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 325 records · Page 18Linked to original sources
PubMed2026

Development and Crossover Evaluation of an Artificial Intelligence-Assisted System for Solid Pancreatic Lesion Detection and Pancreatic Parenchyma Recognition in Endoscopic Ultrasonography (With Video).

BACKGROUND AND STUDY AIMS: Pancreatobiliary endoscopic ultrasonography (EUS) is technically demanding, and supervised training opportunities are limited. We developed an artificial intelligence (AI) overlay system for detecting solid pancreatic lesions (SPL) and recognizing pancreatic parenchyma (PP) and evaluated its effect on reader performance. PATIENTS AND METHODS: Across six centers, two deep learning-based models were trained using expert-annotated EUS frames. We then conducted a randomized, two-sequence, two-period crossover reader study in which eight endosonographers (five novices and three experts) interpreted image sets with and without AI assistance. The primary endpoint was superiority of sensitivity for SPL detection among novices; key secondary endpoints included specificity and PP recognition. RESULTS: From 118 patients, 120 SPL-positive/negative image sets and 160 PP-positive/negative image sets were constructed. Among novices, AI assistance improved SPL detection sensitivity (88.7% vs. 76.8%, p&#x2009;<&#x2009;0.001) and accuracy (86.4% vs. 78.7%), while specificity met the predefined noninferiority criterion (84.2% vs. 80.5%, p&#x2009;<&#x2009;0.001). For PP recognition, sensitivity increased numerically (86.3% vs. 83.3%) but did not meet the predefined superiority criterion (p&#x2009;=&#x2009;0.095); specificity met the noninferiority criterion (87.8% vs. 81.0%), and accuracy increased from 82.1% to 87.0%. Among experts, sensitivity was maintained for both tasks, whereas specificity increased with AI assistance. CONCLUSIONS: AI assistance improved SPL detection among novice endosonographers. For PP recognition, sensitivity increased without reaching statistical superiority, whereas specificity met the predefined noninferiority criterion. These findings support a potential adjunctive role for AI in EUS interpretation.

Humans↗
PubMed1990

Training in short-term psychotherapy.

A new approach to short-term psychotherapy, based on psychoanalytic theory is presented as 'Therapy of limited time and objectives'. Personal experience, psychoanalytic knowledge and flexibility for accepting technical modifications allows for familiarity with concepts such as 'available setting', 'operational time' and 'adequate therapeutic time', which together with supervision and deep psychoanalytic learning, emphasizing counter-transferential knowledge, shall help to achieve a true epistemophilic sentiment, essential for empathic observation and listening, basic for training in short-term psychotherapy.

Curriculum↗
PubMed2025

Network Interactions of Circulating FGF23, HRG-HMGB1, and Cardiac Disease in CKD.

KEY POINTS: Multitrait analysis of genome-wide association study boosts the statistical power to identify novel genetic traits for fibroblast growth factor 23. A functional genomics approach aided network discovery to identify histidine-rich glycoprotein (HRG) and high-mobility group protein box 1 (HMGB1) as key regulators of cardiac disease in CKD. Integration of clinical and genetic data enhances the discovery power and is crucial for understanding the genetic underpinnings of mineral bone disorder related to CKD. BACKGROUND: Genome-wide association studies (GWAS) have identified numerous genetic loci associated with mineral metabolism markers but have exclusively focused on single-trait analysis. In this study, we performed a multitrait analysis of GWAS (MTAG) of mineral metabolism, exploring overlapping genetic architecture between traits to identify novel genetic associations for fibroblast growth factor 23 (FGF23). METHODS: We applied MTAG to variants common to GWAS of five genetically correlated mineral metabolism markers in participants of European ancestry. We integrated UK Biobank GWAS for blood levels for phosphate, 25-hydroxyvitamin D, and calcium (n=366,484) and Cohorts for Heart and Aging Research in Genetic Epidemiology GWAS for parathyroid hormone (n=29,155) and FGF23 (n=13,716). We then used supervised and unsupervised deep machine learning to identify novel associations between genetic traits and FGF23. RESULTS: MTAG increased the effective sample size for mineral metabolism markers to n=50,325 for FGF23. After clumping, MTAG identified independent genome-wide significant single-nucleotide polymorphisms for all traits, including 62 loci for FGF23. Many of these loci have not been previously reported in single-trait analyses. Through a functional genomics approach, we identified histidine-rich glycoprotein (HRG) and high-mobility group box 1 (HMGB1) as master regulators of downstream canonical pathways associated with circulating FGF23, and both genes were highly enriched in hypertrophied cardiac tissue of deceased hemodialysis patients. In addition, we found that DNMT3A was associated with uremic toxin, 8-hydroxy-2-deoxyguanosine, a biomarker of DNA damage. In silico gene perturbation analysis revealed that DNMT3A is protective in patients with heart failure caused by hypertrophied or dilated cardiomyopathy. CONCLUSIONS: Our findings highlight the importance of MTAG analysis of mineral metabolism markers to boost the number of genome-wide significant loci for FGF23 to identify novel genetic traits. Functional genomics revealed novel networks that inform unique cellular functions and identified HRG and HMGB1 as key master regulators of FGF23 and cardiovascular disease in CKD.

