phone (94 [49.0%]). Of the 96 participants randomized to the intervention, 60 (62.5%) engaged in 2 or more navigation sessions, 45 (46.9%) completed the 30-day study visit, and 47 (49.0%) completed the 90-day study visit compared with 44 (46.3%) and 37 (38.5%), respectively, of the 96 randomized to the control arm. No statistically significant differences in treatment linkage were observed at 30 days (24 participants [25.0%] in both arms; risk ratio, 1.00 [95% CI, 0.61-1.63]) or 90 days post enrollment, (32 [33.3%] in intervention vs 24 [25.0%] in control arms; risk ratio, 1.33 [95% CI, 0.85-2.09]). CONCLUSIONS AND RELEVANCE: In this randomized clinical trial, integrating principles of contingency management into the intervention may have increased engagement with a dedicated care navigator but did not increase likelihood of linkage to treatment for methamphetamine use disorder. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT06033365.

Humans↗

PubMed2026

Durvalumab and tremelimumab, with or without lenvatinib, combined with transarterial chemoembolisation in participants with embolisation-eligible hepatocellular carcinoma (EMERALD-3): a global, randomised, open-label, sponsor-blinded, phase 3 study.

BACKGROUND: Transarterial chemoembolisation (TACE), a standard treatment for embolisation-eligible hepatocellular carcinoma (HCC), induces tumour immune responses. Single tremelimumab regular interval durvalumab (STRIDE) is a standard treatment in advanced HCC. In this phase 3 trial, we assessed the efficacy and safety of STRIDE, with or without lenvatinib, plus TACE, in participants with embolisation-eligible HCC. METHODS: EMERALD-3 is a phase 3, randomised, open-label, sponsor-blinded study, conducted at 177 medical sites in 21 countries. Eligible participants were 18 years or older (aged ≥21 years in Egypt or Singapore) at screening and had confirmed HCC (by imaging or histopathologically from biopsy specimen, surgery, or both) not amenable to curative surgery, curative ablation, or transplantation but amenable to TACE. Participants had Child-Pugh class A liver function, an Eastern Cooperative Oncology Group performance status of 0-1, and at least one measurable target intrahepatic lesion per modified Response Evaluation Criteria in Solid Tumours. Participants were randomly allocated in a 1:1:1 ratio to receive STRIDE plus lenvatinib plus TACE, STRIDE plus TACE, or TACE until each group reached its preplanned enrolment target of 175 participants. After the STRIDE plus TACE group reached its enrolment target, randomisation was adjusted to continue in a 1:1 ratio between the STRIDE plus lenvatinib plus TACE group and TACE group until approximately 275 participants were enrolled in each of these two groups. Randomisation used a centrally assigned interactive response technology system, stratified by region, baseline tumour burden, and previous palliative embolisation. In the STRIDE plus lenvatinib plus TACE group, on the first day, participants were given 300 mg tremelimumab intravenously, followed by 1500 mg durvalumab plus oral lenvatinib (8 mg for <60 kg bodyweight or 12 mg for ≥60 kg bodyweight); participants then received 1500 mg durvalumab every 4 weeks plus once-daily lenvatinib for up to 36 cycles. In the STRIDE plus TACE group, participants were given 300 mg tremelimumab and 1500 mg durvalumab intravenously on the first day, followed by 1500 mg durvalumab every 4 weeks. The technique and number of TACE procedures were at the investigators' discretion, with the first procedure administered at least 7 days after the first dose of durvalumab in the two investigation treatment groups and within 7 days of random allocation in the TACE group. The primary endpoint was progression-free survival for STRIDE plus lenvatinib plus TACE versus TACE. Key secondary endpoints were overall survival for STRIDE plus lenvatinib plus TACE versus TACE and progression-free survival and overall survival for STRIDE plus TACE versus TACE. This study was registered with ClinicalTrials.gov (NCT05301842), with enrolment completed. FINDINGS: From March 28, 2022, to Nov 20, 2024, 1124 participants were screened. The full analysis set comprised 760 participants, who were randomly allocated to STRIDE plus lenvatinib plus TACE (n=293), STRIDE plus TACE (n=175), or TACE (n=292). 633 (83%) participants were male and 127 (17%) were female; 548 (72%) were Asian. At the first data cutoff (Sept 2, 2025); the overall median follow-up for progression-free survival was 10·0 months (IQR 4·6-17·2); median follow-up for progression-free survival was 11·0 months (IQR 4·8-18·4) for STRIDE plus lenvatinib plus TACE and 8·3 months (4·1-15·5) for TACE. Median progression-free survival was 13·0 months (95% CI 12·2-16·7) for STRIDE plus lenvatinib plus TACE versus 9·8 months (8·0-11·4) for TACE (HR 0·70 [95% CI 0·57-0·86]; p=0·0007). At the second data cutoff (Feb 23, 2026) and a median follow-up for overall survival of 24·6 months (IQR 16·5-31·5) for STRIDE plus lenvatinib plus TACE and 22·9 months (14·9-30·2) for TACE, median overall survival was 39·5 months (95% CI 34·1-not reached) for STRIDE plus lenvatinib plus TACE and 34·7 months (28·8-not reached) for TACE (HR 0·84 [95% CI 0·65-1·09]; p=0·18). At this data cutoff, median progression-free survival was 12·9 months (95% CI 10·2-15·9) for STRIDE plus TACE and 8·1 months (6·5-10·2) for the first 175 participants randomised to TACE (HR 0·71 [95% CI 0·56-0·91]), with median follow-up of 10·3 months (IQR 4·6-23·7) for STRIDE plus TACE and 7·7 months (3·0-18·5) for the first 175 participants randomly allocated to TACE. The most common adverse events of maximum grade 3 or 4 were hypertension (34 [12%] of 287) for STRIDE plus lenvatinib plus TACE, post-embolisation syndrome and anaemia (ten [6%] of 175 each) for STRIDE plus TACE, and post-embolisation (17 [6%] of 290) for TACE. 184 (64%) participants receiving STRIDE plus lenvatinib plus TACE, 89 (51%) receiving STRIDE plus TACE, and 68 (23%) receiving TACE had serious adverse events. Treatment-related adverse events with an outcome of death during the treatment-emergent period occurred in seven (2%) of 287 participants who received STRIDE plus lenvatinib plus TACE (two for myocarditis; and one each for hepatic failure, haemophagocytic lymphohistiocytosis, septic shock, cardiac failure, and unknown cause), none of 175 participants who received STRIDE plus TACE, and two (1%) of 290 participants who received TACE (one each for acute myocardial infarction and unknown cause). INTERPRETATION: STRIDE plus lenvatinib plus TACE showed a statistically significant progression-free survival improvement versus TACE. These findings support a STRIDE-based regimen as a potential new treatment option for people with embolisation-eligible HCC; additional follow-up is being conducted for final analysis of overall survival across treatment groups. FUNDING: AstraZeneca.

