Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Transfer learning”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 217 records · Page 12Linked to original sources
PubMed2003

Transfer of route learning from virtual to real environments.

The authors investigated the extent to which route learning in a virtual environment (VE) transfers to the real world. In Experiment 1, active VE exploration, on its own or with a map, produced better transfer of training than either no VE training at all or passive VE training; however, transfer was achieved after shorter training times with the map. Experiment 2 demonstrated that VE + map training was not superior to training with a map alone, and Experiment 3 demonstrated that the poorer performances observed after passive VE training were not simply due to a lack of attention but to the lack of active navigational decisions. The authors concluded that the present VE technology does not provide better route learning than studying a map.

Adolescent↗
PubMed1996

Plus-maze retest profile in mice: importance of initial stages of trail 1 and response to post-trail cholinergic receptor blockade.

Recent research has shown that a single undrugged prior experience of the elevated plus-maze produces significant behavioural changes upon 24-h retest in rats and mice. Typically, when reexposed to the maze, animals display an increased avoidance of the open arms and a corresponding preference for the enclosed sections of the apparatus. Using ethological analyses, the present series of experiments sought to further characterize this phenomenon in mice and to determine whether or not it involves cholinergic receptor mechanisms. Results confirmed that behaviour during Trial 2 is markedly different to that seen on initial exposure, and that such changes are independent of the duration of Trial 1 (2 vs. 5 min). Retest behavioural changes included reduced entry latencies, reduced open arm entries, less time on the open arms and centre platform, lower levels of exploratory head-dipping, and increased entries into and time spent in the closed arms. The importance to the retest phenomenon of the first few minutes of initial exposure was further suggested by min-by-min analyses of the behaviour of animals naive to the maze. Results showed that behaviour during the first min is characterized by high levels of risk assessment from the centre platform and relatively low, but equal, levels of open- and closed-arm exploration. From min 2 onwards, however, behaviour showed a marked change with increasing open arm/centre platform avoidance, increasing closed-arm preference, and decreasing levels of risk assessment and exploratory head-dipping. Thus, it would appear that this within-session aversive learning transfers between sessions to account for behavioural profiles on retest. Irrespective of the duration of Trial 1 (2 or 5 min), posttrial administration of the muscarinic antagonist, scopolamine (0.1-1.0 mg/kg), failed to significantly alter the behavioural changes seen between trials. Data are discussed in relation to the apparent sensitization of fear produced by plus-maze exposure, its possible relation to phobia acquisition, and the need for further research on underlying mechanisms.

Animals↗
PubMed2005

Laparoscopic radical prostatectomy--the experience of the German Laparoscopic Working Group.

PURPOSE: To present the current status of laparoscopic radical prostatectomy (LRP) in Germany, Austria and Switzerland with respect to transferability, learning curve, and outcome. MATERIAL AND METHODS: The data of 5824 patients who underwent LRP in 18 centers by 50 urologists from March 1999 to August 2004 were analyzed retrospectively. Three centers performed more than 500, and six more than 250 cases. A transperitoneal descending technique with was used in 2701, a transperitoneal ascending in 1234, an extraperitoneal descending in 1814, and an extraperitoneal ascending modification in 75 cases. Specimen showed pT2 in 3535, pT3a in 1555, pT3b in 623, and pT4 in 111 cases. RESULTS: Mean operating time averaged 211 (131-292) minutes, with shorter duration of the extraperitoneal descending technique. Conversion to open surgery averaged 2.4 (0-14.1) %. Re-intervention rate amounted to 2.7 (0.3-7.7) %. Complication rate averaged 8.9 (1.8-10.8) % including bleeding (0.3-2.5%) and rectal lesion (1.5-2.5%). The rate of positive margins was 10.6 (3.2-18) % for pT2- and 32.7 (20-38.5) % for pT3a-tumors Continence after 12 months was 84.9 (72-94) %. Data about potency (7 centers) revealed 52.5 (35-67) % full erections following bilateral nerve preservation. 5 year-PSA recurrence rate (3 centers) was 8.6 (4-15.3) % for pT2-tumors and 17.5 (15-20.6) % for pT3a-stages. CONCLUSIONS: The results confirm the efficacy of the training program with safe transfer of LRP (i.e. low complication rate), however including all known problems of a retrospective study.

