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 271 records · Page 15Linked to original sources

PubMed2003

Approaches to learning, need for cognition, and strategic flexibility among university students.

BACKGROUND: Considerable research has described students' deep and surface approaches to learning. Other research has described individuals' self-regulated learning and need for cognition. There is a need for research examining the relationships among these constructs. AIMS: This study explored relationships among approaches to learning (deep, surface), need for cognition, and three types of control of learning (adaptive, inflexible, irresolute). Theory suggested similarities among the deep approach, need for cognition, and adaptive control (aspects of self-regulated learning); and among surface, inflexible, and irresolute control (aspects of an ineffective approach to learning). One-factor and two-factor models were proposed. SAMPLE: Participants were 226 Canadian military college students. METHOD: Participants completed the following questionnaires: the Study Process Questionnaire (Biggs, 1978), the Need for Cognition Scale (Cacioppo & Petty, 1982), and the Strategic Flexibility Questionnaire (Cantwell & Moore, 1996). RESULTS: Confirmatory factor analysis supported the identification of the six scale factors. Second order confirmatory factor analysis indicated three factors representing constructs underlying these factors. CONCLUSIONS: Neither the one- nor two-factor models accounted adequately for the data. Self-regulated learning was defined by measures of the deep approach to learning, need for cognition, and adaptive control of learning. The second factor divided into one factor consisting of irresolute control, the surface approach, and negative need for cognition; and another consisting of inflexible and negative adaptive control. Substantial relationships among scales support the need for further theory development.

Adolescent↗

PubMed2003

University students' approaches to learning first-year mathematics.

This study assessed reliability and validity of the Approaches to earning Mathematics Questionnaire, for 218 university students. The study also identified the relationship between subscales. Internal consistency as Cronbach alpha was .77 for the Surface Approach to Learning scale and .88 for the Deep Approach to Learning scale. Principal components analysis yielded a two-factor solution accounting for only 34.6% of variance. The factors were interpreted as Surface Approach and Deep Approach to learning mathematics, as in Australia. The former subscale scores were negatively correlated -.2 with the latter subscale scores.

Adolescent↗

PubMed2004

Epistemological exploration: generalization of learning styles and analytical skills between academic and religious materials.

Among 139 students (mean age 21.8, SD=3.5), use of Schmeck's Deep Processing learning style (looking for conceptual understanding) on academic materials correlated modestly with its use on religious materials. The same was true for Elaborative Processing (looking for associations and applications). Both Deep and Elaborative Processing of academic materials correlated with better Analytical Skills. Only Elaborative Processing of religious materials correlated with Religiousness. Religiousness correlated with poorer Analytical Skills on academic materials and with a more Concrete Divine Concept; however, specific religious affiliation made a difference. Our understanding of the role of contents of materials and characteristics of learners on the types of learning strategies used and competence with cognitive skills is still very limited.

Adult↗

PubMed1999

Intercalated degrees, learning styles, and career preferences: prospective longitudinal study of UK medical students.

OBJECTIVES: To assess the effects of taking an intercalated degree (BSc) on the study habits and learning styles of medical students and on their interest in a career in medical research. DESIGN: Longitudinal questionnaire study of medical students at application to medical school and in their final year. SETTING: All UK medical schools. PARTICIPANTS: 6901 medical school applicants for admission in 1991 were studied in the autumn of 1990. 3333 entered medical school in 1991 or 1992, and 2695 who were due to qualify in 1996 or 1997 were studied 3 months before the end of their clinical course. Response rates were 92% for applicants and 56% for final year students. MAIN OUTCOME MEASURES: Study habits (surface, deep, and strategic learning style) and interest in different medical careers, including medical research. Identical questions were used at time of application and in final year. RESULTS: Students who had taken an intercalated degree had higher deep and strategic learning scores than at application to medical school. Those with highest degree classes had higher strategic and deep learning scores and lower surface learning scores. Students taking intercalated degrees showed greater interest in careers in medical research and laboratory medicine and less interest in general practice than their peers. The effects of the course on interest in medical research and learning styles were independent. The effect of the intercalated degree was greatest in schools where relatively few students took intercalated degrees. CONCLUSIONS: Intercalated degrees result in a greater interest in research careers and higher deep and strategic learning scores. However, the effects are much reduced in schools where most students intercalate a degree. Introduction of intercalated degrees for all medical students without sufficient resources may not therefore achieve its expected effects.

