PubMed Health⌕ Search

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 253 records · Page 14Linked to original sources
PubMed2026

A voyage of reprogrammable metabolic bioengineering reshapes plant defense: from editing tools to synthetic systems.

Metabolic bioengineering has emerged as a transformative approach for reshaping plant defense by targeting intrinsic biosynthetic pathways to enhance immunity in modern agriculture. Moving beyond proof-of-concept metabolomics to broad-spectrum programmable pathway engineering addresses gaps in plant rational design and optimizes resilience in response to diverse environmental cues. This review aims to comprehensively highlight the transition of innovative approaches to phenolics, alkaloids, flavonoids, terpenoids, and benzoxazinoids, inferring adaptive reprogramming that mediates the growth-defense balance and functions as molecular sentinels in plants. Furthermore, decoding the volatile metabolome reveals a dynamic signaling interface that influences defense responses and stress-induced plant-microbe interactions, with the shikimate, jasmonate, and salicylate pathways functioning as central hubs for microbial deterrence and priming immune memory. Recent developments in multi-scalar genome-editing strategies, including CRISPR-driven combinatorial edits, enzyme orthogonalization, fluxomics, and spatially resolved multi-omics, reconfigure central and specialized metabolic fluxes toward improved defense function and regulation. Additionally, emerging tools, such as WUSCHEL2 and BABY BOOM transcriptional modules, and artificial engineering strategies integrating deep learning model-driven predictions facilitate rapid development of synthetic genetic circuits and support a predictive engineering of plants. Moreover, Mass spectrometry imaging (MSI) in spatial metabolomics enables to obtain structures and locations of unidentified endogenous metabolites within cells and tissues. Overall, this review emphasizes a diverse array of primary and secondary metabolites, spanning molecular concepts to recent advances in plant immune mechanisms. It also illustrates new frontiers in programmable metabolic engineering that accelerate the understanding of plant-microbe-metabolite cross-talks, offering strategies to improve plant resistance and advance sustainable agricultural solutions.

metabolic bioengineering↗
PubMed2026

Cross-Device Adaptation of Mirai for Mammography-Based Breast Cancer Risk Prediction.

Fine-tuning can adapt pretrained medical imaging models to new clinical datasets, but device-specific domain shifts may limit generalizability. We evaluated Mirai, a mammography-based deep learning model for breast cancer risk prediction, in a large screening cohort containing Hologic and General Electric (GE) full-field digital mammography systems, including GE Premium View (GE PV) and Tissue Equalization (GE TE) post-processing software. Native Mirai showed lower performance on TE images than on Hologic or PV images. Fine-tuning on TE images improved TE performance, particularly for short-term risk prediction, but substantially reduced performance on Hologic images, consistent with catastrophic forgetting. To mitigate this effect, we developed a device-invariant model using interleaved multi-device sampling and conditional adversarial training. This approach largely restored Hologic performance while maintaining improved TE performance, providing better robustness across heterogeneous imaging platforms. Comparison of cumulative and annual risk AUCs over a five-year time horizon further showed that performance gains were driven mainly by short- and intermediate-term predictions. These findings highlight both the value and dangers of device-specific fine-tuning and support balanced domain-adaptation strategies for deploying mammography-based risk models across diverse clinical imaging environments.

Journal Article↗
PubMed2005

Ambulatory teaching: do approaches to learning predict the site and preceptor characteristics valued by clerks and residents in the ambulatory setting?