Adult↗
PubMed1961

One-trial interhemispheric transfer of a learning engram.

By using spreading depression to decorticate temporarily one hemisphere in rats, a learning trace was established in the opposite functional hemisphere. Spontaneous transfer of the engram from the trained hemisphere to the untrained hemisphere does not occur when both hemispheres are functional. If, however, the animal was allowed to make one reinforced response, complete transfer of the engram to the untrained hemisphere occurred.

Animals↗
PubMedDate not supplied

Home-Based Life Saving Skills in Ethiopia: an update on the second phase of field testing.

Home-Based Life Saving Skills (HBLSS) was integrated over 3 years into a district-level child survival project coordinated through the Ministry of Health and Save the Children Foundation/US in Liben Woreda, Guji Zone, Oromia Region, southern Ethiopia. During late 2004, the second phase of the program was reviewed for performance, home-based management, learning transfer, and program coverage. The immediate posttraining performance score for HBLSS guides for "First Actions" was 87% (a 78% increase over the pretraining baseline) and 79% at 1 year (a 9% decrease from the immediate posttraining score). The home-based management score of women attended by HBLSS guides for "First Actions" was 89%, compared to 32% for women assisted by other unskilled attendants. HBLSS guides teach women and families in the community as they were taught, by using pictorial Take Action Cards, role-play and demonstration, and a variety of venues. Estimates of HBLSS coverage suggest that HBLSS guides attended 24% to 26% of births, and 54% of women giving birth were exposed to HBLSS training. The HBLSS field tests demonstrate a promising program that increases access to basic care for poor, underserved, rural populations who carry the greatest burden of maternal and neonatal mortality.

Educational Measurement↗
PubMed2026

Automated CEAP Classification of Venous Duplex Reports Using Multimodal Artificial Intelligence.

OBJECTIVE: To develop and internally validate a prototype multimodal artificial intelligence system for automated CEAP (Clinical, Etiological, Anatomical and Pathophysiological) classification of venous duplex ultrasound (VDUS) reports, integrating natural language processing of free-text components with computer vision analysis of hand-drawn anatomical diagrams. METHODS: Single centre retrospective observational study using routinely collected clinical data. One thousand consecutive venous duplex ultrasound reports from Cambridge University Hospitals NHS Foundation Trust, UK (July 2024 - May 2025) were labelled according to the CEAP classification, excluding the Etiological component, which could not be reliably determined from duplex reports alone. Transfer learning was applied using ClinicalBERT for text and MobileNetV3 for diagrammatic data. Clinical classes were predicted from request line text. Text- and image-based pathophysiological models were developed for four anatomical territories (Great Saphenous Vein, Small Saphenous Vein, Deep system, Perforators), combined using late fusion with probability averaging. RESULTS: The clinical CEAP model achieved accuracy of 0.91, macro-F1 of 0.82, and macro-AUC of 0.98. Pathophysiological prediction varied, with text models broadly outperforming image models. Fusion yielded heterogeneous benefits, improving SSV performance but reducing Deep system accuracy. The performance of the final pathophysiological CEAP fusion models varied across anatomical territories: accuracy ranged from 0.70-0.92 and macro-AUC from 0.80-0.92. CONCLUSION: This study demonstrates the feasibility of automated CEAP classification from VDUS reports. Despite class imbalance affecting minority class predictions, the strong discriminatory performance validates this multimodal ML model for extracting clinically meaningful information from real-world data. This approach offers potential, pending external validation, to streamline vascular services through automated triage and guideline-compliant decision making.

Artificial intelligence↗
PubMed1997

Intermanual transfer of training: blood flow correlates in the human brain.