Attitude↗

PubMed1995

Differential roles of cerebellar cortex and deep cerebellar nuclei in the learning of the equilibrium behavior: studies in intact and cerebellectomized lurcher mutant mice.

Three- to 6-month-old lurcher mutant mice (+/lc), which exhibit a massive loss of neurons in the cerebellar cortex and in the inferior olivary nucleus but whose deep cerebellar nuclei are essentially intact, were trained daily, for 9 days, to maintain their equilibrium upon a rota rod rotating at 20 or 30 revolutions per minute (rpm). Their scores were measured and their behavior upon the rotating rod quantified in comparison to those of matched control (+/+) mice. Lurcher mice were able to learn to maintain their equilibrium efficiently when rotated at 20 rpm but were not when rotated at 30 rpm. After cerebellectomy, the equilibrium capabilities of the animals were much altered, especially in +/lc. These results show that the deep cerebellar nuclei are sufficient for motor learning, provided the task is not too difficult (20 rpm), but that the cerebellar cortex is required when the task is more difficult (30 rpm). Therefore, it can be concluded that the adaptive motor capabilities of lurcher mice are less developed than those of control animals.

Animals↗

PubMed2025

A comparative study highlights superiority of LSTM in crop genomic prediction.

We systematically evaluated three key determinants affecting prediction accuracy and the algorithm performance differences based on fifteen state-of-the-art GP methods, and found LSTM suitable for capturing additive and epistatic effects. Genomic prediction (GP) has been developed as an important method supporting crop breeding. By utilizing the phenotype values result from GP, breeders could make decisions in the seedling stage that consequently benefit for cost saving. In recent years, machine learning emerged as an efficient technology to solve modeling problems in many fields, including crop breeding. However, numerous modeling approaches have hindered the application of GP since breeders struggle to choose. Therefore, a comprehensively methodological research with guiding significance is extremely necessary. In the present study, we systematically evaluated three key determinants affecting prediction accuracy and the algorithm performance differences based on fifteen state-of-the-art GP methods. As for genomic feature processing, we found feature selection (SNP filtering approach) performed better than feature extraction (PCA method). Specifically, the feature relationship dependent methods (GBLUP, RNN, and LSTM) as well as DNN architecture showed superior performance with feature selection. Marker density analysis showed positive correlation with prediction accuracy in a limited threshold. Comparison on effect of population size demonstrated a positive correlation between trait genetic complexity and the optimal population size required. By testing fifteen modeling methods, we found LSTM network displayed superior performance, achieving the highest average STScore (0.967) across six datasets. Further research using all cell states or the latest cell states of LSTM inputs demonstrated its architecture particularly adept with capturing additive and epistatic QTL effects among SNPs. In conclusion, our findings provide basic principles for implementing GP in breeding project to maximize prediction accuracy while maintaining cost-effectiveness.

Plant Breeding↗

PubMed2026

Investigating cross-organism prediction of prokaryotic essential proteins using unsupervised language model and ensemble strategy.

Cross-organism prediction of essential proteins is a critical task for drug discovery and microbial engineering, yet the generalizability of existing machine learning models across diverse species remains a significant challenge. In this study, we propose DeepPEP, a large language model-based framework designed to reliably transfer essential protein annotations between distantly related organisms. Utilizing 66 curated prokaryotic datasets, we systematically evaluated DeepPEP's cross-organism performance under various conditions. Initial pairwise predictions revealed a correlation between performance and evolutionary distance; however, further investigation demonstrated that integrating training data from multiple organisms yields superior predictive power. In a benchmark scenario designed to simulate real-world applications, DeepPEP outperformed the state-of-the-art tool Geptop 2.0, showcasing a robust ability to identify species-specific essential proteins. Finally, a case study on novel genomes confirmed the model's practical effectiveness. Our results suggest that DeepPEP is a powerful strategy for prokaryotic essential protein prediction, and the rigorous evaluation framework established in this study provides a new benchmark for the field.

