SEARCH · PubMed Health

Results for “Deep Learning”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

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

At least 199 records · Page 11Linked to original sources

PubMed2000

Learning gross anatomy in a clinical skills course.

Recent developments in undergraduate medical education in the United Kingdom have produced changes in the content and delivery of component courses, including human anatomy. Anatomy can retain its place in the medical course in the new world of problem-based learning and clinical skills teaching by gaining recognition as an integral part of the curriculum which underpins much of the practice of clinical medicine. In these new courses, anatomical information is clinically relevant and discussed in the context of medical problems and the acquisition of clinical skills. Students are encouraged to study in a manner in which information is retained (deep learning) and where understanding replaces rote learning of facts. Students take responsibility for their own learning, with appropriate support and resources. In clinical skills courses, anatomy underpins the development and retention of clinical knowledge and skills.

Anatomy↗

PubMed2026

Systematic review of machine learning approaches for predicting sickle cell crisis and mortality risk at the climate-health nexus.

BACKGROUND: Sickle cell anemia (SCA) is a severe genetic blood disorder characterized by recurrent vaso-occlusive crises and increased mortality, with the greatest burden occurring in low- and middle-income countries. Climatic and environmental conditions, including temperature variability, humidity, rainfall, air pollution, and seasonal changes, have been associated with disease exacerbation. However, the extent to which these factors have been incorporated into predictive models remains unclear. This study systematically reviews the application of machine learning (ML) models for predicting SCA crises and mortality in relation to climate and environmental factors. METHODOLOGY: The PRISMA guidelines were used, and 34 peer-reviewed studies published between 2005 and 2026 were analyzed to identify the climate variables, ML approaches employed, and predictive performance. The reviewed studies applied a range of ML techniques, including artificial neural networks, random forests, support vector machines, decision trees, logistic regression, and deep learning models. Temperature, humidity, rainfall, wind speed, air quality indicators, and seasonal patterns were the most frequently examined environmental variables. RESULTS: The findings indicate that most existing models rely predominantly on clinical and demographic data, with limited integration of climate information and inadequate representation of high-burden regions, especially Sub-Saharan Africa. Studies incorporating environmental variables reported improved predictive performance and highlighted the potential of climate-informed early warning systems for SCA management. CONCLUSION: The review recommends development of interdisciplinary, climate-aware ML frameworks, expansion of longitudinal environmental datasets, and increased research in underrepresented regions to support climate-resilient and patient-centered SCA care.

Humans↗

PubMed2003

Evaluation of a formative interprofessional team objective structured clinical examination (ITOSCE): a method of shared learning in maternity education.

Shared learning at undergraduate level provides a potential means of promoting a more multi-professional approach to maternity care. Interprofessional education uses shared interactive sessions to promote different professional groups' understanding of each other and working together. This paper describes the use of a formative objective structured clinical examination as a method of interprofessional education. Mixed groups of student doctors and student midwives rotate through a series of clinical stations based on common labour ward scenarios. After completing each station they are given feedback by a facilitator on their problem-solving skills, knowledge and attitude to team working. The interactive nature of the sessions encourages deep learning, is student centred and promotes a positive attitude to multidisciplinary working. Both student groups felt they benefited from shared learning in this way and that the formative OSCE was an effective method of developing their clinical skills.

Education, Medical, Undergraduate↗

PubMed2026

The Role of Artificial Intelligence Combined With Digital Cholangioscopy for Indeterminant and Malignant Biliary Strictures: A Systematic Review and Meta-analysis.

