SEARCH · Search PubMed

Results for “screening”

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.

91 recordsLinked to original sources

PubMed2026

Artificial intelligence-supported double reading in European population breast cancer screening: A systematic review and meta-analysis of prospective programs.

BACKGROUND: Most European population mammography screening programs rely on double reading with arbitration, a model that delivers mortality benefit but is increasingly challenged by radiologist workload, variable specificity, and interval cancers. Artificial intelligence (AI) is being evaluated to support or optimize these established European screening pathways. PURPOSE: To synthesize prospective or program-embedded evaluations of AI conducted within European-style population screening programs and to estimate exploratory program-level absolute risk differences (RDs) per 1000 examinations for cancer detection rate (CDR) and recall. MATERIALS AND METHODS: We performed a prespecified, focused evidence synthesis of three large studies embedded within routine population screening programs operating under European-relevant workflows: MASAI (randomized AI-supported risk triage within a national program), ScreenTrustCAD (prospective paired-reader evaluation with AI as an independent reader in a double-reading framework), and PRAIM (nationwide decision-referral implementation). Outcomes were harmonized as AI-control RDs per 1000 examinations. Random-effects pooling used Hartung-Knapp-Sidik-Jonkman models. For the paired-reader design, sensitivity analyses applied a Kish effective sample-size approach across plausible within-examination correlations (ρ = 0.3-0.8). Positive predictive value (PPV) and workflow/time outcomes were summarized descriptively. RESULTS: Across 597,419 examinations, the pooled CDR RD was +0.9 per 1000 (95% CI -0.0 to +1.8; I2 ≈ 12%), consistent with a modest directional increase with borderline statistical uncertainty. The pooled recall RD was -0.6 per 1000 (95% CI -3.1 to +2.1; I2 ≈ 41-43%), indicating no consistent recall increase across screening programs. Where reported, PPV was higher with AI-supported screening. Efficiency signals included 44.3% fewer total readings in MASAI and shorter reading times for AI-normal examinations in PRAIM; in PRAIM, a program-level safety-net mechanism recovered 204 cancers that would otherwise have been missed. CONCLUSION: In European population screening programs characterized by double reading and arbitration, prospective program-embedded evidence suggests that AI integration may yield a small absolute increase in cancer detection (≈1/1000) without a consistent increase in recall, alongside improved PPV and efficiency signals. These findings suggestAI primarily as a complementary reader within European screening workflows, with implementation requiring explicit quality assurance and monitoring of interval cancers and stage distribution.

Humans↗

PubMed2026

HPV Testing Versus Cytology for Cervical Cancer Screening Among Women 50 Years and Older: Evidence From the HPV FOCAL Randomized Controlled Trial.

Evidence on the comparative effectiveness of HPV testing versus cytology specifically in women aged ≥ 50 years who are approaching screening cessation remains limited. This analysis included 6471 women aged ≥ 50 at baseline screening in the HPV FOCAL randomized clinical trial. Women were randomly allocated to receive cytology (Control Group, n = 3248, 50.19%) or HPV testing (Intervention Group, n = 3223, 49.81%) at baseline, with co-testing at 48-month exit. We calculated incidence rates and risk ratios for CIN2+ detection over follow-up and compared missed lesions at exit by screening method. At the 48-month exit, CIN2+ detection was lower among HPV baseline-negative women than among those in the cytology group (1.61/1000 [95% CI, 0.52-3.76] vs. 3.15/1000 [95% CI, 1.51-5.78]; risk ratio, 0.51 [95% CI, 0.11-0.91]), reflecting higher baseline detection with HPV testing and fewer prevalent lesions at exit. Even with cytology re-screening at 2 years, 50% of CIN2+ cases were missed compared to 30% with HPV testing. After adjusting for age, education, smoking status, and lifetime sexual partners, the hazard ratio for CIN2+ comparing HPV to cytology was 0.44 (95% CI, 0.22-0.88). Among women aged ≥ 50 years, HPV primary screening was more effective than cytology at detecting CIN2+ lesions and was associated with a continued lower subsequent risk following a negative HPV test, supporting its use in cervical cancer screening programs in this age cohort.