Large Language Models↗

PubMed2026

Integrating histology and spatial transcriptomics via multimodal transformers and contrastive representation learning for accurate gene expression prediction.

Predicting spatial gene expression from Histological images is a fundamental task in understanding tissue organization and molecular phenotypes. However, existing methods often rely on single-model representations or lack effective alignment between image and transcriptomic features. To address these limitations, we propose a unified multimodal learning framework that integrates histological imaging and spatial transcriptomics through a shared latent representation space. Specifically, histological H&E images are encoded by a ResNet50-based convolutional stem and a MobileViT Transformer backbone to extract hierarchical visual representations. Both modalities are projected into a shared latent space via linear-GELU-dropout transformation blocks, enabling cross-modal alignment through a contrastive learning objective that maximizes agreement between the corresponding image and the spot embeddings. Experimental results on the 10x Genomics Visium dataset of human liver tissue demonstrate that MViTGene achieves significantly higher prediction accuracy than existing methods across multiple gene subsets, with improvements of 20%, 33%, and 12% in predicting marker genes, highly expressed genes, and highly variable genes, respectively. The significant improvement in relevance indicates that the model can more accurately capture the true correspondence between tissue morphology and gene expression, therefore enabling more reliable biological interpretation. It provides a computational tool for high-throughput spatial gene expression prediction that balances performance and interpretability.

Humans↗

PubMed1980

Interaction of laminae of the cingulate cortex with the anteroventral thalamus during behavioral learning.

Neurons in deep laminae of the rabbit cingulate cortex develop discriminative activity at an early stage of behavioral discrimination learning, whereas neurons in the anteroventral nucleus of thalamus and neurons in the superficial cortical laminae develop such activity in a late stage of behavioral learning. It is hypothesized that early-forming discriminative neuronal activity, relayed to anteroventral neurons via the corticothalamic pathway, contributes to the construction of changes underlying the late-forming neuronal discrimination in the anteroventral nucleus. The resultant late discriminative activity in the anteroventral nucleus is then relayed via the thalamocortical pathway back to the superficial cortical laminae, promoting disengagement of cortex from further task-processing.

Animals↗

PubMed2005

Administration of epinephrine does not increase learning of fear to tone in rats anesthetized with isoflurane or desflurane.

Previous reports suggest that the administration of epinephrine increases learning during deep barbiturate-chloral hydrate anesthesia in rats but not during anesthesia with 0.4% isoflurane in rabbits. We revisited this issue, using fear conditioning to a tone in rats as our experimental model for learning and memory and isoflurane and desflurane as our anesthetics. Expressed as a fraction of the minimum alveolar anesthetic concentration (MAC) preventing movement in 50% of rats, the amnestic 50% effective dose (ED(50)) for fear to tone in control rats inhaling isoflurane and injected with saline intraperitoneally (i.p.) was 0.32 +/- 0.03 MAC (mean +/- se) compared with 0.37 +/- 0.06 MAC in rats injected with 0.01 mg/kg of epinephrine i.p. and 0.38 +/- 0.03 MAC in rats injected with 0.1 mg/kg of epinephrine i.p. For desflurane, the amnestic ED(50) were 0.32 +/- 0.05 MAC in control rats receiving a saline injection i.p. versus 0.36 +/- 0.04 MAC in rats injected with 0.1 mg/kg of epinephrine i.p. We conclude that exogenous epinephrine does not decrease amnesia produced by inhaled isoflurane or desflurane, as assessed by fear conditioning to a tone in rats.

Anesthesia, Inhalation↗

PubMed2005

Genetic susceptibility to scrapie in sheep: a clinically relevant theme in veterinary medical education.