bones, stones, and mineral metabolism↗
PubMed2025

AlphaGenome Enhances Personal Gene Expression Prediction but Retains Key Limitations.

In recent years, numerous genome AI models have been developed to elucidate the relationship between DNA sequence and gene expression. However, these models have faced criticism for their limited accuracy in predicting individual-specific gene expression. AlphaGenome, the current state-of-the-art in genome AI, achieves exceptional performance across a range of sequence-based predictive tasks, but its utility for personal expression prediction has not yet been assessed. In this study, we evaluate AlphaGenome's ability to predict personal gene expression and find that it significantly outperforms its predecessor. Using GTEx data, AlphaGenome improves the prediction of expression direction over Enformer, achieving an odds ratio of 3.0. In some cases, it even reverses previously observed negative correlations into positive ones. Moreover, AlphaGenome demonstrates improved performance for genes with known nonlinear sequence-expression relationships, though it uncovers mechanisms distinct from those identified by tree-based models.

deep learning↗
PubMed2026

Development of a PCR-based technique for genotyping UGT1A1 gene and distribution of rs3064744 alleles in the Russian population.

BACKGROUND: Accurate determination of tandem thymine-adenine (TA) repeat numbers in the UGT1A1 promoter region (rs3064744) is essential for diagnosing Gilbert's syndrome and personalizing therapy with toxic agents like irinotecan and atazanavir. However, traditional polymerase chain reaction (PCR) assays face severe limitations due to the AT-rich sequence and overlapping melting temperatures (Tm) of the highly homologous 7TA and 8TA alleles. In this context, melting curve analysis (MCA) employing fluorophore-quencher systems has emerged as a promising alternative. The purpose of this study was to develop a novel genotyping approach combining optimized aPCR-MCA analysis with an automated classifier to overcome the limitations posed by the differentiation of highly homologous alleles and to demonstrate its practical application, providing the distribution of rs3064744 genotypes across four regional cohorts of the Russian population. METHODS: A specialized Dual Head 1D-convolutional neural network (1D-CNN) ensemble with Test-Time Augmentation (TTA) was developed. The model was trained and internally validated on 1,620 engineered plasmid samples, and independently evaluated on an external clinical test set of 440 unique patient genomic DNA specimens. Real-time PCR was performed on CFX96 and DTprime platforms. Additionally, population-wide screening was conducted on 997 archival clinical samples from Moscow, Sakha (Yakutia), Dagestan, and Rostov regions. RESULTS: While 5TA and 6TA alleles were easily separated, absolute Tm distributions of 7TA and 8TA alleles overlapped significantly, and non-uniform Tm shifts of 0.8&#xa0;&#xb0;C-1.4&#xa0;&#xb0;C occurred across platforms. Conventional absolute Tm thresholding was therefore inadequate. By assessing relative morphological curve divergence against co-amplified 7TA/7TA and 7TA/8TA reference anchors, the 1D-CNN ensemble neutralized instrument noise. It achieved 100% accuracy on internal validation and 100% concordance (440/440) with clinical reference pyrosequencing. Population screening revealed that Dagestan, Yakutia, and Rostov cohorts closely align with the European population. Rare 5TA and 8TA alleles were detected at low frequencies in Yakutia and Moscow. CONCLUSION: Combining LNA-modified aPCR-MCA with a comparative 1D-CNN model successfully circumvents thermodynamic limitations and eliminates human operator bias. This integrated system offers an accessible, high-throughput, and clinically valid solution for routine UGT1A1 pharmacogenetic testing.