Aged↗

PubMed2026

AI Health message intervention: The role of message customization and message source in breast cancer screening among women of color.

OBJECTIVES: To examine the effectiveness of breast cancer screening messages with varying levels of customization (generic, targeted, and tailored) and to compare AI-generated versus human-generated messages. METHODS: A between-subjects experimental design with a control condition was employed. Message content followed a standardized structure and varied by level of customization: generic, targeted (demographic-based), and tailored (perceived susceptibility- and barrier-based). Messages were developed by either the authors or GenAI (ChatGPT-4o). A total of 391 participants recruited via Prolific were randomly assigned to five groups (generic, targeted-human, targeted-AI, tailored-human, and tailored-AI). Self-efficacy, behavioral intentions, attitudes, and message believability were measured using different scales. RESULTS: Customized (tailoring and targeting) health messages performed comparably to generic messages in shaping positive health outcomes. GenAI-generated messages also produced outcomes comparable to those of human-generated messages under standardized conditions. Significant negative indirect effects through message believability for the human-tailored condition was found relative to the generic condition. CONCLUSIONS: GenAI may be a useful tool for developing and customizing scalable health messages. Its effectiveness depends not only on customization but also on maintaining message quality, including readability, clarity, coherence, naturalness, and credibility. PRACTICAL IMPLICATIONS: GenAI may support health practitioners in developing customized and scalable breast cancer messages. However, professional review remains necessary to ensure that the message is culturally appropriate, responsive to patient concerns, and suitable for use alongside patient-provider communication.

Humans↗

PubMed2026

Effects of Digital Mental Health Screening Alone and With the Online MINDBODYSTRONG CBT-Based Program on Burnout, Depression, Anxiety, Healthy Behaviors, and Suicidal Ideation in at-Risk Nurses at 3- and 6-Months Post-Intervention: An RCT.

BACKGROUND: Burnout and mental distress among nurses are global public health epidemics that adversely affect nurse well-being and healthcare quality. Evidence-based, scalable mental health interventions are urgently needed. AIMS: To evaluate the 3- and 6-month outcomes of a randomized controlled trial (RCT) comparing a psychologically safe, digital mental health screening and referral program alone versus the same screening and referral program combined with the video-based online MINDBODYSTRONG (MBS) cognitive behavioral therapy (CBT)-based skills-building program among nurses at risk for mental distress. METHODS: 501 nurses were recruited from professional organizations and healthcare systems across the United States by email and randomized to either mental health screening and referral (standard care) or standard care plus the MBS cognitive behavioral skills-building intervention (the intervention). All study activities were conducted remotely. Follow-up surveys administered at 3- and 6-months assessed anxiety, depression, suicidal ideation, burnout, healthy lifestyle beliefs, and healthy lifestyle behaviors using valid and reliable scales. RESULTS: Compared with the screening and referral only group, participants in the intervention group had greater reductions in anxiety and depression and significantly greater increases in healthy lifestyle beliefs and behaviors at 3 and 6 months post-intervention. After controlling baseline risk, the intervention group had a lower risk of suicidal ideation than the screening and referral group at 3 months (relative risk ratio [RRR] = 0.717; 95% CI: 0.320-1.606) and 6 months (RRR = 0.329; 95% CI: 0.101-1.072). The intervention group also had a significantly lower risk of burnout at 6 months (RRR: 0.698, 95% CI: 0.528, 0.929, p = 0.012). Nurses who completed more MBS sessions had less suicidal ideation at 6 months and those who completed more MBS skills-building activities had less burnout at 3 and 6 months. LINKING ACTION TO EVIDENCE: Integrating psychologically safe mental health screening combined with the scalable online CBT-based intervention, MBS, can produce sustained improvements in burnout, mental health symptoms, including suicidality, and healthy lifestyle beliefs and behaviors among nurses experiencing mental distress.