RATIONALE FOR THIS STUDY: This article describes and evaluates two molecular biology practical classes based around the theme of genetic susceptibility to scrapie in sheep. These practical classes allow students to experience a range of molecular biology techniques in the context of a clinically based genetic disease. METHODOLOGY: The two molecular biology practical classes described are evaluated in terms of their perceived usefulness to study by first-year veterinary medicine students. The students' ratings are then assessed in relation to the approaches to studying (i.e., deep, strategic, and surface). These dimensions of learning are measured using the 52-item Approaches to Studying Inventory (ASI). RESULTS: The overall ratings from students in relation to both the practical classes were found to be positive. The scrapie genotyping practical was the highest-ranking laboratory-based practical in the first-year curriculum. Ratings in terms of usefulness to studies for both practical classes were found to be significantly higher for students with higher deep learning scores. CONCLUSION: The practical classes described here provide a clinically relevant scenario within which molecular biology concepts and methods can be illustrated to veterinary students. The positive correlation with deep learning is more evident for the scrapie genotyping practical than for the DNA extraction practical. This may reflects the complexity of the former, which is greater both technically and conceptually.

Animals↗

PubMed2004

Unlearning in health care.

Learning in health care is essential if healthcare organisations are to tackle a challenging quality of care agenda. Yet while we know a reasonable amount about the nature of learning, how learning occurs, the forms it can take, and the routines that encourage it to happen within organisations, we know very little about the nature and processes of unlearning. We review the literature addressing issues pivotal to unlearning (what it is, why it is important, and why it is often neglected), and go further to explore the conditions under which unlearning is likely to be encouraged. There is a difference between routine unlearning (and subsequent re-learning) and deep unlearning--unlearning that requires a substantive break with previous modes of understanding, doing, and being. We argue that routine unlearning merely requires the establishment of new habits, whereas deep unlearning is a sudden, potentially painful, confrontation of the inadequacy in our substantive view of the world and our capacity to cope with that world competently.

Humans↗

PubMed2005

The effect of a metacognitive intervention on approach to and self-regulation of learning in baccalaureate nursing students.

There is little empirical support in the nursing education literature related to the process of learning. This quasi-experimental study examined the effect of a metacognitive intervention (concept mapping) on approach to learning and self-regulation of learning in a sample of baccalaureate nursing students. Significant group differences were found in the concept mapping group, with an increase in deep approach to learning and adaptive control belief mean scores at the end of the semester. Students in the control group experienced a decrease in the deep approach to learning mean score and an increase in the surface approach to learning mean score at the end of the semester. Therefore, students who used concept mapping demonstrated an increase in deep approach to learning and the self-regulation of that learning, compared with students who did not use concept mapping. The results of this study provide empirical support for the use of concept mapping as a metacognitive intervention.

Adult↗

PubMed2005

[Perceptions of classroom goal structures, personal achievement goal orientations, and learning strategies].

We examined the relations among students' perceptions of classroom goal structures (mastery and performance goal structures), students' achievement goal orientations (mastery, performance, and work-avoidance goals), and learning strategies (deep processing, surface processing and self-handicapping strategies). Participants were 323 5th and 6th grade students in elementary schools. The results from structural equation modeling indicated that perceptions of classroom mastery goal structures were associated with students' mastery goal orientations, which were in turn related positively to the deep processing strategies and academic achievement. Perceptions of classroom performance goal stractures proved associated with work avoidance-goal orientations, which were positively related to the surface processing and self-handicapping strategies. Two types of goal structures had a positive relation with students' performance goal orientations, which had significant positive effects on academic achievement. The results of this study suggest that elementary school students' perceptions of mastery goal structures are related to adaptive patterns of learning more than perceptions of performance goal structures are. The role of perceptions of classroom goal structure in promoting students' goal orientations and learning strategies is discussed.

Child↗

PubMed2004

[Effects of test format on learning strategy and perceived utility].

This study investigated effects of test format on use of different learning strategies and their perceived utility. It was conducted in a classroom setting. Sixty seven (67) eighth grade students participated in the study, and were randomly assigned to one of two experimental conditions: fill-in test or essay test conditions. They took a history class for five days, and at the end of each session, took a test about the lecture with the format of their conditions. Results showed that in comparison with fill-in format, essay format facilitated use of deep-processing learning strategy, and decreased use of such strategies as rote memory. No significant effect was found for perceived utility of learning strategies, suggesting that it did not mediate the format effect. Underlying mechanism of the format effect and practical value of the current research were discussed.

Adolescent↗

PubMed2003

Bilateral buccal radicular groove in maxillary incisors: case report.