In a previous study, we found that relearning of a task with one hand might negatively be influenced by previous, opposite hand training of the analogue task, Thut G., et al., Exp. Brain Res., 108 (1996) 321-327. Drawing of a figure with the right hand, following left hand training, was slower than right hand drawing of an unknown figure. These conditions were termed right hand transfer learning (rTL) and right hand original learning (rOL). The present study aimed to identify the cerebral areas associated with these influences by measuring regional cerebral blood flow (rCBF) in 16 right-handed, healthy subjects during rTL and rOL. Positron emission tomography and statistical parametric mapping were used. Compared with rOL, rTL was associated with increased rCBF in the left medial prefrontal cortex and the right prefrontal convexity. Individual rCBF changes in the area homotopic to the right prefrontal convexity furthermore correlated with individual changes in rTL performance. While the smallest rCBF increases were found in subjects with weakest slowing of rTL relative to rOL, highest rCBF increases were present when rTL slowing dominated. Comparisons between rTL and rOL, however, revealed on average no performance differences. Our data suggest that relearning after previous opposite hand training activates neural mechanisms within the prefrontal convexity which might have an inhibitory function but that inhibition does not have to be the net final behavioral result.

Adult↗
PubMed1980

A computer-based therapy for the treatment of aphasic subjects with writing disorders.

A computer-controlled rehabilitation for aphasics with writing impairments is presented. Subjects were asked to type words under dictation. Each time a letter was typed in its correct position, it was displayed on a screen. If the contrary, the error was not displayed, thus avoiding visual reinforcement of false choices. This method of rehabilitation has proved efficient as concerns typewriting. More importantly, some learning transfer to handwriting was observed at the completion of experimental training. The results showed a significant reduction in the number of misspelled words as well as in the erroneous choice and serial ordering of letters. The stability of the observed improvement is discussed in relationship to variables such as the time elapsed since brain damage and the type of writing difficulty.

Adult↗
PubMed2001

Recovery from moderate aphasia in the first year poststroke: effect of type of therapy.

OBJECTIVES: (1) To determine whether 2 model-based remediation programs affect writing performance in unselected subjects with moderate aphasia and whether there is consequent improvement in everyday life, and (2) to interpret the potential changes observed by recourse to a theoretical model. DESIGN: Consecutive sample, multiple baseline, within subject crossover study. SETTING: Ambulatory care units. PARTICIPANTS: Eight subjects with moderate aphasia from 6 to 12 months postonset. INTERVENTION: A standardized test for reading and writing skills was given at the beginning and the end of each therapy program and 1 month after therapy stopped. MAIN OUTCOME MEASURES: Functional outcome measures were the Communicative Abilities in Daily Living (CADL) test and subtests from standardized aphasia assessment. RESULTS: After the 2 programs, there was improved writing performance, which was maintained after therapy stopped. Patterns of improvement corresponded to each of the 2 programs. Learning transfer was observed on the CADL test and functional writing, but gains on oral language were limited. Only 1 program was effective for 6 of the 8 patients. CONCLUSION: Specific rehabilitation programs aid recovery from aphasic symptoms from 6 to 12 months postonset. Individual response is linked to type of treatment. The interpretation is linked to a model-based description of aphasic symptoms and mechanisms of functional recovery.

Activities of Daily Living↗
PubMed2025

Predicting dynamic expression patterns in budding yeast with a fungal DNA language model.

Predicting gene expression from DNA sequence remains challenging due to complex regulatory codes. We introduce a masked DNA language model pretrained on 165 fungal genomes closely related to budding yeast that captures conserved regulatory grammar. Fine-tuning the LM on yeast RNA-seq data-including high-resolution transcriptional regulator induction time courses generated in this study-yielded Shorkie, a model that substantially improves gene expression prediction compared to baselines trained without self-supervision. Shorkie identified canonical transcription factor (TF) binding motifs and tracked their usage across induction experiments. Furthermore, Shorkie accurately predicted variant effects, outperforming leading sequence-to-expression models in cis-eQTL classification and achieving high concordance with massively parallel reporter assays. Interpretability analyses revealed Shorkie's ability to resolve promoter dynamics, splicing signals, and temporal changes in regulatory motif usage. This framework demonstrates that evolutionary-scale pretraining combined with transfer learning substantially improves our ability to decode gene regulation from sequence, providing insights into noncoding variants and regulatory networks.

Journal Article↗
PubMed2006

The management of new primary care organizations: an international perspective.