BACKGROUND: Current endoscopic retrograde cholangiopancreatography (ERCP) and cholangioscopic-based diagnostic sampling for indeterminant biliary strictures remain suboptimal. Artificial intelligence (AI)-based algorithms by means of computer vision in machine learning have been applied to cholangioscopy in an effort to improve diagnostic yield. The aim of this study was to perform a systematic review and meta-analysis to evaluate the diagnostic performance of AI-based diagnostic performance of AI-associated cholangioscopic diagnosis of indeterminant or malignant biliary strictures. METHODS: Individualized searches were developed in accordance with PRISMA and MOOSE guidelines, and meta-analysis according to Cochrane Diagnostic Test Accuracy working group methodology. A bivariate model was used to compute pooled sensitivity and specificity, likelihood ratio, diagnostic odds ratio, and summary receiver operating characteristics curve (SROC). RESULTS: Five studies (n=675 lesions; a total of 2,685,674 cholangioscopic images) were included. All but one study analyzed a deep learning AI-based system using a convoluted neural network (CNN) with an average image processing speed of 30 to 60 frames per second. The pooled sensitivity and specificity were 95% (95% CI: 85-98) and 88% (95% CI: 76-94), with a diagnostic accuracy (SROC) of 97% (95% CI: 95-98). Sensitivity analysis of CNN studies (4 studies, 538 patients) demonstrated a pooled sensitivity, specificity, and accuracy (SROC) of 95% (95% CI: 82-99), 88% (95% CI: 72-95), and 97% (95% CI: 95-98), respectively. CONCLUSIONS: Artificial intelligence-based machine learning of cholangioscopy images appears to be a promising modality for the diagnosis of indeterminant and malignant biliary strictures.

Humans↗

PubMed2003

[The interactive lecture. A simple form of student-activating learning].

INTRODUCTION: Activation of students in lectures to enhance learning by means of questions to be answered in buzz groups has been described in pedagogic handbooks and articles. We tried the concept in three lectures in biochemistry in order to evaluate use of time, training requirements of the lecturer, and method acceptance by students. MATERIAL AND METHODS: The experiment was carried out with a group of 87 medical students from a 4th semester course in biochemistry. Evaluation was made by direct observation and analysis of quantitative and qualitative data from a student questionnaire. RESULTS: Buzz groups and questions took less time than anticipated and not more than ten minutes of the lecture time. The lecturer needed supervision from a colleague to function well. Acceptance of the procedure was high among the students. Qualitative data indicate that students used more time for self-studies and moved towards deep learning. DISCUSSION: We conclude that interactive lecture could be implemented without major problems in lecture based educational programmes and that it is useful for the learning of the students.

Biochemistry↗

PubMed2025

Leveraging functional annotations to map rare variants associated with Alzheimer disease with gruyere.

Increased availability of whole-genome sequencing (WGS) has facilitated the study of rare variants (RVs) in complex diseases. Multiple RV association tests are available to study the relationship between genotype and phenotype, but most do not fully leverage the availability of variant-level functional annotations. We propose genome-wide rare variant enrichment evaluation (gruyere), an empirical Bayesian framework that complements existing methods by learning global, trait-specific weights for functional annotations to improve variant prioritization. We apply gruyere to WGS data from the Alzheimer's Disease Sequencing Project to identify Alzheimer disease (AD)-associated genes and annotations. Growing evidence suggests that the disruption of microglial regulation is a key contributor to AD risk, yet existing methods have not examined rare non-coding effects that incorporate such cell-type-specific information. To address this gap, we (1) define per-gene non-coding RV test sets using predicted enhancer and promoter regions in microglia and other brain cell types (oligodendrocytes, astrocytes, and neurons) and (2) include cell-type-specific variant effect predictions (VEPs) as functional annotations. gruyere identifies 13 significant genetic associations not detected by other RV methods, four of which remain significant in omnibus tests. We find that deep-learning-based VEPs for splicing, transcription factor binding, and chromatin state are highly predictive of functional non-coding RVs. Our study establishes a robust framework incorporating functional annotations, coding RVs, and cell-type-associated non-coding RVs to perform genome-wide association tests, uncovering AD-relevant genes and annotations.

Alzheimer Disease↗

PubMed2026

Community-driven advances in computational mass spectrometry: The perspective of EuBIC-MS members.