Humans↗

PubMed2026

Compound muscle action potential amplitudes in newborn screen positive spinal muscular atrophy.

OBJECTIVE: To evaluate the utility of compound muscle action potential (CMAP) amplitudes as biomarkers of disease severity in newborn screening (NBS)-positive infants with spinal muscular atrophy (SMA). METHODS: We conducted a retrospective review of 21 infants identified through SMA NBS (11 with 2 SMN2 copies and 10 with 3 SMN2 copies). Baseline and serial right median, ulnar, and fibular motor nerve CMAP amplitudes (millivolts, mV) were obtained during the study. Functional outcomes were assessed using the Children's Hospital of Philadelphia Infant Test of Neuromuscular Disorders (CHOP-INTEND). RESULTS: At baseline, infants with 2 SMN2 copies demonstrated significantly lower median, ulnar, and fibular CMAP amplitudes compared with infants with 3 SMN2 copies (p < 0.05). In contrast, baseline CHOP-INTEND scores did not differ significantly between the two groups. Prior to genetic confirmation, a right median CMAP amplitude ≥3.2 mV predicted ≥3 SMN2 copies. Following treatment, right median and fibular CMAP amplitudes demonstrated significant improvement over time, including in analyses accounting for SMN2 copies number. CONCLUSION: CMAP amplitudes obtained from multiple upper- and lower-extremity motor nerves provided objective electrophysiological measures that distinguished infants with two versus three SMN2 copies, despite similar baseline CHOP-INTEND scores. Furthermore, CMAP abnormalities were detectable in some cases before confirmatory genetic testing results became available. Serial CMAP measurements demonstrated significant longitudinal changes following treatment, whereas functional assessments approached ceiling values, supporting the potential value of electrophysiological monitoring in the era of disease-modifying therapies. SIGNIFICANCE: CMAP assessment is a useful adjunct in the evaluation of infants identified through SMA NBS.

Humans↗

PubMed2026

Bioactive peptides for meat quality and preservation: Integrating peptidomics and computational screening.

Bioactive peptides generated from meat proteins, fermented meat products, and slaughter by-products have attracted increasing attention as functional molecules for improving meat quality and preservation. In meat systems, peptides can be produced through endogenous postmortem proteolysis, microbial fermentation, gastrointestinal digestion, or controlled enzymatic hydrolysis of underutilized animal by-products. These peptides are closely associated with key meat science endpoints, including postmortem tenderization, oxidative stability, color retention, flavor development, microbial inhibition, and the valorization of processing by-products. However, although high-resolution peptidomics has greatly expanded the identification of meat-derived peptide sequences, their translation into practical meat applications remains limited by matrix interactions, processing stability, sensory constraints, safety concerns, and insufficient validation in real meat systems. This review synthesizes recent advances in meat-related peptidomics and computational screening, including sequence-based prediction, machine learning, molecular docking, molecular dynamics, stability assessment, and safety-oriented filtering. Particular attention is given to how these approaches can prioritize peptides with antioxidant, antimicrobial, flavor-modulating, and preservation-related functions under meat-specific technological constraints. By integrating peptide generation pathways, mass spectrometry-based identification, in silico prioritization, and meat quality endpoints, this review proposes a stage-gated framework for translating meat-derived bioactive peptides from discovery to application. Future research should strengthen matrix-specific validation, standardized peptidomic reporting, and safety assessment to support the use of bioactive peptides in meat quality improvement, clean-label preservation, and circular utilization of meat industry by-products.

Animals↗

PubMed2026

Evaluation of three Aspergillus antibody assays for screening of chronic pulmonary aspergillosis: prospective diagnostic accuracy study.