Management practice arising from parallel policies for modernizing health systems is examined across a purposive sample of 16 countries. In each, novel organizational developments in primary care are a defining feature of the proposed future direction. Semistructured interviews with national leaders in primary care policy development and local service implementation indicate that management strategies, which effectively address the organized resistance of medical professions to modernizing policies, have these four consistent characteristics: extended community and patient participation models; national frameworks for interprofessional education and representation; mechanisms for multiple funding and accountabilities; and the diversification of non-governmental organizations and their roles. The research, based on a two-year fieldwork programme, indicates that at the meso-level of management planning and practice, there is a considerable potential for exchange and transferable learning between previously unconnected countries. The effectiveness of management strategies abroad, for example, in contexts where for the first time alternative but comparable new primary care organizations are exercising responsibilities for local resource utilization, may be understood through the application of stakeholder analyses, such as those employed to promote parity of relationships in NHS primary care trusts.

Community Health Planning↗
PubMed1994

Change in wheelchair transfer performance during rehabilitation of men with cerebrovascular accident.

OBJECTIVES: A retrospective study was conducted to investigate differences in the functional outcome between patients with left cerebrovascular accident (left CVA) and right cerebrovascular accident (right CVA) in wheelchair transfer performance. It was hypothesized that the level of improvement in left CVA and right CVA groups of patients would not be the same, due to the different constellation of neurological impairments. METHOD: Data were obtained from the admission and discharge scores on the Functional Independence Measure of 100 male patients, 50 with left CVA and 50 with right CVA. RESULTS: Both groups demonstrated a statistically significant improvement between admission and discharge for wheelchair to bed, wheelchair to toilet, and wheelchair to tub or shower transfers (p < .001). However, there were no statistically significant differences between the two groups at admission or discharge for any of the three transfers, even after statistical removal of the effects of age, number of days from onset of stroke to rehabilitation, and number of days in rehabilitation. CONCLUSION: The lack of difference between the groups at time of discharge may have obscured a difference in the rate of transfer learning during the course of rehabilitation.

Activities of Daily Living↗
PubMed1990

Managing the education function.

The challenge of hospital-based education is to motivate change in the participant to benefit the organization as a whole, while increasing individual competence and positively affecting the quality of service. Enabling people to transfer learning into effective work behavior is the goal of hospital education staff. To ensure a successful future, the manager of the education function needs to be passionately committed to service of the organization. The manager must be able, be seen as able, and be unafraid to assume the evolving role. On the cover of Jim Lundy's book, Lead, Follow or Get Out of the Way, he addresses a memo to those who lead or aspire to lead. He recommends the following: Share your goals and aspirations. Let the implementers be involved in the planning. Achieve clear understanding of expected results. Evaluate progress periodically and fairly. Reinforce the importance of others. Coach your subordinates for growth. Emphasize and reemphasize teamwork. Search constantly for improvement in understanding, performance, and results! (Lundy, 1986) These leadership strategies will serve the manager of the education function well. Choose to lead, follow, or get out of the way! Marilyn Ferguson shared in her 1986 address to the ASHET annual conference, "We can fear the future . . . or by taking courage from each other and tapping our hidden resources, we can help create it" (Ferguson, 1986).

Education Department, Hospital↗
PubMedDate not supplied

The health promotion internship: a sponsor's perspective.

Health promotion internships are an essential link between the university classroom and the corporate or public service environment. A well structured internship experience will provide the future health promotion professional with an opportunity to transfer learned theory into practical application within a controlled and supervised environment. The success of the internship experience is contingent upon the mutual effort of the student, academic institution, and sponsoring organization working together to develop a model that will reflect the goals and objectives of each group with primary long-term benefit accruing to the future health promotion professional. The purpose of this article is to explore the components of a successful health promotion internship model developed from the perspective of a sponsoring organization.

Commerce↗
PubMedDate not supplied

Symmetry diagnosis and treatment in coordination dynamics therapy.

Forward-backward movement symmetry of the human CNS was investigated through pattern change during exercising on a special coordination dynamics therapy and recording device in the forward and backward direction. A mirror-image symmetry shift was found in a patient with cerebellar injury for the attractor patterns (movement patterns with the highest temporal stability) for exercising on the special coordination dynamics therapy device in forward and backward direction, with respect to pace and trot gait coordinations. Symmetry diagnosis and symmetry treatment are of importance in CNS repair because of the learning transfer to the movement symmetry counterpart: training of backward walking improves forward walking The theoretical and practical basis of Schalow Coordination Dynamics Therapy is given in the Method section including symmetry considerations with respect to CNS self-organization.

Ataxia↗