BACKGROUND: In a study to determine the site and preceptor characteristics most valued by clerks and residents in the ambulatory setting we wished to confirm whether these would support effective learning. The deep approach to learning is thought to be more effective for learning than surface approaches. In this study we determined how the approaches to learning of clerks and residents predicted the valued site and preceptor characteristics in the ambulatory setting. METHODS: Postal survey of all medical residents and clerks in training in Ontario determining the site and preceptor characteristics most valued in the ambulatory setting. Participants also completed the Workplace Learning questionnaire that includes 3 approaches to learning scales and 3 workplace climate scales. Multiple regression analysis was used to predict the preferred site and preceptor characteristics as the dependent variables by the average scores of the approaches to learning and perception of workplace climate scales as the independent variables. RESULTS: There were 1642 respondents, yielding a 47.3% response rate. Factor analysis revealed 7 preceptor characteristics and 6 site characteristics valued in the ambulatory setting. The Deep approach to learning scale predicted all of the learners' preferred preceptor characteristics (beta = 0.076 to beta = 0.234, p < .001). Valuing preceptor Direction was more strongly associated with the Surface Rational approach (beta = .252, p < .001) and with the Surface Disorganized approach to learning (beta = .154, p < 001) than with the Deep approach. The Deep approach to learning scale predicted valued site characteristics of Office Management, Patient Logistics, Objectives and Preceptor Interaction (p < .001). The Surface Rational approach to learning predicted valuing Learning Resources and Clinic Set-up (beta = .09, p = .001; beta = .197, p < .001). The Surface Disorganized approach to learning weakly negatively predicted Patient Logistics (beta = -.082, p = .003) and positively the Learning Resources (beta = .088, p = .003). Climate factors were not strongly predictive for any studied characteristics. Role Modeling and Patient Logistics were predicted by Supportive Receptive climate (beta = .135, p < .001, beta = .118, p < .001). CONCLUSION: Most site and preceptor characteristics valued by clerks and residents were predicted by their Deep approach to learning scores. Some characteristics reflecting the need for good organization and clear direction are predicted by learners' scores on less effective approaches to learning.

Adult↗
PubMed2002

Postal survey of approaches to learning among Ontario physicians: implications for continuing medical education.

OBJECTIVES: To understand the approaches to learning of practising physicians in their workplace and to assess the relation of these approaches to their motivation for, preferred methods of, and perceived barriers to continuing medical education. DESIGN: Postal survey of 800 Ontario physicians. PARTICIPANTS: 373 physicians who responded. MAIN OUTCOME MEASURES: Correlations of approaches to learning and perceptions of workplace climate with methods, motives, and barriers to continuing medical education. RESULTS: Perceived heavy workload was significantly associated with the surface disorganised (r=0.463, P<0.01) and surface rational approach (r=0.135, P<0.05) to learning. The deep approach to learning was significantly correlated with a perception of choice-independence and a supportive-receptive climate at work (r=0.341 and 0.237, P<0.01). Physicians who adopt a deep approach to learning seem to be internally motivated to learn, whereas external motivation is associated with surface approaches to learning. Heavy workload and a surface disorganised approach to learning were correlated with every listed barrier to continuing medical education. The deep approach to learning was associated with independent learning activities and no barriers. CONCLUSIONS: Perception of the workplace climate affects physicians' approaches to learning at work and their motivation for and perceived barriers to continuing medical education. Younger, rural, family physicians may be most vulnerable to feeling overworked and adopting less effective approaches to learning. Further work is required to determine if changing the workplace environment will help physicians learn more effectively.

Attitude of Health Personnel↗
PubMed2025

Transfer learning with multiomics integration and deep neural networks reveals drug resistance mechanisms in cancer.