OBJECTIVES: Chronic pulmonary aspergillosis (CPA) is a frequent complication of pulmonary tuberculosis (PTB), particularly in high-burden settings where access to reliable serological diagnostics remains limited. We evaluated the diagnostic performance of two immunochromatographic technology (ICT) lateral flow assays (LFAs) and an ELISA for CPA screening among patients with active or previously treated PTB. METHODS: In this two-year prospective multicentre diagnostic evaluation, serum from adults with prior or active PTB was tested using the Era Biology Aspergillus IgG ICT LFA, LDBio Aspergillus IgG/IgM ICT LFA, and Bordier Aspergillus fumigatus IgG ELISA. CPA diagnosis was established using a consensus composite reference standard incorporating clinical, immunological, radiological, and microbiological criteria. The Bordier ELISA was used as part of the immunological component of the consensus CPA diagnosis, with a cutoff optical density of ≥1.0. Diagnostic accuracy, agreement statistics, receiver operating characteristic analysis, and latent class analysis (LCA) were performed. RESULTS: Among 340 participants, 24 (7.06%) had CPA. Proportion of participants with positive antibody tests among all tested individuals were 6.76% for LDBio ICT LFA, 20.0% for Era Biology ICT LFA, and 11.47% for Bordier ELISA. Against consensus CPA diagnosis, Bordier ELISA showed 87.50% sensitivity and 94.30% specificity, LDBio ICT LFA 58.33% sensitivity and 97.15% specificity, and Era Biology LFA 66.67% sensitivity and 83.54% specificity. LCA estimated CPA prevalence at 7.72%. LCA-derived sensitivities and specificities were 86.58% and 99.92% for LDBio ICT LFA, 83.39% and 85.31% for Era Biology LFA, and 79.10% and 94.19% for Bordier ELISA. CONCLUSIONS: The Bordier ELISA showed high sensitivity and specificity, while the LDBio ICT LFA demonstrated very high specificity with strong LCA-derived performance. These findings support the use of ELISA for laboratory diagnosis and ICT as a point-of-care screening tool for CPA in resource-limited settings. Era Biology Aspergillus IgG LFA demonstrated moderate sensitivity and acceptable diagnostic performance, indicating its potential utility as a supplementary screening assay for CPA in settings where rapid, point-of-care testing is required.

Humans↗

PubMed2026

The identification of growth-promoting lncRNAs in oral cavity squamous cell carcinoma.

Oral Cavity Squamous Cell Carcinoma (OCSCC) is an aggressive tumor that develops within the mouth of patients. Tumor-suppressor gene loss and genomic arrangements fuel tumorigenesis and transcriptional reprogramming. Understanding how these alterations contribute to OCSCC growth and cell survival may identify new therapeutic vulnerabilities or biomarkers. We profiled the role of long non-coding RNAs (lncRNAs) in the growth of three OCSCC cell lines using a CRISPRi-screen and identified 19 lncRNAs that contribute to OCSCC proliferation. By comparing these lncRNAs to other screens, we find that these lncRNAs are uniquely required in OCSCC and not other malignancies. We show that these lncRNAs are abundantly expressed in OCSCC cells and tumors. Independent testing of candidate lncRNAs confirms their role in supporting OCSCC growth. Our results show that a novel subset of lncRNAs are required for the growth of OCSCC cancer cells and that these lncRNAs are cell lineage specific.

CRISPRi↗

PubMed2026

Innovations in microbial physical mutagenesis for food fermentation: An overview from traditional to emerging technologies.