1D-CNN↗
PubMed2026

Fast, accurate construction of multiple sequence alignments from protein language embeddings.

Multiple sequence alignment (MSA) is a foundational task in computational biology, underpinning protein structure prediction, evolutionary analysis, and domain annotation. Traditional MSA algorithms rely on pairwise amino acid substitution matrices derived from conserved protein families. While effective for aligning closely related sequences, these scoring schemes struggle in the low-identity "twilight zone." Here, we present a new approach for constructing MSAs leveraging amino acid embeddings generated by protein language models (PLMs), which capture rich evolutionary and contextual information from massive and diverse sequence datasets. We introduce a windowed reciprocal-weighted embedding similarity metric that is surprisingly effective in identifying corresponding amino acids across sequences. Building on this metric, we develop ARIES (Alignment via RecIprocal Embedding Similarity), an algorithm that constructs a PLM-generated template embedding and aligns each sequence to this template via dynamic time warping in order to build a global MSA. Across diverse benchmark datasets, ARIES achieves higher accuracies than existing state-of-the-art approaches, especially in low-identity regimes where traditional methods degrade, while scaling almost linearly with the number of sequences to be aligned. Together, these results provide the first large-scale demonstration of the power of PLMs for accurate and scalable MSA construction across protein families of varying sizes and levels of similarity, highlighting the potential of PLMs to transform comparative sequence analysis.

Deep Learning↗
PubMed1981

[Non-medicinal treatment in psychogeriatrics (author's transl)].

Of the triad of possible concepts of psychogeriatric therapy which includes pharmacotherapy, psychotherapy and social therapy, the non-medicinal treatment of the elderly is dealt with here, starting from the experience of our own team in the outpatient and part inpatient sphere and in the interdisciplinary team. For psychotherapy in its strict sense, approaches are described which arise from the elaboration of deep psychology, learning theory or client-centered therapy. Forms of treatment are described and possibilities for concrete aids in the patients' environment outlined.

Aged↗
PubMed2026

Survival prediction for clear cell renal cell carcinoma based on deep multimodal synergistic survival network.

Objective.To propose a deep multimodal synergistic survival analysis framework (Deep Multimodal Synergistic Survival Network, DMSSN) to achieve accurate prognostic analysis for clear cell renal cell carcinoma (ccRCC).Methods.This study (DMSSN) utilized matched multimodal data from the Cancer Genome Atlas-KIRC database, including CT imaging data, whole slide images, copy number variation (CNV) features, and clinical data. Deep Canonical Correlation Analysis was employed to map heterogeneous modalities into a shared latent space. Contrastive learning was introduced to enhance semantic consistency across multimodal features, and a gating network was utilized for the adaptive fusion of multimodal information to achieve precise survival risk prediction for patients.Results.Experimental results demonstrated that DMSSN achieved a Concordance Index (C-index) of 0.8153 &#xb1; 0.0994, with a Log-rank testp-value of 1.6553&#xd7;10-11. DMSSN exhibited significant performance advantages over traditional statistical methods like Log-rank-Cox (0.7055 &#xb1; 0.0670) and machine learning methods such as Random Survival Forest (RSF) (0.6836 &#xb1; 0.1048). Furthermore, in comparison with similar deep learning approaches, DMSSN outperformed late fusion strategies (0.7493 &#xb1; 0.1211) and discrete-time survival models such as DeepHit (0.7655 &#xb1; 0.1041) and Nnet-surv (0.7694 &#xb1; 0.0635). Notably, DMSSN still achieved the best predictive performance when compared to the classic deep survival model DeepSurv (0.7919 &#xb1; 0.0978) and advanced state-of-the-art multimodal fusion frameworks like Context-Aware Transformer (0.7735 &#xb1; 0.0818) and Multimodal Co-Attention Transformer (0.8102 &#xb1; 0.0972). Ablation studies showed that removing any single modality led to a decline in performance, with the largest numerical decrease occurring after removing CT imaging features (C-index decreased to 0.7327), validating the complementarity of multimodal data and the pivotal role of radiomic features in prognostic assessment. Module ablation experiments further confirmed the effectiveness of the core components.Conclusion:By effectively integrating imaging, pathology, genomic, and clinical features, the DMSSN framework demonstrates superior performance and robustness in the survival prediction of ccRCC.

