SEARCH · PubMed Health

Results for “Large Language Models”

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 721 records · Page 40Linked to original sources

PubMedDate not supplied

Four-year language outcomes of children exposed to cocaine in utero.

A large cohort of children exposed to cocaine in utero (n=189) were followed prospectively from birth to 4 years of age and compared to nonexposed children (n=185) on the Clinical Evaluation of Language Fundamentals-Preschool (CELF-P), a measure of receptive and expressive language abilities. Children exposed to cocaine in utero performed more poorly on the expressive and total language measures than nonexposed children after controlling for confounding variables, including prenatal exposure to alcohol, marijuana, and tobacco, as well as medical and sociodemographic variables. Children exposed to cocaine had more mild receptive language delays than nonexposed children and were less likely to have higher expressive abilities. Also, maternal factors such as language ability, performance IQ, race, and education correlated with child language abilities. Prenatal cigarette and marijuana exposure were related to deficits in specific language skills. Children placed in adoptive or foster care who were cocaine exposed demonstrated superior language skills compared to children exposed to cocaine who remained in biological relative or mother's care. These findings support a cocaine-specific effect on language skills in early childhood that may be modified with an enriched environment.

Analysis of Variance↗

PubMed2004

Neuropsychological characterization of cognitively-impaired-not-demented (CIND) individuals: clinical comparison data.

The primary objective of the present investigation was to characterize the neuropsychological test performance of a large sample of clinic-referred individuals diagnosed as Cognitively-Impaired-Not-Demented (CIND). Participants classified as Not-Cognitively-Impaired (NCI; n = 68) differed from CIND individuals (n = 205) on a number of demographic, clinical, and neuropsychological measures. A backward stepwise logistic regression analysis revealed that measures of learning and memory, visuoconstruction abilities, and cognitive flexibility provided the best discrimination between NCI and CIND participants. Clinical comparison data for CIND participants were generated for various demographically defined groups. The amount of inter-test scatter (highest minus lowest sample-based z-score) and the overall number of cognitive impairments (impairment being defined as performance equal to or greater than 1 standard deviation below the sample mean) in CIND individuals are reported. The results support the impression that CIND is a cognitively heterogeneous condition.

Adult↗

PubMed2005

Noise-robust soft clustering of gene expression time-course data.

Clustering is an important tool in microarray data analysis. This unsupervised learning technique is commonly used to reveal structures hidden in large gene expression data sets. The vast majority of clustering algorithms applied so far produce hard partitions of the data, i.e. each gene is assigned exactly to one cluster. Hard clustering is favourable if clusters are well separated. However, this is generally not the case for microarray time-course data, where gene clusters frequently overlap. Additionally, hard clustering algorithms are often highly sensitive to noise. To overcome the limitations of hard clustering, we applied soft clustering which offers several advantages for researchers. First, it generates accessible internal cluster structures, i.e. it indicates how well corresponding clusters represent genes. This can be used for the more targeted search for regulatory elements. Second, the overall relation between clusters, and thus a global clustering structure, can be defined. Additionally, soft clustering is more noise robust and a priori pre-filtering of genes can be avoided. This prevents the exclusion of biologically relevant genes from the data analysis. Soft clustering was implemented here using the fuzzy c-means algorithm. Procedures to find optimal clustering parameters were developed. A software package for soft clustering has been developed based on the open-source statistical language R. The package called Mfuzz is freely available.

Algorithms↗

PubMed2006

Cell_motility: a cross-platform, open source application for the study of cell motion paths.