Advances in data acquisition, artificial intelligence, and integrative bioinformatics are driving the rapid evolution of computational mass spectrometry, and in turn, transforming modern proteomics, metabolomics, and lipidomics. These developments have greatly increased the scale and complexity of mass spectrometry data, underscoring the importance of evolving accurate, transparent, efficient and reproducible data processing workflows. Addressing these challenges requires collaborative innovation that brings together expertise in software engineering, statistics, and biology. The European Bioinformatics Community for Mass Spectrometry (EuBIC-MS), an initiative of the European Proteomics Association (EuPA), fosters a culture of open, community-driven development through its biennial Developers Meetings and Winter Schools. This commentary summarizes the scientific background and outcomes of the EuBIC-MS Developers Meeting 2025, which took place in Novacella, Italy. Three keynote presentations highlighted major frontiers in the field: deep proteome and phosphoproteome profiling, text mining for protein-protein interaction extraction, and scalable proteomics for AI-driven drug discovery. Seven community-selected hackathons addressed emerging challenges such as single-cell proteomics data analysis, FAIR metadata extraction, deep learning frameworks, R-Python interoperability, and DIA validation. Together, these efforts demonstrate the potential for scientific and technical innovation to arise from open collaboration, and highlight how community-driven initiatives can accelerate progress in computational mass spectrometry. SIGNIFICANCE: Modern proteomics increasingly depends on computational advances to translate complex, high-dimensional data into biological knowledge. The EuBIC-MS Developers Meeting 2025 exemplifies how community-driven collaboration can directly accelerate this process by bringing together experts from bioinformatics, statistics, and experimental proteomics to co-develop open, interoperable, and reproducible analytical tools. By fostering shared software frameworks, transparent benchmarking, and collaborative problem solving, the EuBIC-MS community helps ensure that technological innovation translates into reliable biological insights. This collaborative model strengthens the foundation for quantitative, system-level understanding of proteomes and establishes a sustainable path for integrating artificial intelligence and next-generation data acquisition into routine biological discovery. This commentary shows some current highlights in the field of computational mass spectrometry and community-based approaches undertaken during the most recent Developers Meeting to solve these challenges. The approaches discussed and initiated during the meeting - ranging from deep proteome profiling and phosphosite mapping to text mining, single-cell data analysis, and FAIR metadata extraction - address key bottlenecks that currently limit the biological interpretability and comparability of proteomics data.

Mass Spectrometry↗

PubMed2026

Diagnostic performance of machine learning models versus established risk stratification for intracranial aneurysm rupture: a systematic review and bivariate meta-analysis.

BACKGROUND: Machine learning (ML) models have been proposed to improve the discrimination of intracranial aneurysm rupture status beyond established clinical risk stratification tools. However, reported performance is heterogeneous and the relative contribution of model architecture and feature dominance remains unclear. METHODS: We performed a Preferred Reporting Items for Systematic Reviews and Meta-Analyses-diagnostic test accuracy systematic review and diagnostic meta-analysis of studies evaluating ML models for intracranial aneurysm rupture discrimination. PubMed, Embase and CENTRAL were searched to February 2026. Sensitivity and specificity were pooled using a bivariate random-effects model, with summary receiver operating characteristic curves generated across training, internal testing and external validation datasets. Models were compared with regression-based approaches and Population, Hypertension, Age, Size of aneurysm, Earlier subarachnoid haemorrhage, Site of aneurysm (PHASES) scores. Subgroup and meta-regression analyses explored associations between algorithm family and feature domain. RESULTS: Sixty-two retrospective cohorts (29 709 patients 209 models) met the inclusion criteria. In training datasets, pooled sensitivity and specificity for ML were 0.81 (95% CI 0.75 to 0.85) and 0.83 (0.80-0.86), with an area under the curve (AUC) of 0.878, exceeding PHASES (AUC 0.667). In testing datasets, ML retained higher discrimination (AUC 0.837) than regression models (0.806) and PHASES (0.646). In external validation, sensitivity was preserved (0.82), but specificity declined (0.66). Deep learning demonstrated the highest AUCs (training and testing). Incorporation of haemodynamic or radiomic features improved pooled discrimination relative to morphology alone. Evidence of small-study effects and mostly unclear Prediction Model Risk Of Bias Assessment Tool ratings were observed. CONCLUSIONS: ML approaches demonstrate higher pooled discrimination for aneurysm rupture status than conventional risk scores in retrospective datasets, but reduced external validation specificity and heterogeneity limit confidence for clinical translation. Prospective, externally validated, calibrated models are required before integration into routine cerebrovascular risk stratification.