Adult↗

PubMed2026

Statistical test to compare the linkage model and the admixture model based on central limit results.

In the Admixture Model, the probability that an individual carries a certain allele at a specific marker depends on the allele frequencies in K ancestral populations and the proportion of the individual's genome originating from these populations. The markers are assumed to be independent. The Linkage Model is a Hidden Markov Model that extends the Admixture Model by incorporating linkage between neighboring loci. We prove consistency and asymptotic normality of maximum likelihood estimators for the ancestry of individuals in the Linkage Model, complementing earlier results by (Pfaff et al., 2004; Pfaffelhuber and Rohde, 2022; Heinzel, 2025) for the Admixture Model. These results are used to prove that a statistical test that allows for model selection between the Admixture Model and the Linkage Model is an asymptotic level-α-test. Finally, we demonstrate the practical relevance of our results by applying the test to real-world data from The 1000 Genomes Project Consortium (2015).

Genetic Linkage↗

PubMed2026

Modelling the effects of biological intervention in a dynamical gene network.

Cellular response to environmental and internal signals can be modeled by dynamical gene regulatory networks (GRN). In the literature, three main classes of gene network models can be distinguished: (1) non-quantitative (or data-based) models which do not describe the probability distribution of gene expressions; (2) quantitative models which fully describe the probability distribution of all genes co-expression; and (3) mechanistic models which allow for a causal interpretation of gene interactions. We propose two rigorous frameworks to model gene alteration in a dynamical GRN, depending on whether the network model is quantitative or mechanistic. We explain how these models can be used for design of experiment, or, if additional alteration data are available, for validation purposes or to improve the parameter estimation of the original model. We apply these methods to the Gaussian graphical model, which is quantitative but non-mechanistic, and to mechanistic models of Bayesian networks and penalized linear regression.