BACKGROUND: Migration is an important aspect of cellular behaviour and is therefore widely studied in cell biology. Numerous components are known to participate in this process in a highly dynamic manner. In order to obtain a better insight in cell migration, mutants or drugs are used and their motive phenotype is then linked with the disturbing factors. One of the typical approaches to study motion paths of individual cells relies on fitting mean square displacements to a persistent random walk function. Since the numerous calculations involved often rely on diverse commercial software packages, the analysis can be expensive, labour-intensive and error-prone work. Additionally, due to the nature of algorithms employed the calculations involved are not readily reproducible without access to the exact software package(s) used. RESULTS: We here present the cell_motility software, an open source Java application under the GNU-GPL license that provides a clear and concise analysis workbench for large amounts of cell motion data. Apart from performing the necessary calculations, the software also visualizes the original motion paths as well as the results of the calculations to help the user interpret the data. The application features an intuitive graphical user interface as well as full user and developer documentation and both source and binary files can be freely downloaded from the project website at http://genesis.UGent.be/cell_motility . CONCLUSION: In providing a free, open source software solution for the automated processing of cell motion data, we aim to achieve two important goals: labs can greatly simplify their data analysis pipeline as switching between different computational software packages becomes obsolete (thus reducing the chances for human error during data manipulation and transfer) and secondly, to provide scientists in the field with a freely available common platform to perform their analyses, enabling more efficient data quality control through peer reviewing.

Animals↗

PubMed2005

Testing the agreement/tense omission model: why the data on children's use of non-nominative 3psg subjects count against the ATOM.

One of the most influential recent accounts of pronoun case-marking errors in young children's speech is Schütze & Wexler's (1996) Agreement/Tense Omission Model (ATOM). The ATOM predicts that the rate of agreeing verbs with non-nominative subjects will be so low that such errors can be reasonably disregarded as noise in the data. The present study tests this prediction on data from 12 children between the ages of 1; 8.22 and 3 ; 0.10. This is done, first, by identifying children who produced a reasonably large number of non-nominative 3psg subjects; second, by estimating the expected rate of agreeing verbs with masculine and feminine non-nominative subjects in these children's speech; and, third, by examining the actual rate at which agreeing verb forms occurred with non-nominative subjects in those areas of the data in which the expected error rate was significantly greater than 10%. The results show, first, that only three of the children produced enough non-nominative subjects to allow a reasonable test of the ATOM to be made; second, that for all three of these children, the only area of the data in which the expected frequency of agreeing verbs with non-nominative subjects was significantly greater than 10% was their use of feminine case-marked subjects; and third, that for all three of these children, the rate of agreeing verbs with nonnominative feminine subjects was over 30%. These results raise serious doubts about the claim that children's use of non-nominative subjects can be explained in terms of AGR optionality, and suggest the need for a model of pronoun case-marking error that can explain why some children produce agreeing verb forms with non-nominative subjects as often as they do.

Child↗

PubMed2001

Long-term antibiotics for preventing recurrent urinary tract infection in children.

BACKGROUND: Acute urinary tract infection (UTI) is common in children. By the age of seven years, 8.4% of girls and 1.7% of boys will have suffered at least one episode. Symptoms are systemic rather than localised in early childhood and consist of fever, lethargy, anorexia, and vomiting. UTI is caused by E. coli in over 80% of cases and treatment consists of a course of antibiotics. Due to the unpleasant acute illness caused by UTI and the risk of pyelonephritis-induced permanent kidney damage, many children are given long-term antibiotics aimed at preventing recurrence. However these medications may cause side effects and promote the development of resistant bacteria. OBJECTIVES: To determine the efficacy and side effects of long-term antibiotics given to prevent recurrent UTI in children. SEARCH STRATEGY: A search of MEDLINE (1966 to Jan 2001), EMBASE (1988 to Jan 2001) and the Cochrane Controlled Trials Register for relevant randomised controlled trials without language restriction; reference lists of review articles; contact with content experts. SELECTION CRITERIA: Randomised comparisons of two or more antibiotics and placebo with one or more antibiotics to prevent recurrent UTI. DATA COLLECTION AND ANALYSIS: Two reviewers independently assessed and extracted information. For each trial, information was collected on the methods of the trial, participants, interventions and outcomes. A random-effects model was used to estimate a summary relative risk (RR) and a summary risk difference (RD) for recurrent UTI. Heterogeneity tests and subgroup analyses were carried out based on a priori hypothesis of plausible effect modification. MAIN RESULTS: There were three trials (n = 151) comparing antibiotics with placebo/no treatment. The duration of antibiotic prophylaxis treatment varied among the studies (10 weeks to 12 months). The method of allocation concealment in the three trials was inadequate, unclear and adequate. The overall rate of recurrent UTI in the placebo/no treatment group was 63% (48/76). Compared to placebo/no treatment, antibiotics reduced the risk of recurrent UTI (RR 0.36, 95% CI 0.16 to 0.77; RD -46%, 95% CI -59% to -33%). No side effects were described in any of these three trials. There was one double-blinded trial (n = 120) with unclear allocation concealment that compared two different types of antibiotics to prevent recurrent UTI. Nitrofurantoin was more effective than trimethoprim in preventing recurrent UTI over a six month period (RR 0.48, 95% CI 0.25 to 0.92; RD -18%, 95% CI -34% to -3%). However, patients receiving nitrofurantoin were more likely to discontinue the antibiotic due to side effects (mainly gastrointestinal) than patients receiving trimethoprim (RR 3.17, 95% CI 1.36 to 7.37; RD 22%, 95% CI 8% to 36%). REVIEWER'S CONCLUSIONS: Most published studies to date have been poorly designed with biases known to overestimate the true treatment effect. Large, properly randomised, double blinded trials are needed to determine the efficacy of long-term antibiotics for the prevention of UTI in susceptible children.