Humans↗

PubMed2026

The molecular landscape of chordoma: Current frontiers from multi-omics to artificial intelligence.

Chordoma is a rare and aggressive malignant bone tumor of the axial skeleton that has historically challenged clinicians due to its complex anatomical locations and a high recurrence rate of up to 85%. This review synthesizes the most recent advances in chordoma research and offers an overview of how multi-omics, advanced immunology, and artificial intelligence are reshaping the treatment paradigm. Central to its pathogenesis is the T-box transcription factor Brachyury, which this review highlights as both the pathognomonic diagnostic marker and the primary therapeutic vulnerability. Cutting-edge innovations targeting this driver include covalent small-molecule binders, targeted protein degradation, and peptide-centric CAR-T cells designed to attack the intracellular oncoprotein. The tumor immune microenvironment is functionally dynamic, and new dimensions in cellular therapy, such as dual-specific CAR constructs and NK-cell platforms, are being engineered to neutralize immunosuppressive factors. Beyond biological insights, the review emphasizes the role of computational biology, specifically how deep-learning and machine-learning models achieve expert-level precision in tumor segmentation and personalized survival forecasting. By integrating genomic, transcriptomic, epigenomic, and proteomic data, multiomics approaches can fully elucidate chordoma subtypes and underlying resistance mechanisms, ultimately paving the way for more precise and personalized therapeutic strategies.

Humans↗

PubMed2026

Use of synthetic data, a novel paradigm for immunopathology.

The complexity and heterogeneity of autoimmune diseases are only partially captured by current analytic tools, even when deep learning techniques are employed to intercept patterns beyond existing dogma. Synthetic data offer a newer paradigm through machine-generated reconstructions of real-world data that faithfully attempt to recapitulate biological and clinical patterns without creating duplicates and maintaining the privacy of the original ones. Synthetic data act as a magnifying lens, allowing predictions otherwise not possible on disease classification, progression, and therapeutic response. This approach has several advantages and is currently underutilized. Firstly, it provides cohort enrichment and equilibrates group imbalances. Second, it generates synthetic arms for both in vitro studies and human clinical trials, relevant to disentangle the rarity and heterogeneity of autoimmune diseases. Third, the platform allows applications beyond tabular registries, including medical images, genomics, and flow cytometry data. Last, 'digital twins' act through dynamic bidirectional links with the biological/clinical system counterpart, lending themselves to transformative opportunities for precision medicine. Herein, we discuss the current status of this fast-moving novel component of artificial intelligence and its implications for autoimmune diseases.

Humans↗

PubMed2026

Digital pathology, image analysis, and artificial intelligence in liver disease.

Advances in digital pathology, image analysis, and artificial intelligence (AI) are rapidly transforming how pathologists and researchers interact with tissue samples and enable the development of diagnostic tools that harness high-resolution whole-slide images; these advances are in turn creating new opportunities for research, education, and routine clinical care globally. Liver disease is no exception, and digital pathology and AI have many applications in the diagnosis of liver cancer and liver diseases and in the assessment and management of transplantation. Although quantitative image analysis techniques have been applied to liver disease in research settings for over 50 years, recent improvements in image resolution, data storage, and the availability of advanced AI methods such as deep learning have driven multiple exciting developments. In this Review, we summarise the advancements in digital pathology, image analysis, and AI in liver disease. Key challenges such as access to and the logistics of using digital solutions, quality issues, and appropriate guidance in research and clinical use are reviewed, along with potential solutions to these challenges in the context of liver pathology and liver disease. Digital technologies are well established in liver pathology research, and access in clinical practice is increasing, with potential to address current laboratory challenges. Further evaluation is required to assess real-world effectiveness, clinical safety, and implementation of AI tools in liver pathology.