Gene Regulatory Networks↗

PubMed2026

The landscape of pruning for large language models: A systematic review and unified taxonomy.

Confronting the inherent tension between the exceptional capabilities and the immense computational costs of Large Language Models (LLMs), pruning has become a crucial technique for achieving efficient deployment. However, a systematic analytical framework dedicated specifically to LLM pruning remains absent. In this paper, we aim to bridge this gap. We first elucidate the theoretical foundations that underpin the effectiveness of pruning, namely overparameterization and redundancy, and then propose a multidimensional taxonomy that organizes existing approaches along the axes of granularity, timing, and criteria. Building upon this unified perspective, we further analyze performance recovery mechanisms and the broader evaluation ecosystem, while also exploring forward-looking challenges such as interpretability, automation, and hardware-algorithm co-design. Through this comprehensive synthesis, we seek to provide an integrated and coherent analytical lens for advancing both research and practice in LLM pruning.

Large Language Models↗

PubMed2025

An integrated multiscale air quality modelling framework for industrial park pollution: Linking local emissions to regional transport.

Capturing the spatiotemporal distribution of pollutants in industrial parks remains challenging for regional air quality models because of their coarse resolution (3 km), resulting in uncertainties in local emission quantification. To address this, we developed the Integrated Multiscale Air Quality Modelling System for Industry (IAQMS-Industry), coupling the regional Nested Air Quality Prediction Modelling System (NAQPMS) with a city-scale chemical transport model. This framework integrates point-source locations and Gaussian plume dispersion to simulate particulate matter with a diameter smaller than 2.5 micrometres (PM2.5) at 100 m resolution. Applied to the Beijing Yi Zhuang and Tangshan industrial parks and evaluated against observations. The coupled model achieved a normalized mean bias (NMB) ranging from 3.1 % to 6.2 %, improving upon NAQPMS (-16.9 % to -7.7 %). Spatial analysis revealed that coarse regional grids underestimated the PM2.5 concentrations at industrial sites by smoothing gradients, whereas IAQMS-Industry successfully resolved spatial patterns. Industrial point emissions accounted for 22.9 %-26.4 % of PM2.5 in the coupled model, which was significantly greater than the regional model estimates of 1.6 %-13.7 %. These findings indicate that regional models overestimate pollutant dispersion processes in industrial parks while underestimating local industrial impacts. By explicitly resolving point-source dynamics and linking them to regional transport, IAQMS-Industry provides a robust tool for designing targeted emission controls in industrial cities and balancing local air quality improvements with minimized regional pollution outflow. This study underscores the necessity of multiscale modelling for accurate source apportionment and informed environmental governance in industrial zones.

Air Pollution↗

PubMed2026

Snow carps as an emerging model for polyploid genome evolution.

Explore the source record for details and available documents.

Editorial↗

PubMed2026

Construction of precision clinical-proteomics risk model based on machine learning for predicting heart failure in type II diabetes mellitus.

BACKGROUND AND AIMS: Heart failure (HF) is a severe complication in type 2 diabetes mellitus (T2DM), but current risk stratification scores have limited predictive accuracy. We aimed to develop novel prediction tools integrating clinical variables with proteomics to improve risk stratification of hospitalization for HF in T2DM. METHODS AND RESULTS: In this study, we included 2111 UK Biobank participants with T2DM but no prior HF, and profiled 2920 proteins to predict 10-year incident HF hospitalization. Participants were randomly divided into training (70%), tuning (10%), and validation (20%) sets.Three prediction models were developed: a Clinical model based on demographic characteristics, comorbidities, medication use, and laboratory indices; a Protein model based on 40 proteins selected by the Light Gradient Boosting Machine (LGBM); and the Clinical OMics and Protein ASSessment for Heart Failure (COMPASS-HF) model, which integrated both clinical variables and the LGBM-selected proteins. Models were evaluated for area under the curve (AUC), sensitivity, and specificity. During follow-up, 168 participants (7.96%) developed incident HF. The COMPASS-HF model showed better discrimination than the Clinical model, with an AUC of 0.897 (95% CI: 0.850-0.945) versus 0.790 (95% CI: 0.723-0.856). It also demonstrated higher sensitivity (0.882; 95% CI: 0.725-0.967) and consistent performance in subgroups. COMPASS-HF effectively stratified risk of hospitalization for HF, with cumulative incidence rates of 31.9% in the high-risk group and 1.2% in the low-risk group. CONCLUSIONS: By combining clinical and proteomic variables, we developed a high-performance HF prediction model for T2DM, enabling precise risk stratification and informing early intervention strategies.