Microbial strains serve as an important factor affecting fermentation efficiency and product quality. To obtain superior strains, mutation breeding is a classic strategy. Compared to chemical mutagenesis, physical mutagenesis directly induces genomic changes, providing notable advantages such as the elimination of chemical residues and environmental sustainability, hence rendering it a favored method for enhancing food-grade microorganisms. Conventional physical mutagenesis mostly depends on UV, rays, high pressure, or space radiation. As physical technologies advance, emerging methods such as ion implantation, plasma, microwave, ultrasound, and pulsed light are widely utilized for genetic modification. Mutagenesis technologies are progressively transitioning from single-effect to multi-effect synergy. Recent evaluations indicate that emerging technologies can enhance microbial mutation efficiency at the application level relative to established technologies. Nonetheless, the systematic clarification and comparative analysis at the mechanistic level remain inadequate, hindering intuitive comprehension of the qualities and distinctions across techniques. Furthermore, physical mutagenesis encounters several significant obstacles, such as cellular damage, limited rates of advantageous mutations, and laborious screening processes. This review carefully elucidates the mechanisms and properties of physical mutagenesis technology and delineates the distinctions among approaches through comparative analysis. Simultaneously, solutions for optimizing mutagenesis are presented to tackle the principal challenges mentioned above. This review aims to offer a theoretical foundation and practical guidance for the enhanced application of physical mutagenesis technologies in microbial breeding.

Mutagenesis↗

PubMed2026

Retinal microstructural alterations as early phenotypes of depression in radiogenomics analysis.

BACKGROUND: With the increasing prevalence of depression, there is an urgent clinical need for early screening in depression. The retina offers a promising window for early screening in depression due to its rapid, non-invasive, objective, eye-brain correlated characteristics, but previous research has yielded conflicting alterations in retinal microstructure in depression. METHODS: We screened retinal optical coherence tomography and brain magnetic resonance imaging data in the UK Biobank to enroll 23,225 participants for retinal study of depression occurrence, and 1475 participants for the eye-brain association study. We also used genetic data (ID: ebi-a-GCST90014267 and ukb-d-20,448) from the Integrative Epidemiology Unit Open Genome-Wide Association Study for Mendelian randomization analysis. We used Cox regression to assess the association between retinal microstructure and depression risk, Mendelian randomization to infer causality, and mediation analysis to explore retina-brain pathway association. RESULTS: The Cox regression analysis showed that retinal ganglion cell-inner plexiform layer (GCIPL) thickness remained a significant predictor of depression. The Mendelian randomization analysis indicated a positive statistical association between GCIPL thickness and depression. Moreover, there was a significant positive correlation (all p < 0.001) between the volume of specific depression-related brain regions and the GCIPL thickness. Adjusting for age, sex, and head size, the mediation analysis provided preliminary evidence for a potential anatomical pathway linking retinal GCIPL thickness to depression-related brain regions through primary visual cortex and secondary visual cortex volumes. CONCLUSION: Thickened retinal GCIPL is a potential early phenotype of depression and has a potential association pathway with depression-related brain regions using a radiogenomics approach.

Humans↗

PubMed2026

Prevalence of Theileria luwenshuni in goats (Capra hircus) on Tarama Island, Okinawa, Japan.

Caprine theileriosis is an economically important tick-borne disease caused by various Theileria species, particularly Theileria lestoquardi, Theileria luwenshuni, and Theileria uilenbergi, in goats (Capra hircus). Goat farming plays an economically and culturally important role on Tarama Island, Okinawa, Japan. Because goats on the island are mainly managed under an extensive grazing system, tick infestation is common. However, Theileria infections have not previously been investigated in goats on Tarama Island. To address this, archived DNA samples prepared from blood collected from 44 goats on Tarama Island were screened using a universal PCR assay targeting 18S rRNA sequences of Theileria and Babesia species. Two DNA samples were positive, and sequencing analysis of the amplicons identified T. luwenshuni. To further investigate the epidemiology of T. luwenshuni on Tarama Island, blood samples were subsequently collected from 96 goats across 19 farms. From each blood sample, a thin blood smear was prepared and genomic DNA was extracted. Microscopic examination of Diff-Quik-stained smears detected intraerythrocytic Theileria-like organisms in 35 (36.5%) goats. In addition, screening of DNA samples using a newly developed T. luwenshuni-specific PCR assay detected 77 (80.2%) positive goats, and the subsequent sequencing analysis confirmed the PCR results. Given that T. luwenshuni can cause severe disease in small ruminants, our findings highlight the importance of managing T. luwenshuni infection in goats on Tarama Island.