Anti-Infective Agents, Urinary↗

PubMed1992

Early clinical differentiation of cerebral infarction from severe atherosclerotic stenosis and cardioembolism.

BACKGROUND AND PURPOSE: Hyperacute cerebral infarction trials require early differentiation of infarction subtype. Our aim was to determine clinical factors predictive of infarction subtype from data collected in the early hours of admission. METHODS: Using the 1,273 patients enrolled in the Stroke Data Bank, stroke risk factors and demographic, clinical, and radiological features were compared between the 246 cardioembolic and 113 large-vessel atherosclerotic cerebral infarcts. RESULTS: Stroke Data Bank definitions ensured more transient ischemic attacks in atherosclerotic infarcts and more cardiac disease in cardioembolic infarcts, but the diagnosis was distinguished further using a logistic regression model. Fractional arm weakness (shoulder different from hand) (odds ratio 3.1, 95% confidence interval [CI] 1.6-5.8), hypertension (odds ratio 2.8, CI 1.4-5.3), diabetes (odds ratio 2.5, CI 1.2-5.1) and male gender (odds ratio = 2.2, CI 1.2-4.1) occurred more frequently in patients with atherosclerotic than cardioembolic infarcts. Reduced consciousness (odds ratio = 3.2, CI 1.4-7.3) was more frequent in cardioembolism. For a male patient with hypertension, diabetes, and fractional arm weakness, the estimated odds of an atherosclerotic infarction were 47-fold that of a cardioembolic infarction. Patients with atherosclerotic infarcts were more likely to have a fractional arm weakness regardless of infarct size, whereas, for those with cardioembolic infarctions, fractional weakness was more frequent in infarcts less than 20 cc in volume. CONCLUSIONS: Clinical features that are observed at stroke onset can help distinguish cerebral infarction subtypes and may allow for early stratification in therapeutic trials.

Aged↗

PubMed2005

SNPHunter: a bioinformatic software for single nucleotide polymorphism data acquisition and management.

BACKGROUND: Single nucleotide polymorphisms (SNPs) provide an important tool in pinpointing susceptibility genes for complex diseases and in unveiling human molecular evolution. Selection and retrieval of an optimal SNP set from publicly available databases have emerged as the foremost bottlenecks in designing large-scale linkage disequilibrium studies, particularly in case-control settings. RESULTS: We describe the architectural structure and implementations of a novel software program, SNPHunter, which allows for both ad hoc-mode and batch-mode SNP search, automatic SNP filtering, and retrieval of SNP data, including physical position, function class, flanking sequences at user-defined lengths, and heterozygosity from NCBI dbSNP. The SNP data extracted from dbSNP via SNPHunter can be exported and saved in plain text format for further down-stream analyses. As an illustration, we applied SNPHunter for selecting SNPs for 10 major candidate genes for type 2 diabetes, including CAPN10, FABP4, IL6, NOS3, PPARG, TNF, UCP2, CRP, ESR1, and AR. CONCLUSION: SNPHunter constitutes an efficient and user-friendly tool for SNP screening, selection, and acquisition. The executable and user's manual are available at http://www.hsph.harvard.edu/ppg/software.htm

Algorithms↗

PubMed2004

A web services choreography scenario for interoperating bioinformatics applications.