Drug resistance remains one of the primary challenges in effective cancer therapy. In this study, we employed a deep neural network (DNN)-based transfer learning (TL) approach to predict drug response and uncover drug resistance mechanisms. We integrated gene expression, somatic mutation, and copy number aberration (CNA) data with drug response profiles using multi-omics integration (MI). We used the Genomics of Drug Sensitivity in Cancer (GDSC) data for training and incorporated drugs with same pathways into the training models. We then evaluated drug response predictions on independent in-vivo PDX Encyclopedia (PDX) and ex-vivo the Cancer Genome Atlas (TCGA) datasets. In addition, we conducted pathway enrichment analyses to elucidate the mechanisms underlying drug resistance for paclitaxel, 5-fluorouracil (5-FU), gemcitabine, and cetuximab. We also applied Fisher's exact test (FET) to assess potential associations between drug resistance and the presence of mutations or CNAs. Our pan-drug models outperformed other methods based on the area under the precision-recall curve (AUCPR). Our pathway enrichment analyses revealed LDHB-mediated pyruvate metabolism and FYN-mediated focal adhesion might have pivotal roles in paclitaxel resistance, while PINK1-mediated mitophagy might be critical in 5-FU resistance. In addition to transcriptional activation, FET suggested that CNAs in LDHB and PINK1 may also be associated with resistance to paclitaxel and 5-FU, respectively. Furthermore, enrichment results for paclitaxel and cetuximab indicated shared resistance mechanisms between the two drugs. Importantly, our findings are consistent with prior experimental studies, providing literature-based validation of our results. Overall, our DNN-based TL approach achieved strong predictive performance across PDX & TCGA datasets and enrichment analyses provided valuable biological insights into drug resistance mechanisms.

Humans↗
PubMed1998

Differential roles of cerebellar cortex and deep cerebellar nuclei in learning and retention of a spatial task: studies in intact and cerebellectomized lurcher mutant mice.

Lurcher mutant mice (+/Lc) exhibit a massive loss of neurons in the cerebellar cortex and the inferior olivary nucleus, while deep cerebellar nuclei are essentially intact. To discriminate the relative participation of the cerebellar cortex and deep structures in learning and memory, 3 to 6-month-old +/Lc mice were subjected to a spatial learning task derived from the Morris water escape. They were able to learn to escape as well as their strain-matched controls (+/+). Seven days later, their scores showed that they had memorized the spatial environment but not as accurately as +/+ mice. Cerebellectomy before training did not significantly alter the escape learning capabilities of either group, whereas cerebellectomy performed after learning completely abolished retention in +/+, as well as in +/Lc, mice. These results suggest that the cerebellum, although not necessary for learning a spatial task, plays a crucial role in its retention, and that the storing structure of spatial information differs in +/+ and +/Lc mice.

Analysis of Variance↗
PubMed2005

Can we influence medical students' approaches to learning?

Students use three approaches to learning and studying: deep, surface and strategic. These are influenced by the learning environment. In response to the General Medical Council's report 'Tomorrow's Doctors', the second year of the medical course at the University of Edinburgh was changed to promote deep learning, with learning objectives constructed according to the SOLO taxonomy, learning methods such as problem-based learning and constructively aligned written assignments and examinations. The Approaches to Study Skills Inventory for Students (ASSIST) was used to evaluate the effect of these changes. Scores were highest for deep approaches and lowest for surface approaches and showed almost no change during the course. There are various possible explanations. The students already scored highly on deep approaches at the beginning of Year 2 and it may be difficult to increase the deep scores further, particularly over the relatively short period of the study. Alternatively, the effect of the changes in learning environment may not be strong enough to change entrenched approaches which have hitherto been successful.

Cohort Studies↗
PubMed2002

Attitudes among students and teachers on vertical integration between clinical medicine and basic science within a problem-based undergraduate medical curriculum.

Important elements in the curriculum at the Faculty of Health Sciences in Linköping are vertical integration, i.e. integration between the clinical and basic science sections of the curriculum, and horizontal integration between different subject areas. Integration throughout the whole curriculum is time-consuming for both teachers and students and hard work is required for planning, organization and execution. The aim was to assess the importance of vertical and horizontal integration in an undergraduate medical curriculum, according to opinions among students and teachers. In a questionnaire 102 faculty teachers and 106 students were asked about the importance of 14 different components of the undergraduate medical curriculum including vertical and horizontal integration. They were asked to assign between one and six points to each component (6 points = extremely important for the quality of the curriculum; 1 point = unimportant). Students as well as teachers appreciated highly both forms of integration. Students scored horizontal integration slightly but significantly higher than the teachers (median 6 vs 5 points; p=0.009, Mann-Whitney U-test), whereas teachers scored vertical integration higher than students (6 vs 5; p=0.019, Mann-Whitney U-test). Both students and teachers considered horizontal and vertical integration to be highly important components of the undergraduate medical programme. We believe both kinds of integration support problem-based learning and stimulate deep and lifelong learning and suggest that integration should always be considered deeply when a new curriculum is planned for undergraduate medical education.