Animals↗

PubMed2026

Long-term mortality in pediatric sepsis: a systematic review and meta-analysis.

BACKGROUND: Pediatric sepsis represents a significant factor in the mortality rates among children, with survivors remaining highly fragile during the period following discharge. While in-hospital and short-term mortality have been widely studied, the long-term mortality of pediatric sepsis is not adequately synthesized or appreciated. This study aims to estimate the long-term mortality associated with pediatric sepsis, providing a basis for optimizing post-discharge surveillance and care protocols. METHODS: This systematic review and meta-analysis followed PRISMA guidelines and was registered in PROSPERO (CRD420251137504). Exhaustive searches were conducted in PubMed, Embase, the Cochrane Library, and Web of Science for studies published from the inception of each database to June 30, 2025. Studies reporting long-term mortality in pediatric sepsis patients diagnosed using international consensus criteria were included. After literature screening, long-term mortality was pooled using a random effects meta-analysis in R statistical software. RESULTS: A total of 72,065 records were identified through database searching. After removing duplicates and screening, six studies comprising 11,318 pediatric sepsis patients were included. The pooled long-term mortality in pediatric sepsis was 11% (95% CI: 7-16%), though significant heterogeneity was observed (I2 = 98.2%, p < 0.001). Sensitivity analyses yielded similar results, and evidence of publication bias was limited. CONCLUSION: Long-term mortality after pediatric sepsis was 11%, highlighting the persistent risk of mortality after hospital discharge. Further high-quality longitudinal studies are required to identify modifiable risk factors and guide evidence-based follow-up and personalized care.

Humans↗

PubMed2026

A conserved distal-tail helical extension defines a tailspike attachment architecture in Gram-negative siphophages.

Rapid growth of bacteriophage genome collections has outpaced functional annotation of tail-tip proteins, limiting comparative analysis of host-recognition structures. Starting from a shared distal-tail gene organization in the Salmonella phages 9NA and Jersey, I developed a morphogenetic bioinformatic framework integrating gene synteny, sequence comparison, profile hidden Markov model (HMM) screening, structural evidence, structure-aware searching, and AlphaFold modeling. Comparison with the experimentally characterized lambda and Sf11 tail assemblies identified a predominantly alpha-helical C-terminal extension of the distal-tail (DT) protein associated with tailspike attachment, termed the distal-tail helical extension (DT-helix). Screening 541,986 proteins from 5167 complete NCBI RefSeq tailed-phage genomes, followed by evidence-based evaluation of sequence, genomic context, and structural architecture, identified 165 curated DT-helical-extension-associated phages. Their DT proteins segregated into six sequence groups. In the four principal multi-member groups, cognate tailspikes showed group-specific conservation in proximal N-terminal regions but substantially greater downstream diversity, consistent with sequence constraint at the DT-tailspike attachment boundary. A complementary ProstT5/Foldseek search supported the established groups but revealed no convincing additional highly divergent family. Together with the experimentally characterized Sf11 attachment interface, these findings define a recurrent morphogenetic architecture linking conserved distal-tail scaffolds to more variable receptor-binding proteins across siphophages infecting Gram-negative bacteria. Although universal exchangeability is not established, the identified scaffold-receptor-binding boundaries provide a framework for molecular characterization and rational phage engineering. Accession-level information for the 165 curated phages is available through PhageTailDB.

Viral Tail Proteins↗

PubMed2026

Privacy, security, and reliability risks of artificial intelligence in healthcare: a systematic review of empirical evidence.