BACKGROUND: Very often genome-wide data analysis requires the interoperation of multiple databases and analytic tools. A large number of genome databases and bioinformatics applications are available through the web, but it is difficult to automate interoperation because: 1) the platforms on which the applications run are heterogeneous, 2) their web interface is not machine-friendly, 3) they use a non-standard format for data input and output, 4) they do not exploit standards to define application interface and message exchange, and 5) existing protocols for remote messaging are often not firewall-friendly. To overcome these issues, web services have emerged as a standard XML-based model for message exchange between heterogeneous applications. Web services engines have been developed to manage the configuration and execution of a web services workflow. RESULTS: To demonstrate the benefit of using web services over traditional web interfaces, we compare the two implementations of HAPI, a gene expression analysis utility developed by the University of California San Diego (UCSD) that allows visual characterization of groups or clusters of genes based on the biomedical literature. This utility takes a set of microarray spot IDs as input and outputs a hierarchy of MeSH Keywords that correlates to the input and is grouped by Medical Subject Heading (MeSH) category. While the HTML output is easy for humans to visualize, it is difficult for computer applications to interpret semantically. To facilitate the capability of machine processing, we have created a workflow of three web services that replicates the HAPI functionality. These web services use document-style messages, which means that messages are encoded in an XML-based format. We compared three approaches to the implementation of an XML-based workflow: a hard coded Java application, Collaxa BPEL Server and Taverna Workbench. The Java program functions as a web services engine and interoperates with these web services using a web services choreography language (BPEL4WS). CONCLUSION: While it is relatively straightforward to implement and publish web services, the use of web services choreography engines is still in its infancy. However, industry-wide support and push for web services standards is quickly increasing the chance of success in using web services to unify heterogeneous bioinformatics applications. Due to the immaturity of currently available web services engines, it is still most practical to implement a simple, ad-hoc XML-based workflow by hard coding the workflow as a Java application. For advanced web service users the Collaxa BPEL engine facilitates a configuration and management environment that can fully handle XML-based workflow.

Computational Biology↗

PubMed2006

Telephone follow-up, initiated by a hospital-based health professional, for postdischarge problems in patients discharged from hospital to home.