Attitude of Health Personnel↗
PubMed1993

Automatic knowledge base refinement: learning from examples and deep knowledge in rheumatology.

MESICAR is a second generation expert system which contains very general descriptions of rheumatological disorders in the primary medical care field. With the help of a detailed hierarchical description of the human anatomy the system is able to support diagnostic decisions. The paper describes how machine learning techniques are used to automatically construct more specific disease descriptions for common, frequently occurring cases. The system MESICAR-LEARN implements a learning method which integrates analytical and empirical learning techniques. Cases diagnosed by MESICAR form the training examples, and MESICAR's knowledge base is used as domain theory. The learned concepts are integrated into a hierarchy of disease descriptions. They support efficient and fast reasoning on common cases in addition to the general diagnostic support afforded by MESICAR's deep knowledge.

Algorithms↗
PubMed1997

Learning in dental education.

Learning in dental education has long been a subject of academic research. There are different types of student learning which directly or indirectly contribute to the learning outcomes. In this paper, the various classifications of different types of learning; superficial, deep, individual, collaborative and resource-based learning are explored. In order to achieve the learning objectives, different teaching methods such as group-work, rôle-play and problem-based learning are employed. The types of resources used in present day teaching are diverse and multiple: such as print-based, computer-based technologies and audio-visual technologies. In order to assess the effectiveness of such teaching, different methods of assessment have been adopted for use in higher education. In the context of good teaching in higher education, there are a number of constraints placed upon the individual and the institution. These constraints are discussed and recommendations made whereby they may be overcome.

Education, Dental↗
PubMed2003

Approaches to learning: psychometric testing of a study process questionnaire.

BACKGROUND: One method of evaluating students' learning is to measure surface, deep and achieving approaches to learning using a questionnaire. In comparison with research on student nurses' learning styles, there has been little examination of their 'approaches to learning'. Much of the 'approaches to learning' research has been conducted with higher education students in Australia and Hong Kong and this kind of measurement is viewed as a valid and reliable way to assess learning. AIM: The aim of study reported here was to establish the validity of an 'approaches to learning' questionnaire, the study process questionnaire, for use with student nurses by undertaking psychometrical testing, including exploratory factor analysis. METHOD: The study process questionnaire is a 42-item questionnaire measuring surface, deep and achieving approaches to learning. It was distributed to 300 student nurses attending a common foundation programme in a higher education establishment in the United Kingdom (UK) in July 2000. Principal components analysis was conducted to determine the validity of the deep, surface and achievement scales in the questionnaire. RESULTS: A new factor structure was identified comprising three main scales which were similar in content but not identical to the original questionnaire. The deep factor correlated positively and significantly with grade performance average and sociology examination results. CONCLUSION: The study process questionnaire is a valid and useful tool for nurse teachers to gain knowledge about student nurses' approaches to learning. Deep learning appears to influence academic performance. More work is required to elucidate the complex nature of deep learning.

Achievement↗
PubMed2026

ClearDepthIAS enables automated high-throughput quantification of roots in soil-grown taproot crops.