Carcinoma, Renal Cell↗
PubMed2002

An interactive, student-centered approach, adopting the SOLO taxonomy, for learning to analyze laboratory data in veterinary clinical pathology.

INTRODUCTION: The purpose of this study was to describe and evaluate an interactive, student-centered teaching strategy for learning to analyze laboratory data in veterinary clinical pathology. The strategy was designed to operate in tutorials of approximately one hour duration and adopted the structure of the observed learning outcome (SOLO) taxonomy in order to align with outcomes and assessment components of unit of study design and to encourage a deep approach to learning. METHODOLOGY: The teaching strategy adopted group discussion and reflective observation as core activities. Students worked alone in identifying abnormal laboratory data, in pairs in discussing possible reasons for the abnormalities, and in two larger groups in deciding on conclusions for, and further investigation and management of, the case. The final debriefing brought the two groups together to reflect, question, and reach a consensus about the case. The teaching strategy was evaluated on the basis of its success in encouraging interaction through discussion, developing self confidence in analyzing laboratory data, and enhancing understanding as to how the disciplines of veterinary clinical pathology and veterinary medicine interrelate. Evaluation used self-reflection, peer feedback, and a student questionnaire. CONCLUSION: The teaching strategy provided the opportunity for students to develop and practice an approach to the analysis of laboratory data in a manner consistent with current educational thinking on student-centered learning. The use of group discussion and significant reflective practice not only enhanced interpersonal skills but also encouraged a deep approach to learning, leading to ownership of knowledge and increased awareness of the worth of veterinary clinical pathology in the investigative process.

Animals↗
PubMed2006

The effect of problem-based learning on students' approaches to learning in the context of clinical nursing education.

The effect of problem-based learning (PBL) on nursing students' approaches to learning has received scanty attention in nursing education. The purpose of the present study was to evaluate the effect of PBL on students' approaches to learning in clinical nursing education. Using a one-group before-after quasi-experimental design, the revised two-factor Study Process Questionnaire (R-SPQ-2F) was administered to compare students' approaches to learning before and after a period of clinical education in which PBL was implemented. Focus group interviews were used to elicit from students their PBL experience. Of the 237 students who participated in the study, 187 returned the R-SPQ-2F, representing a response rate of 78.9%. Twenty-eight of the students also participated in focus group interviews. The R-SPQ-2F scores indicated that for the deep approach to learning, the post-test mean score was noticeably higher than that at the pre-test (p=0.005). No significant difference was observed between the pre-test and post-test mean scores for the surface approach to learning (p>or=0.05). The four themes inductively derived from students' descriptions of their clinical education experience (motivated to learn; self-direction in learning; active, interactive and student-centred learning; and enjoyment in learning) also suggested that the students adopted a deep approach to learning during a period of clinical education in which PBL was implemented.

Adult↗
PubMedDate not supplied

[Learning approaches used by undergraduate interns in the development of a medical specialty].

INTRODUCTION: Two learning approaches are examined: a superficial and a profound one. The purpose of the first one is to pass the subjects with the minimum effort; in the second one, there is a genuine interest in grasping knowledge and being acknowledged for one's achievements. The objective of this research was to identify the learning approach of undergraduate interns, according to sex, age, specialty and grade. MATERIAL AND METHODS: An analysis of conglomerates was used as a statistical tool. The sample was formed by 179 undergraduate interns of 19 medical and surgical specialties offered at a concentration hospital. The form R-SPQ-2F, proposed by Biggs, was used as an instrument. The reliability and validity of the instrument was estimated by means of Cronbach's alpha and factorial analysis of learning. RESULTS: 60.3% of the students had a deep approach to learning; women and students older than 28 years old in greater proportion, as well as those studying internal medicine and those in the third year. CONCLUSIONS: Undergraduate interns are students with a high level of participation in their own learning process. This is probably due to the demands of the profession, the filters of selection and the nature of the subjects.

Adult↗
PubMed1991

Serial processing and the "phonetic route:" lessons learned in the functional reorganization of deep dyslexia.