BACKGROUND: It is known that many patients encounter a variety of problems in the first weeks after they have been discharged from hospital to home. In recent years many projects have addressed discharge planning, with the aim of reducing problems after discharge. Telephone follow-up (TFU) is seen as a good means of exchanging information, providing health education and advice, managing symptoms, recognising complications early, giving reassurance and providing quality aftercare service. Some research has shown that telephone follow-up is feasible, and that patients appreciate such calls. However, at present it is not clear whether TFU is also effective in reducing postdischarge problems. OBJECTIVES: To assess the effects of follow-up telephone calls in the first month post discharge, initiated by hospital-based health professionals, to patients discharged from hospital to home. SEARCH STRATEGY: We searched the following databases from their start date to July 2003, without limits as to date of publication or language: the Cochrane Consumers and Communication Review Group's Specialised Register, the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library), PubMed, EMBASE (OVID), BiomedCentral, CINAHL, ERIC (OVID), INVERT (Dutch nursing literature index), LILACS, Picarta (Dutch library system), PsycINFO/PsycLIT (OVID), the Combined Social and Science Citation Index Expanded (SCI-E), SOCIOFILE. We searched for ongoing research in the following databases: National Research Register (http://www.update-software.com/nrr/); Controlled Clinical Trials (http://www.controlled-trials.com/); and Clinical Trials (http://clinicaltrials.gov/). We searched the reference lists of included studies and contacted researchers active in this area. SELECTION CRITERIA: Randomised and quasi-randomised controlled trials of TFU initiated by a hospital-based health professional, for patients discharged home from an acute hospital setting. The intervention was delivered within the first month after discharge; outcomes were measured within 3 months after discharge, and either the TFU was the only intervention, or its effect could be analysed separately. DATA COLLECTION AND ANALYSIS: Two review authors independently assessed studies for inclusion and for methodological quality. The methodological quality of included studies was assessed using the criteria from the Cochrane Effective Practice and Organisation of Care Review Group. The data-extraction form was based on the template developed by the Cochrane Consumers and Communication Review Group. Data was extracted by one review author and checked by a second author. For as far it was considered that there was enough clinical homogeneity with regard to patient groups and measured outcomes, statistical pooling was planned using a random effects model and standardised mean differences for continuous scales and relative risks for dichotomous data, and tests for statistical heterogeneity were performed. MAIN RESULTS: We included 33 studies involving 5110 patients. Predominantly, the studies were of low methodological quality. TFU has been applied in many patient groups. There is a large variety in the ways the TFU was performed (the health professionals who undertook the TFU, frequency, structure, duration, etc.). Many different outcomes have been measured, but only a few were measured across more than one study. Effects are not constant across studies, nor within patient groups. Due to methodological and clinical diversity, quantitative pooling could only be performed for a few outcomes. Of the eight meta-analyses in this review, five showed considerable statistical heterogeneity. Overall, there was inconclusive evidence about the effects of TFU. AUTHORS' CONCLUSIONS: The low methodological quality of the included studies means that results must be considered with caution. No adverse effects were reported. Nevertheless, although some studies find that the intervention had favourable effects for some outcomes, overall the studies show clinically-equivalent results between TFU and control groups. In summary, we cannot conclude that TFU is an effective intervention.

Aftercare↗

PubMed1999

Cross-cultural differences in health-related quality of life of people with epilepsy: findings from a European study.

PURPOSE: To examine between-country differences in health-related quality of life (HRQOL) of adults with epilepsy across a large number of European countries. METHODS: Self-completion postal questionnaire sent to large sample of adults with epilepsy, recruited from epilepsy support groups or epilepsy outpatient clinics. The questionnaire was developed in English and translated. Back-translations from each language were checked for accuracy. The questionnaire sought information on clinical and socio-demographic details, and contained a number of previously validated scales of psychosocial well-being (the SF-36, the perceived impact of epilepsy scale, and a feelings of stigma scale). RESULTS: Controlling for socio-demographic and clinical characteristics, significant between-country differences were found in scores on the perceived impact of epilepsy scale, on seven of the eight SF-36 domains, and on the feelings of stigma scale. Respondents in Spain and the Netherlands fared consistently better, whilst those in France fared poorest, compared to those in other countries in terms of the various HRQOL measures used. CONCLUSION: Several possible reasons for the cross-cultural differences in HRQOL are proposed. Clearly, there is no single explanation and there may also be reasons which we have overlooked. This study emphasises the need for further comprehensive research in order that the position of people with epilepsy in different countries be more thoroughly understood in the social context.

Adolescent↗

PubMed2005

A neurocognitive approach to music reading.

Music reading offers a unique perspective on the acquisition of a notational system. Many people cannot read music, but a large proportion are motivated to learn. Musical literacy is therefore amenable to studies of acquisition in a way that language literacy is not. The studies reviewed here investigate how musical symbols on the page are decoded into a musical response. The studies address the nature of the mental representations used in music reading, as well as their instantiation within the brain. The results of a musical Stroop paradigm are described, in which musical notation was present but irrelevant for task performance. The presence of musical notation produced systematic effects on reaction time, demonstrating that reading of the written note, like the written word, is obligatory for those who are musically literate. Spatial interference tasks are also described that suggest that music reading, at least for the pianist, can be characterized as a set of vertical to horizontal mappings. These behavioral findings are mirrored by the results of an fMRI training study in which musically untrained adults were taught to read music and play piano keyboard over a period of three months. Specific learning-related changes were seen in the superior parietal cortex and fusiform gyrus, for melody reading and rhythm reading, respectively. These changes are suggested to correspond to the acquisition of processes that deal with the extraction of spatial and featural properties of notation, respectively.