Journal Article↗

PubMed2026

Disentangling covariate effects on single-cell-resolved epigenomes with DeepDive.

Understanding the effects of individual biological factors from single-cell-resolved epigenomic data is hindered by multicollinearity, particularly in human cohorts. We introduce DeepDive, a deep-learning framework designed to systematically disentangle known and unknown sources of variation in single-nucleus ATAC-seq data. DeepDive accurately reconstructs chromatin accessibility, outperforms state-of-the-art methods with incomplete covariate information, and robustly recovers true biological signals from even highly entangled covariates, unlocking counterfactual, "what-if," analyses. Applying DeepDive to pancreatic islet cells, we perform counterfactual analyses to prioritize covariates associated with a type 2 diabetes-linked beta-cell subtype and nominate transcription regulators. DeepDive offers a powerful and unbiased tool for mechanistic discovery in complex human disease cohorts.

disentanglement↗

PubMed1997

Nursing students' approaches to studying.

The Approaches to Study Inventory (ASI), developed by Entwistle & Ramsden (1983), was administered to all nursing students at an Australian university (response rate = 67%). The purpose was to find out whether ASI constructs also apply to nursing students and to see whether nursing students change in their study approaches in the course of their nursing education. The ASI was construct validated through factor analysis. While it was possible to reconstruct a majority of the subscales based on individual items, only the meaning and reproducing study orientations were supported. These two orientations also demonstrated satisfactory levels of internal consistency for group comparisons. The authors conclude that the ASI is a useful and robust instrument for use in nursing education with respect to the two main study orientations. Ideally, nursing education should successively pave the way for an increase in meaning orientation scores (deep learning) and a reduction in reproducing orientation scores (surface learning). However, in this study there was no change in study orientations from first to third year. The association between meaning orientation scores and academic performance was weak.

Adolescent↗

PubMed2026

Conditional Diffusion Model-Based Method for Annotation of Antibiotic Resistance Gene Properties.

The crisis of bacterial antibiotic resistance, which has led to a decline in the effectiveness of antibiotics originally used to combat bacterial infections, has emerged as an urgent challenge for public health. Antibiotic resistance genes (ARGs) are one of the key reasons for bacteria to develop resistance to antibiotics. Therefore, accurately identifying and annotating the critical properties of ARGs is of great importance for addressing the antibiotic resistance emergency. Although existing deep learning models demonstrate remarkable effectiveness in extracting local features from sequence data, they still face limitations in the capacity to further gain the enriched latent representations within the data. To address the critical challenge of extracting higher-quality representations from ARGs sequence data, we propose a novel ARGs properties annotation method based on the conditional diffusion model which is used to learn latent representations through domain-specific knowledge injection. Specifically, during the conditional information integration phase, we systematically incorporate ARGs' domain knowledge to guide the diffusion process in generating high-quality latent representations. To overcome information redundancy caused by direct concatenation of conditional information and intermediate features, we design a cross-attention mechanism that enables feature fusion between heterogeneous information sources, thereby enhancing further the quality of obtained representations. Experimental results on widely used data sets demonstrate the framework's effectiveness in achieving superior prediction performance compared to existing methods.

Anti-Bacterial Agents↗

PubMed2026

In vitro reconstitution of chromatin replication recapitulates symmetric histone recycling.

Symmetric histone recycling is vital for maintaining epigenetic inheritance upon eukaryotic DNA replication. Recent genome-wide studies have uncovered key determinants of this process, but how these factors collectively support parental histone transfer remains incompletely understood. Here, we successfully reconstitute histone recycling with 24 purified proteins and analyze the products digested by Micrococcal nuclease with Repli-pore-seq, the newly developed pipeline combining nanopore sequencing and deep-learning-based classification. As a result, we identify histones symmetrically recycled as tetrasomes or hexasomes on nucleosome-favorable sequences. We also observe the discordance of the recycled position between lagging and leading strands on the GC-rich DNA sequences. Moreover, removal of Pol δ, Pol32, Dpb3/4, Ctf4, Csm3/Tof1, or Mrc1 disrupts the balance of histone recycling between the two daughter strands, whereas removal of Ctf4, Csm3/Tof1, or Mrc1 additionally alters the positions at which histones were recycled. Furthermore, addition of the lagging-strand maturation factors Fen1 and Cdc9 enhances histone recycling to the lagging strand. These findings provide critical insights into the molecular players and mechanisms underlying symmetric histone recycling.