BACKGROUND: Artificial intelligence (AI) is increasingly integrated into healthcare information systems, supporting clinical decision-making, imaging analysis, and predictive modeling. While these applications offer operational and clinical benefits, they also introduce emerging risks to patient privacy, data security, and system reliability. OBJECTIVE: To systematically review empirical evidence on privacy breaches, security vulnerabilities, and misuse associated with AI applications in healthcare settings. METHODS: PubMed, Embase, Web of Science, Scopus, IEEE Xplore, and ACM Digital Library were searched for empirical studies published between January 2015 and November 2025 that evaluated AI use or misuse in clinical diagnosis, treatment, or decision-making. Two reviewers independently screened studies and extracted data using a standardized form. Findings were synthesized narratively due to heterogeneity in study designs, AI methods, and reported outcomes. RESULTS: Of 7,285 records identified through database searches and 205 through citation screening, 22 empirical studies met the inclusion criteria, spanning multiple clinical domains and data modalities, predominantly medical imaging applications. Five recurring threat categories were identified: patient re-identification, membership inference, unauthorized access and adversarial exploitation, input manipulation, and misuse or overinterpretation of AI outputs. Across studies, AI models were shown to encode latent biometric signals across diverse data types, limiting the effectiveness of traditional anonymization and synthetic data approaches. Adversarial attacks and input manipulation were also shown to compromise diagnostic performance and system integrity. CONCLUSION: This systematic review provides empirical evidence suggesting that contemporary AI systems in healthcare introduce privacy and security risks that may challenge traditional assumptions about data protection. These findings underscore the need for privacy- and security-by-design approaches and governance frameworks that address risks across the AI lifecycle.

Humans↗

PubMed2026

Data-centric, robust, and explainable multimodal deep learning for clinical decision support: A systematic review.

PURPOSE: Multimodal deep learning is increasingly proposed for clinical decision support (CDS) under a "data-centric" framing that prioritizes label quality, missing-modality robustness, distribution shift, calibration, and explainability. Prior reviews have examined multimodal medical AI, CDS, and data-centric methods separately, but none address their intersection. We mapped the modalities, fusion strategies, and data-centric and explainability techniques used in this recent literature, quantified how often each is implemented rather than merely mentioned, assessed deployment-relevant evidence (external validation, clinical-outcome measurement, equity), and formally appraised study-level risk of bias. METHODS: Following the PRISMA 2020 statement (PROSPERO CRD420261427815; registered retrospectively), we screened 150 records and included primary, clinical, multimodal studies that applied machine or deep learning to a decision-support task and reported at least one quantitative result. Two reviewers screened and extracted data with consensus adjudication. Each study was coded against pre-specified operational definitions, separating implemented or empirically evaluated techniques from those only mentioned. Study-level risk of bias was assessed with PROBAST + AI. Synthesis was narrative. RESULTS: Thirty-one studies met inclusion; 30 (97%) were published between 2024 and 2026, with a median of three modalities (range 2-6), most commonly structured EHR (71%) and imaging (39%). Data-centric techniques were frequently reported (74-84% across label-noise, distribution-shift, calibration, missing-modality and class-imbalance handling; equity 61%). However, external validation was reported in only 4/31 studies (13%), a clinical or provider outcome in 3/31 (10%), and no study reported routine deployment. Overall risk of bias was high in 27/31 studies (87%), driven by the analysis domain. CONCLUSION: Within this recent, self-selected slice of the field, technical robustness and explainability techniques are widely reported but rarely validated out-of-distribution or against clinical outcomes, and the underlying evidence is at high risk of bias. Progress requires external multi-site validation, clinical-outcome measurement, formal bias appraisal, and adherence to AI reporting standards (e.g., TRIPOD + AI) before deployment can be justified.

Deep Learning↗

PubMed2026

Machine learning-based prediction of unplanned readmission and construction of an online calculator for elderly patients with mild ischemic stroke.