Cognition↗

PubMed2004

Inferring pathways from gene lists using a literature-derived network of biological relationships.

MOTIVATION: A number of omic technologies such as transcriptional profiling, proteomics, literature searches, genetic association, etc. help in the identification of sets of important genes. A subset of these genes may act in a coordinated manner, possibly because they are part of the same biological pathway. Interpreting such gene lists and relating them to pathways is a challenging task. Databases of biological relationships between thousands of mammalian genes can help in deciphering omics data. The relationships between genes can be assembled into a biological network with each protein as a node and each relationship as an edge between two proteins (or nodes). This network may then be searched for subnetworks consisting largely of interesting genes from the omics experiment. The subset of genes in the subnetwork along with the web of relationships between them helps to decipher the underlying pathways. Finding such subnetworks that maximally include all proteins from the query set but few others is the focus for this paper. RESULTS: We present a heuristic algorithm and a scoring function that work well both on simulated data and on data from known pathways. The scoring function is an extension of a previous study for a single biological experiment. We use a simple set of heuristics that provide a more efficient solution than the simulated annealing method. We find that our method works on reasonably complex curated networks containing approximately 9000 biological entities (genes and metabolites), and approximately 30,000 biological relationships. We also show that our method can pick up a pathway signal from a query list including a moderate number of genes unrelated to the pathway. In addition, we quantify the sensitivity and specificity of the technique.

Algorithms↗

PubMed2000

Folic acid and folinic acid for reducing side effects in patients receiving methotrexate for rheumatoid arthritis.

OBJECTIVES: To assess the effects of folic acid and folinic acid in reducing the mucosal and gastrointestinal (GI) and haematologic side effects of low-dose of Methotrexate (MTX) in patients with Rheumatoid Arthritis (RA) and to determine whether or not folate supplementation alters MTX efficacy. SEARCH STRATEGY: We searched the Cochrane Controlled Clinical Trial's Register (CCTR), the Cochrane Musculoskeletal Group Specialized Register and Medline up to and including June 1999, using the search strategy developed by the Cochrane Collaboration (Dickersin 1994). We also handsearched the following: (i) bibliographic references; (ii) current contents of the last 6 months; (iii) abstracts of the rheumatology meetings; and (iv) all issues of four journals; Journal of Rheumatology, Arthritis & Rheumatism, Clinical and Experimental Rheumatology, and British Journal of Rheumatology. All languages were included. Principal investigators were also contacted in order to look for unpublished literature. SELECTION CRITERIA: We selected all double-blind, randomized, placebo-controlled, clinical trials (RCTs), in which adult RA patients were treated with a low dose of MTX (<20 mg / week) concurrently with folate supplementation. DATA COLLECTION AND ANALYSIS: Two observers extracted the data and assessed the quality of the trials. (BS, Z0) The overall treatment effect across trials was calculated using a fixed effect model. Disease activity was evaluated using standardized mean differences to ensure comparability across outcome measures. Results are presented with 95% Confidence Inervals (95% CI). Subgroup analyses were conducted evaluating different doses and sensitivity analysis looking at the quality of the trials. Publication bias was assessed with an inverted funnel plot technique. Heterogeneity of the trials was measured using a standard chi square test. Costs per month in different countries were compared. MAIN RESULTS: Of the 12 trials retrieved, 7 met the inclusion criteria. The total sample included 307 patients, of which 147 were treated with folate supplementation, 80 patients with folinic acid and 67 patients with folic acid. A 79% reduction in mucosal and GI side effects was observed for folic acid [OR = 0.21 (95% CI 0.10 to 0.44)]. For folinic acid, a clinically but non-statistically significant reduction of 43% was found [OR = 0. 57 (95% CI 0.28 to 1.15)]. No major differences were observed between low and high doses of folic or folinic acid. Haematologic side effects could not be analyzed, since details of each haematologic side effect by patients were not provided. No consistent differences in disease activity parameters were observed when comparing placebo and folic or folinic acid at low or high doses, although patients on high dose folinic acid had an increase in the number of tender joints, but not swollen joints. Large differences in costs across countries were found, but folinic acid was more expensive in all. REVIEWER'S CONCLUSIONS: The results support the protective effect of folate supplementation in reducing MTX side effects related to the oral and GI systems. We could not determine if folic was different from folinic acid. Therefore, for folinic acid to be considered cost-effective it must be found more effective than folic acid at reducing MTX side effects.