AIM: To present the rare localization of a radicular groove on the buccal aspect of a tooth and to discuss the pathology and management of the concomitant endo-periodontal defect. SUMMARY: Bilateral buccal radicular grooves were found on the maxillary central incisors of a 60-year-old female Caucasian. One groove was associated with deep local pocketing resulting in pulp necrosis and the formation of a periodontal-endodontic lesion. After endodontic treatment of the affected tooth, periodontal surgery was performed during which an apicoectomy was carried out on the root-filled tooth. Both the buccal grooves were removed by grinding, the roots were planed with curettes and a guided-tissue regeneration technique applied using amelogenin (Emdogain, Biora, Sweden). Following a period of 2 years, re-examination showed excellent healing with the complete elimination of the periodontal pocket on both incisors and significant radiographic evidence of bone regeneration. KEY LEARNING POINTS: Deep radicular grooves can predispose to pulp necrosis and the establishment of combined periodontal-endodontic lesions. Evaluation of clinical signs and appropriate diagnostic tests are of paramount importance in order to prevent incorrect diagnosis and treatment. Endodontists must be capable of performing advanced periodontal regeneration techniques during endodontic surgery.

Alveolar Bone Loss↗

PubMed2025

Foundation model based multimodal transformer framework for survival analysis in HER2 stratified breast cancer.

Objective. To improve survival prediction for HER2-positive breast cancer by integrating histopathological, molecular, and clinical data using a multimodal transformer framework.Approach. We propose a multimodal transformer framework for breast cancer survival prediction using HER2 stratified (SurvMBC), a foundation model-enhanced architecture that fuses three data modalities: whole-slide images, clinical narratives, and molecular features. Tumor microenvironment features are extracted using a pathology language and image pre-training (PLIP), clinical narratives are processed with BioBERT, and miRNA expression plus DNA methylation data are embedded using Gen2Vec. These representations are integrated through a cross-modal transformer with attention mechanisms for survival prediction.Main results. The model was evaluated on 1,095 HER2-positive breast cancer patients from The Cancer Genome Atlas. SurvMBC achieved a concordance index (C-index) of 0.857 (95% CI: 0.834, 0.880), a low integrated Brier score, and a strong inverse negative binomial log-likelihood. Risk stratification based on model outputs significantly separated high- and low-risk groups (log-rankp< 0.01) and showed strong associations with tumor stage, grade, and hormone receptor status (allp< 0.05).Significance. SurvMBC demonstrates the effectiveness of multimodal fusion in addressing tumor heterogeneity and improving prognostic accuracy. The attention-based integration enables context-aware learning of survival-relevant features across modalities, supporting individualized risk stratification and risk-adaptive treatment planning for HER2 stratified breast cancer patients.

Breast Neoplasms↗

PubMed2026

Multimodal artificial intelligence and machine learning in oncology: from data integration to precision cancer care.

Cancer remains a major global health burden, with approximately 20 million new cases and 9.7 million cancer-related deaths reported globally in 2022. While advances in radiological imaging, molecular profiling, and clinical data have enhanced the interpretation of disease progression, the availability of multiple such modalities still does not meet the needs of a large patient population. This narrative review focuses on the role of multimodal artificial intelligence and machine learning in bridging the gap in interpreting heterogeneous modalities to improve risk prediction, prognostic assessment, and treatment decision-making in precision oncology. Multimodal frameworks such as Pathomic Fusion illustrate how complementary histopathological and genomic information can be integrated for cancer diagnosis and prognostic modeling. Multimodal models have demonstrated potential in virtual biopsy, cancer screening, prognostic prediction, radiotherapy planning, intraoperative guidance, and clinical-trial design using digital twins and synthetic control arms. The major limitations of incorporating multimodal artificial intelligence and machine learning in oncology include data heterogeneity, demographic or institutional biases, and reproducibility challenges that hinder translation. Accordingly, appropriate data-governance strategies, fairness audits, and privacy-preserving approaches such as federated learning should be considered where appropriate. Future progress will depend on the development of standardized benchmarking datasets, robust external validation, seamless integration with electronic health records and picture archiving and communication systems, and the implementation of explainable, secure, and clinically validated multimodal artificial intelligence frameworks that support precision oncology in routine clinical practice.

deep learning↗