Histones↗

PubMed2026

Towards efficient perturbation for the noncoding genome.

Deciphering the functionality of the noncoding genome, which includes important cis-regulatory elements (CREs) and transcribed noncoding RNA genes, remains technically challenging. Here, using massively parallel genetic screening, we systematically benchmark the performance of five representative loss-of-function perturbation tools, including single-guide RNA (gRNA) mediated SpCas9 cleavage or CRISPR interference, and paired gRNA (pgRNA) involved dual-SpCas9, Big Papi (paired SpCas9 and SaCas9) or dual-enAsCas12a fragment deletion methods, in decoding the roles of the noncoding genome. For targeting CREs such as enhancers, dual-SpCas9 outperforms other methods with superior efficiency in destroying functional genomic regions. For perturbing noncoding RNA genes, in addition to dual-SpCas9, other RNA-targeting methods such as RNA interference are recommended to discriminate transcript-dependent or -independent roles. A deep learning model, DeepDC, with an associated web server, is built to facilitate optimal dual-SpCas9 pgRNA design for efficiently deleting a genomic fragment. Together, our work provides practical guidance on selecting appropriate loss-of-function tools to resolve the functional complexity of the noncoding genome.

CRISPR-Cas Systems↗

PubMed2026

The periphery of nuclear speckles defines a spatially and temporally regulated compartment of long-lived intron-retained RNAs that resolves during mitosis.

RNA localization adds a fundamental layer to gene expression by determining when and where translation-ready mRNAs become available, yet how this timing is coordinated with nuclear architecture and cell-cycle progression remains unclear. Here we identify a subnuclear RNA niche at the nuclear speckle periphery that couples intron retention to cell-cycle-timed RNA release. Using compartment-resolved transcriptional inhibition, sequence-based deep learning and single-molecule and super-resolution RNA imaging in human pluripotent stem cells, we define a class of nuclear RNAs with long-lived retained introns that persist for hours and are enriched in transcripts encoding regulators of genome maintenance and mitosis, including centromere and kinetochore assembly, DNA repair and telomere maintenance. Long-lived retained introns exhibit elevated GC content, predicted structural stability and enrichment for nuclear speckle-associated RNA-binding proteins. In interphase, these RNAs localize to a distinct nuclear speckle-peripheral RNA niche in a spatial arrangement conserved across cell types. During mitotic remodelling, they undergo coordinated, kinase-dependent splicing and are released into the cytoplasm of early G1 daughter cells. Together, these findings link cis-encoded intronic features, subnuclear organization and mitotic remodelling to temporal control of RNA fate.

Mitosis↗

PubMed2026

Translating functional molecular knowledge into crop-breeding success.

Historical plant breeding, which optimizes phenotypes through selective crossing guided by phenotypic evaluation and molecular markers, is limited by evolutionary constraints that hinder rapid crop improvement. A new paradigm, precision breeding, circumvents these limitations by targeting genetic variants through functional molecular knowledge. To generate this knowledge at scale, sequence-based deep learning leverages high-quality genome sequence data to predict variant effects at base-pair resolution. When linked to agronomically important traits, these predictions enable breeders to prioritize variants for precision selection or editing. Although it is still in the early stages of development, we foresee three key applications for this approach: introgressing genes from distant breeding pools, purging deleterious mutations and designing new plant ideotypes. Looking ahead, refined computational models will facilitate targeted editing and the systematic redesign of complex physiological processes to address emerging breeding goals under shifting environmental conditions.

Crops, Agricultural↗