Antirheumatic Agents↗

PubMed2004

Comparison of character-level and part of speech features for name recognition in biomedical texts.

The immense volume of data which is now available from experiments in molecular biology has led to an explosion in reported results most of which are available only in unstructured text format. For this reason there has been great interest in the task of text mining to aid in fact extraction, document screening, citation analysis, and linkage with large gene and gene-product databases. In particular there has been an intensive investigation into the named entity (NE) task as a core technology in all of these tasks which has been driven by the availability of high volume training sets such as the GENIA v3.02 corpus. Despite such large training sets accuracy for biology NE has proven to be consistently far below the high levels of performance in the news domain where F scores above 90 are commonly reported which can be considered near to human performance. We argue that it is crucial that more rigorous analysis of the factors that contribute to the model's performance be applied to discover where the underlying limitations are and what our future research direction should be. Our investigation in this paper reports on variations of two widely used feature types, part of speech (POS) tags and character-level orthographic features, and makes a comparison of how these variations influence performance. We base our experiments on a proven state-of-the-art model, support vector machines using a high quality subset of 100 annotated MEDLINE abstracts. Experiments reveal that the best performing features are orthographic features with F score of 72.6. Although the Brill tagger trained in-domain on the GENIA v3.02p POS corpus gives the best overall performance of any POS tagger, at an F score of 68.6, this is still significantly below the orthographic features. In combination these two features types appear to interfere with each other and degrade performance slightly to an F score of 72.3.

Abbreviations as Topic↗

PubMed2003

Pentoxifylline for neonatal sepsis.

BACKGROUND: Although the overall incidence of neonatal sepsis has declined over the past decade, mortality remains high in the pre term infant. The high level of mortality and morbidity from sepsis despite the use of potent anti-microbial agents, and the global emergence of antibiotic resistance, have led to the search for new modalities to boost new born host defences. Pentoxifylline, a xanthine derivative and a phosphodiesterase inhibitor, has been shown to possess a broad spectrum of activity modulating inflammation. OBJECTIVES: The primary objective was to assess the effect on mortality and the safety of intravenous pentoxifylline as an adjunct to antibiotic therapy in neonates with suspected or confirmed sepsis. SEARCH STRATEGY: The Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 4, 2002), MEDLINE, EMBASE and CINAHL were searched in October 2002 and again in March 2003. Science Citation Index for articles referencing Lauterbach 1996 and Lauterbach 1999 was searched as well as proceedings of the Pediatric Academic Societies which were published in Pediatric Research from 1980. Doctoral dissertations and theses were searched from 1980. The reference lists of identified RCTs, and personal files were searched. No language restriction was applied. SELECTION CRITERIA: Studies were included if they were randomised or quasi-randomised trials, assessing the efficacy of pentoxifylline compared to placebo or no intervention as an adjunct to antibiotic therapy of suspected or confirmed sepsis in newborn infants less than 28 days old. Eligible trials were required to report treatment effects on at least one of the following outcomes: all cause mortality during initial hospital stay, neurological development at two years of age or later, length of hospital stay, duration of ventilation via endotracheal intubation, chronic lung disease in survivors, periventricular leukomalacia, necrotising enterocolitis, or adverse events. DATA COLLECTION AND ANALYSIS: Two reviewers independently abstracted information for the outcomes of interest. Any differences were resolved by mutual discussion. Typical Relative Risk (RR) and Risk Difference (RD) with 95% confidence intervals (CI) using fixed effects model are reported for dichotomous outcomes. NNT was calculated for outcomes for which there was a statistically significant reduction in RD. MAIN RESULTS: Two RCTs enrolled a total of 140 preterm (< 36 weeks) neonates with suspected late onset (> 7 days) sepsis to evaluate the effect of pentoxifylline on neonatal outcomes. However, the two studies reported outcomes of only the 107 randomised patients with confirmed sepsis. The results showed a reduction in 'all cause mortality during hospital stay' following pentoxifylline treatment [typical RR 0.14 (95% CI 0.03, 0.76), RD -0.16 (95% CI -0.27, - 0.04), NNT 6 (95% CI 4, 25)]. No adverse effects due to pentoxifylline were observed in the two included trials. No other outcomes of interest were reported. REVIEWER'S CONCLUSIONS: Current evidence suggests that the use of pentoxifylline as an adjunct to antibiotics in neonatal sepsis reduces mortality without any adverse effects. But the number of neonates studied is small and considerable methodological weaknesses exist in the included trials. Hence these results should be interpreted with caution. Researchers are encouraged to undertake large well-designed trials to confirm or refute the effectiveness of pentoxifylline to reduce mortality and adverse outcomes in neonates with suspected or confirmed neonatal sepsis. Researchers might also compare pentoxifylline with other adjuncts to antibiotics which modulate inflammation (e.g. intravenous immunoglobulins, haematopoetic colony stimulating factors among others) in reducing mortality and morbidity due to neonatal sepsis.