OBJECTIVE: To screen for independent risk factors for unplanned readmission in elderly patients with mild ischemic stroke, and to construct and validate an online risk prediction calculator based on an interpretable machine learning model, thereby providing a promising practical tool for accurate clinical assessment of 30‑day all‑cause unplanned readmission risk in this population. METHODS: A prospective cohort study was conducted, including 1050 patients aged ≥ 60 years with mild ischemic stroke admitted between August 2023 and September 2024. Participants were randomly divided into a training set (840 cases) and a test set (210 cases) at a ratio of 8:2. Risk factors were screened by univariate analysis and multivariable Logistic regression. Four machine learning models, namely LightGBM, XGBoost, Random Forest, and K‑Nearest Neighbors (KNN), were developed and their performance was evaluated using AUC, accuracy, sensitivity, and specificity as metrics. The SHAP framework was used for interpretability analysis, and an online calculator was subsequently developed based on the optimal model. RESULTS: Univariate analysis showed significant differences (P < 0.05) in 13 factors including age, smoking, AIP, TyG index, HALP score, etc. Multivariable Logistic regression identified age (OR = 9.752), smoking (OR = 5.171), AIP (OR = 6.691), TyG index (OR = 4.393), HALP score (OR = 2.831), and ≥ 2 comorbidities (OR = 3.664) as independent risk factors. All four machine learning models demonstrated good predictive performance. Based on a comprehensive evaluation of multiple metrics and computational efficiency, the LightGBM model exhibited the best predictive performance (AUC = 0.884, accuracy = 0.829, sensitivity = 0.812, specificity = 0.875). SHAP analysis showed that age, AIP, TyG index, smoking, and HALP score were key predictors. An online calculator developed based on this model enables individualized risk predictions. CONCLUSION: Key risk factors associated with 30‑day unplanned readmission in elderly patients with mild ischemic stroke were identified. The LightGBM model demonstrated high predictive accuracy, and together with the interpretability analysis and online calculator, offers a practical tool to support clinical risk assessment. However, this tool requires future external validation.

Humans↗

PubMed2026

Integrated transcriptomic and metabolomic analysis of fluoride tolerance-related pathways and differentially expressed genes in silkworm strain XSKD.

XueSong KD (XSKD) silkworm strain exhibits prominent fluoride tolerance, yet the underlying molecular mechanisms of fluoride tolerance remains unclear. In the present study, fourth-instar pre-molting XSKD silkworms were used as experimental materials for integrated transcriptomic and untargeted metabolomic analyses. In total, 572 differentially expressed genes and 90 differential metabolites were screened. GO enrichment and KEGG enrichment based on the hypergeometric distribution model revealed that 13-Hydroxy-9Z,11E-octadecadienoic acid (13-(S)-HODE) acts as the core differential metabolite, which is significantly enriched in the linoleic acid metabolism pathway. Within this pathway, LOC101737302 and CYP338A1 display opposite expression trends and show correlations with pathway metabolites. Based on multi-omics data, this study preliminarily characterizes the lipid metabolic response under fluoride stress, providing omics dataset support for further in-depth exploration of the molecular mechanism of fluoride tolerance in silkworms.

Animals↗

PubMed2026

Algae-to-host horizontal gene transfer in Paramecium bursaria is associated with host adaptation during endosymbiosis.

Paramecium bursaria maintains a stable endosymbiosis with green algae, yet the evolutionary consequences of this association remain unclear. Here, we screened the host genome for algal-derived horizontally transferred genes (HTGs) using a lineage-aware workflow designed to detect horizontal gene transfer (HGT) between two defined lineages. We identified 16 candidate HTGs, including four putative newly transferred genes and 12 homologous transferred genes, most of which were functionally associated with redox homeostasis and metabolism. Five HTGs showed symbiosis-dependent expression. RNAi knockdown of GH32s and SATs reduced host proliferation, total cell area, and motility, while GH32s knockdown also reduced endosymbiont load. Duplication patterns suggest that most transfers may have occurred after the P. bursaria lineage diverged from the sampled Paramecium species but before its lineage-specific whole-genome duplication (WGD). The HTGs also showed host-associated shifts in GC content and gene length, while representative HTGs retained conserved domains and functional motifs. Together, our results support algae-to-host HGT in P. bursaria and suggest that some transferred genes may contribute to metabolic integration during endosymbiosis.

Gene Transfer, Horizontal↗