Humans↗

PubMed2026

Beyond predictive performance: A systematic review and critical methodological appraisal of AI/ML and conventional modelling strategies in breast, colorectal, and pancreatic Cancer.

BACKGROUND: Predictive modelling for cancer risk, treatment-related complications, and survival is central to precision oncology. Conventional logistic regression (LR) and Cox proportional hazards (CoxPH) regression remain widely used but are limited when modelling nonlinear interactions, high-dimensional imaging features, and multimodal clinical-metabolic predictors. Artificial intelligence (AI) and machine learning (ML) methods offer expanded capability through automated feature extraction, ensemble learning, and flexible survival modelling, but the evidence on when AI/ML adds value over conventional models across cancer sites and predictive tasks remains fragmented. OBJECTIVE: To systematically evaluate the methodological performance, validation strategies, and translational limitations of AI/ML models compared with conventional statistical models in published predictive-modelling studies for breast, colorectal, or pancreatic cancer. METHODS: PubMed, Scopus, and Web of Science were searched for studies published between January 2019 and March 2025. Two reviewers independently conducted title-and-abstract screening, full-text eligibility assessment, and PROBAST risk-of-bias assessment. Sixty-five studies (n = 907,567 participants) were narratively synthesised by cancer site, predictive task, model family, comparator, validation strategy, predictor modality, and calibration or explainability reporting. RESULTS: The 65 studies comprised breast cancer (n = 35), colorectal cancer (n = 21), and pancreatic cancer (n = 9). AI/ML superiority over LR and CoxPH was task- and data-dependent. CNN- and U-Net-based models predominated in imaging and body-composition tasks, tree-based ensembles consistently outperformed LR for tabular perioperative complication prediction, and CoxPH remained competitive, and in the largest pancreatic risk study, superior to XGBoost (C-index 0.802 vs 0.723) in well-structured datasets. PROBAST analysis-domain risk was moderate in 54 of 65 studies (83%), driven by limited external validation, sparse calibration reporting (11/65), and few decision-curve analyses (7/65). CONCLUSION: AI/ML adds the most methodological value in imaging-derived feature extraction and nonlinear perioperative prediction, while conventional regression remains preferable in large, structured datasets with linear predictors. Clinical translation requires standardised body-composition definitions, external validation, calibration assessment, decision-curve analysis, and explainability, in line with TRIPOD+AI and CLAIM standards.

Humans↗

PubMed2026

Corrigendum to "The impact of diverse ancestry on polygenic risk score-based prediction models for psychiatric and neurodevelopmental disorders" [Psychiatry Research 2026 Oct:364:117322].

Explore the source record for details and available documents.

Published Erratum↗

PubMed2026

Integrative modeling of the genome structure and dynamics in fission yeast.