Understanding root system architecture is critical for improving crop productivity and resilience, yet phenotyping root traits such as root growth angle and rooting depth remains technically challenging, especially at high throughput. Here, we present ClearDepthIAS, a high-throughput imaging and analysis platform that enables nondestructive, automated quantification of root architecture traits in taproot system crops. By capturing and stitching 360&#xb0; images of roots growing along the transparent walls of pots and applying deep learning-based segmentation (ClearDepth-WRT), we measured wall root shallowness (WRS)-a proxy for root growth angle-with high precision. We demonstrated for the tap root systems of soybean and canola that the system accurately detects root tips, quantifies their vertical distribution, and extracts biologically meaningful traits such as root area, distribution indices, and growth angles. Validation experiments in canola and soybean demonstrated that WRS can correlate with root crown architecture in mature plants, both in greenhouse and field settings. Furthermore, WRS and root distribution indices derived from ClearDepthIAS are predictors of early root architecture and can be correlated with root biomass distribution across soil depths under field conditions; however, environmental interactions may influence these relationships and weaken or even negate such correlations, as observed when comparing field to field variation in root system architecture. Our system enables efficient phenotyping of genetically diverse populations, with medium to high trait heritability, supporting its utility for genome-wide association studies and breeding. ClearDepthIAS accelerates the development of root ideotypes for improved resource acquisition and carbon sequestration, offering a scalable tool for supporting climate-resilient agriculture.

Plant Roots↗
PubMed2026

Biallelic loss of RB1 in hepatocellular carcinoma as synthetic lethal target for artificial intelligence-guided therapy.

The retinoblastoma (RB1) gene is a critical tumor suppressor that regulates cell cycle progression and genomic stability. Although RB1 alterations have been reported in hepatocellular carcinoma (HCC), the biological and clinical consequences of biallelic RB1 inactivation (RB1-Bi) remain poorly defined. We performed a comprehensive allele-specific genomic analysis of HCC patients from the TCGA-LIHC (n&#x2009;=&#x2009;355) and in-house AMC (n&#x2009;=&#x2009;206) cohorts, collectively comprising the AMC-TCGA discovery cohort. In this combined cohort, RB1-Bi was identified in 14.6% of tumors, was enriched in poorly differentiated HCCs and was independently associated with significantly reduced overall survival (adjusted hazard ratio 3.32, 95% CI 1.93-5.72, p&#x2009;<&#x2009;0.001). Additionally, a deep learning-based histopathology model using hematoxylin and eosin-stained slides (i.e., FR-MIL model) accurately predicted RB1-Bi status (F1 score 84.39% [95% CI, &#xb1;0.02]), making it readily identifiable in routine clinical practice. The prevalence and prognostic impact of RB1-Bi, as well as FR-MIL model performance, were consistent across independent validation cohorts, including advanced-stage tumors and external institutions. High-throughput drug screening in isogenic HCC models revealed that RB1-Bi HCC cells were particularly sensitive to inhibitors targeting mitotic regulators (e.g., AURKA, PLK1, KSP) and DNA damage response pathways (e.g., PARP inhibitors). Synthetic lethal interactions between RB1-Bi and these compounds were demonstrated in vitro and in vivo, and combination treatment with mitotic and PARP inhibitors had synergistic effects with acceptable tolerability. We conclude that RB1-Bi represents a clinically actionable biomarker that identifies a high-risk HCC subtype with specific therapeutic vulnerabilities, offering new opportunities for precision medicine.

Humans↗
PubMed1996

Deep assessment of machine learning techniques using patient treatment in acute abdominal pain in children.

Learning from patient records may aid knowledge acquisition and decision making. Existing inductive machine learning (ML) systems such us NewId, CN2, C4.5 and AQ15 learn from past case histories using symbolic and/or numeric values. These systems learn symbolic rules (IF... THEN like) which link an antecedent set of clinical factors to a consequent class or decision. This paper compares the learning performance of alternative ML systems with each other and with respect to a novel approach using logic minimization, called LML, to learn from data. Patient cases were taken from the archives of the Paediatric Surgery Clinic of the University Hospital of Crete, Heraklion, Greece. Comparison of ML system performance is based both on classification accuracy and on informal expert assessment of learned knowledge.

Abdomen, Acute↗
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↗