A patient with chronic agrammatic Broca's aphasia presented with symptoms of deep dyslexia, in which it is presumed that phonologic processing of the written message is disrupted. Functional reorganization of the "nonlexical" or phonetic route of reading was undertaken according to principles of A. R. Luria, whereby the patient must learn consciously to control behaviors that had formerly been unconscious and automatic. This patient's responses in treatment made it clear that the phonetic route encompasses at least two dissociable functions, grapheme-phoneme conversion and sequential analysis. This is discussed in light of Luria's ideas regarding the functions of the precentral and postcentral regions of sensori-motor cortex.

Adult↗
PubMed2005

Could proximal white subungual onychomycosis be a complication of systemic spread? The lessons to be learned from Maladie Dermatophytique and other deep infections.

There are several published cases where dermatophyte infections have spread systemically, resulting in widespread internal dissemination as well as spread to local lymphatics and lymph nodes. The best example is provided by the condition known as Maladie Dermatophytique. In this commentary the arguments are discussed for a potential role of lymphatic dissemination in the development of proximal white subungual onychomycosis, where invasion of the nail plate by fungi proceeds from the proximal nail fold.

Dermatomycoses↗
PubMed1989

Anxiety and study methods in preclinical students: causal relation to examination performance.

Stress and anxiety are substantially raised in many preclinical students in their first year at medical school. Although correlated with poor end-of-year examination performance, anxiety levels did not cause poor performance, but were themselves caused by previous poor performance in sessional examinations. Study habits showed declining deep and strategic approaches, and increasing surface ('rote-learning') approaches. Surface learning correlated with poor end-of-year examination performance, and was a result of previous poor sessional examination performance. Deep learning did not correlate with performance, whereas strategic learning correlated positively with examination success, even when measured 2 years previously during application to medical school.

Achievement↗
PubMed2004

Assessing the relationship of learning approaches to workplace climate in clerkship and residency.

PURPOSE: To determine what approaches to learning are adopted by clinical clerks and residents and whether these approaches are associated with demographic factors, specialty, level of training, and perceptions of the workplace climate. METHOD: In 2001-02, medical clerks (n = 532) and residents (n = 2,939) at five medical schools in Ontario, Canada, were mailed the Workplace Learning Questionnaire. The correlation between the approaches to learning at work and perceived workplace climate and the influence of gender, age, location, residency program and level of training on outcomes were measured. RESULTS: A total of 1,642 clerks and residents responded (47%). The factor structure and reliability of the Workplace Learning Questionnaire were confirmed for these respondents. A surface-disorganized approach to learning was correlated with perception of heavy workload (r = .401, p < .001). The deep approach to learning was correlated with perception of choice-independence in the workplace and a supportive-receptive workplace (r = .32, p < .001; r = .23, p < .001). The climate factors, perception of choice-independence and supportive-receptive workplace, were correlated (r = .60, p < .001). There were significant differences among the mean scores for scales based on residency, year of training, and location of training. CONCLUSIONS: Perception of the workplace climate was associated with the approach to learning in the workplace of clerks and residents. Perception of heavy workload was associated with less effective approaches to learning. These associations varied with the residency program and the level of training.

Adult↗
PubMed1988

Learning in mature mice (Peromyscus leucopus) subjected to deep hypothermia as neonates.

In certain species of nonhibernating rodents, although young nestlings cease breathing and heart action when their body temperature is lowered to near freezing, the nestlings need only be rewarmed to recover. This remarkable capacity for immediate recovery has been known many years, but long-range consequences of deep neonatal hypothermia have never before been investigated. Mice (Peromyscus leucopus) that had been exposed to four 2.5-hr episodes of deep (2-4 degrees C) hypothermia when 4-10 days old were later compared with littermate controls in their performance on two learning tasks. The two groups did not differ in their acquisition or extinction of a lithium-induced learned taste aversion to sucrose. Nor did they differ in learning to find a hidden platform in a swimming pool. Thus in a nonhibernating rodent species, deep hypothermia experienced neonatally--unlike similar hypothermia administered in adulthood--seems not to induce deficits in subsequent learning capabilities. The resistance of neonates to damage probably represents an adaptation, for their modest thermoregulatory abilities render them vulnerable to deep hypothermia in frigid environments.

Animals↗