Genome organization in the nucleus is highly structured and dynamic. Recent advances in genomic technology have enabled the measurement of genome-wide architecture and locus-specific motion, yielding contact maps and live-cell trajectories. However, these outcomes are derived from different modalities and are not directly comparable, with their quantitative integration being a key challenge. Here we establish a genome-wide live-cell imaging platform in fission yeast Schizosaccharomyces pombe, tracking 131 chromosomal loci, along with the spindle pole body (SPB) and nucleolus, to construct a quantitative map of locus dynamics. By integrating these dynamics with contact data through polymer modeling of Hi-C data, we build a physics-based "digital twin" of the S. pombe genome consistent with the spatiotemporal dynamics of interphase chromatin. We validate it against genome-wide mobility patterns and known architectural features, including centromere and telomere clustering. The model also identifies distinct dynamical regimes: centromere- and telomere-proximal loci relax within [Formula: see text]150 s, whereas the remaining loci relax within [Formula: see text]70 s. We measure semiperiodic dynamics of SPB motion, including a characteristic peak near 225 s and [Formula: see text] fluctuations. We use the model with SPB-directed forcing to show how these low-frequency components propagate through the genome to drive genome-wide chromatin displacements. Together, this predictive physics-based modeling framework integrates genome structure and dynamics to reveal how nuclear mechanical driving forces shape chromosome motion, linking mechanically driven chromatin responses to genome maintenance and regulation.

Schizosaccharomyces↗

PubMed2026

Revealing the Shared Genetic Architecture of Metabolic Dysfunction-Associated Steatotic Liver Disease-Related Traits Through Genomic Structural Equation Modeling.

Although individual traits related to metabolic dysfunction-associated steatotic liver disease (MASLD) have been investigated through large-scale genome-wide association studies (GWASs), the shared genetic susceptibility across these traits remains unclear. We therefore conducted a multivariate GWAS of key MASLD-related traits to elucidate their common genetic architecture. We applied genomic structural equation modeling to model a latent genetic factor (MASLD-F) underlying genetically correlated MASLD-related traits, leveraging their GWAS-derived genetic correlations. We then performed functional annotations, including fine-mapping, transcriptome-wide association study, and cell- and tissue-type-specific enrichment analyses, and conducted Mendelian randomization analyses to identify modifiable risk factors. Our multivariate MASLD-F GWAS identified 50 independent variants across 48 genomic loci. Transcriptomic imputation identified several MASLD-F-associated genes, including ARNTL, NPC1, BTBD10, VDAC2, TSKU, SFMBT1, and ABHD17C. We observed significant enrichment of MASLD-F-related genetic signals predominantly in brain tissues, pancreatic islets, and the adrenal gland. Additionally, six modifiable risk factors and four modifiable protective factors for MASLD-F were identified. These findings reveal a complex shared genetic architecture underlying MASLD components, thereby expanding our understanding of disease pathogenesis and providing novel insights for precision medicine and public health interventions.

Humans↗

PubMed2026

Evaluation of a cornea-specialized large language model for diagnostic and management accuracy in complex corneal cases.

PURPOSE: To evaluate whether a cornea-specialized large language model (LLM) enhanced with retrieval-augmented generation (RAG) improves clinicians' diagnostic and management accuracy in complex corneal cases compared to a general-purpose GPT-4o model and unaided clinician performance. METHODS: This prospective, randomized, masked evaluation study involved three cornea trainees who each independently reviewed 39 real-world corneal cases under three experimental conditions: unaided, GPT-4o-assisted, and assisted by a cornea-specialized GPT-4o model. The cornea-specialized model was constructed by embedding over 200 publicly available Wikipedia articles into GPT-4o's RAG framework. Participants provided open-ended diagnoses and selected the next-step management options (multiple choice). They were allowed up to three GPT-4o queries per case, and the AI-assisted arms were randomized to minimize bias. Accuracy for both tasks was compared against expert reference standards using McNemar's test. RESULTS: Diagnostic accuracy was 48.7%, 20.5%, and 38.5% unaided, improving to 69.2%, 46.2%, and 59.0% with general GPT-4o (p<0.04). The cornea-specialized GPT-4o further improved accuracy to 71.8%, 48.7%, and 74.4%, with improvements over unaided performance for all clinicians (p<0.01). For next-step decisions, unaided accuracy was 76.9%, 87.2%, and 59.0%. With the specialized model, Ophthalmologist 3 improved to 71.8% (p<0.05), Ophthalmologist 1 remained high at 82.1%, and Ophthalmologist 2 declined to 64.1% (p<0.05). CONCLUSIONS: A cornea-specialized LLM enhanced with RAG improved diagnostic accuracy in complex corneal cases, particularly among clinicians with lower baseline performance. Effects on management accuracy were inconsistent. Future studies should explore the use of open-ended management tasks and examine whether smaller, curated retrieval corpora yield better model performance.

Humans↗