Anti-Bacterial Agents↗

PubMed2006

Defining the behavioral phenotype in children with fetal alcohol spectrum disorders: a review.

While there exists a large body of literature on cognitive functions in children with prenatal alcohol exposure, it remains undetermined if these children exhibit a unique profile of cognitive-behavioral functioning or a behavioral phenotype. Researchers have consistently found that intellectual functioning, as assessed by IQ tests, of children with prenatal alcohol exposure is deficient. There is increasing evidence that prenatal alcohol exposure is associated with slow information processing and attentional problems, in particular inattentiveness. Studies examining specific cognitive abilities such as language, visual perception, and memory in alcohol-affected children have shown performance decrements associated with increased task complexity. Children with prenatal alcohol exposure have also been found to exhibit significant deficits in daily functional skills or adaptive behavior, with deficits in socialization becoming pronounced during adolescence. The above findings point to the conclusion that a generalized deficit in complex information processing constitutes the central cognitive-behavioral characteristic of children with prenatal alcohol exposure. Researchers have consistently documented that specific brain regions are more affected by alcohol than other regions. The problem of mapping focal anomalies of the brain with a generalized pattern of deficits can be solved by taking developmental processes into consideration.

Animals↗

PubMed2007

Immigration-based disparities in adolescent girls' vulnerability to dating violence.

OBJECTIVES: Little data exists on dating violence among immigrant adolescents. The present study assessed disparities in experiences of physical and sexual dating violence based on immigrant status and language spoken at home among a large representative sample of adolescent girls. METHODS: Data from the 1997-2003 Massachusetts Youth Risk Behavior Surveys (N = 7,970) were analyzed. Adjusted logistic regression analyses were conducted among the full sample and sexually active sub-sample. To further clarify immigration-based effects, separate analyses were conducted within age and racial/ethnic groups. RESULTS: Being an immigrant was found to be protective against dating violence (OR 0.77, CI 0.60-0.98), but not among those reporting sexual intercourse. Stratified analyses revealed important differences in these effects based on age and race/ethnicity; only immigrant girls age 16 or older (OR 0.69, CI 0.48-0.99) and Hispanic immigrant girls (ORs 0.39-0.54) reported reduced risk for dating violence as compared to their non-immigrants peers. No differences in vulnerability to dating violence were detected based on immigrant status for Asian, Black, or White adolescents in stratified analyses. CONCLUSIONS: The social context of immigration may offer protection regarding adolescent girls' vulnerability to dating violence, but effects are not uniform across age, sexual experience, or race and ethnicity. Additional research is needed to understand how immigration, social behavior, age, race and ethnicity may interact to produce disparities in vulnerability to gender-based violence.

Adolescent↗