Connectionism, phonology, reading, and regularity in developmental dyslexia

Tests of the "phonological deficit" account of developmental dyslexia have produced apparently inconsistent results. We show how a connectionist approach to dyslexic reading development can resolve the paradox. A "dyslexic" model of reading was created by reducing the quality of the phonological representations available to the model during learning. The model behaved similarly to dyslexic children in that it had a selectively reduced ability to process nonwords, but showed normal effects of words' spelling-to-sound regularity. An experimental test of the model's predictions confirmed that dyslexic children perform similarly, in that they are impaired on irregular words to the same extent as nondyslexic children. It is concluded that developmentally dyslexic reading can indeed be understood in terms of impaired phonological representations and that the adoption of a modeling approach resolves an apparent paradox in the experimental literature.

Isolated-brain parallels to simple types of learning and memory in Tritonia

Physiological studies of learning and memory often require reduced nervous system preparations that can be trained by stimulation of neural pathways in a manner that mimics behavioral training. In the isolated brain preparation of the seaslug Tritonia, fictive swimming can be activated with a few electrical pulses applied to the cut end of a nerve, and learning experiments can be simulated with combinations of nerve stimuli. Repeated application of a nerve stimulus produced changes in fictive swimming that resembled habituation and iterative enhancement of multiple components of the swimming behavior. Many repetitions of the nerve stimulus led to fictive swimming failure. Stimulation of another nerve then restored the ability of the original nerve to activate fictive swimming, thus reproducing dishabituation results. Sensitization of threshold and latency were also simulated. Parallels between the behavior and the isolated brain suggest that the neuronal modifications underlying simple types of learning can be induced by nerve stimulation. The activity pattern of interneuron C2 in these experiments suggests that increased C2 synaptic efficacy may underlie sensitization of swimming latency.

Word-frequency effects on short-term memory tasks: evidence for a redintegration process in immediate serial recall

Four experiments investigated the mechanisms responsible for the advantage enjoyed by high-frequency words in short-term memory tasks. Experiment 1 demonstrated effects of word frequency on memory span that were independent of differences in speech rate. Experiments 2 and 3 showed that word frequency has an increasing effect on serial recall across serial positions, but Experiment 4 showed that this effect was abolished for backward recall. A model that includes a redintegration process that operates to "clean up" decayed short-term memory traces is proposed, and the multinomial processing tree model described by R. Schweickert (1993) is used to provide a quantitative fit to data from Experiments 2, 3, and 4.

Conservation of the H-2 BF1 binding motif 5' of the H-2Ds, Ks and Dq genes

The biological consequences of radiation leukaemia virus (RadLV) infection include the stimulation of H-2 antigen expression soon after injection of the virus. Early studies demonstrated that resistance to RadLV-induced leukaemia in certain mouse strains is mediated by genes in the H-2D region of the major histocompatibility complex (MHC). Recent studies have shown that elevated H-2Dd expression on the thymocyte cell surface of resistance mouse strains results from increased mRNA transcription and is correlated with elevated levels of a DNA-binding activity that recognizes a short DNA sequence 5' of the start of transcription for the H-2Dd gene. This binding activity has been termed H-2 binding factor 1 (H-2 BF1) and is found exclusively in the thymus. In an effort to examine the H-2 genes of RadLV-susceptible mice for the presence of the H-2 BF1 binding target, we have cloned class I genes from the highly susceptible B10.S mouse strain and have identified both the Ds and the Ks genes. The entire genomic sequence for the Ds gene has been determined and is reported here. In addition, the 5' regulatory region of the previously cloned Dq gene has been sequenced; mice of the Dq haplotype are also susceptible to RadLV-induced leukaemia. In this report, we show that the H-2 BF1 DNA binding sequence is present 5' of each of these three class I genes.

Electronic communication with patients. Evaluation of distance medicine technology

OBJECTIVE

To evaluate controlled evidence on the efficacy of distance medicine technologies in clinical practice and health care outcome.

DATA SOURCES

Systematic electronic database and manual searches (1966-1996) were conducted to identify clinical trial reports on distance medicine applications.

STUDY SELECTION

Three eligibility criteria were applied: prospective, contemporaneously controlled clinical trial with random assignment of the intervention; electronic distance technology application in the intervention group and no similar intervention in the control group; and measurement of the intervention effect on process or outcome of care.

DATA EXTRACTION

Data were abstracted by independent reviewers using a standardized abstraction form and the quality of methodology was scored. Distance technology applications were described in 6 categories: computerized communication, telephone follow-up and counseling, telephone reminders, interactive telephone systems, after-hours telephone access, and telephone screening.

DATA SYNTHESIS

Of 80 eligible clinical trials, 61 (76%) analyzed provider-initiated communication with patients and 50 (63%) reported positive outcome, improved performance, or significant benefits, including studies of computerized communication (7 of 7), telephone follow-up and counseling (20 of 37), telephone reminders (14 of 23), interactive telephone systems (5 of 6), telephone access (3 of 4), and telephone screening (1 of 3). Significantly improved outcomes were demonstrated in studies of preventive care, management of osteoarthritis, cardiac rehabilitation, and diabetes care.

CONCLUSIONS

Distance medicine technology enables greater continuity of care by improving access and supporting the coordination of activities by a clinician. The benefits of distance technologies in facilitating communication between clinicians and patients indicate that application of telemedicine should not be limited to physician-to-physician communication.

Activation transcription factor 1 involvement in the regulation of murine H-2Dd expression

Resistance to radiation leukemia virus-induced leukemia is correlated with an increase in H-2D expression on the thymocyte surface. Recently, it has been shown that elevated H-2Dd expression on the infected thymocyte is a result of elevated mRNA transcription and that the transcriptional increase is correlated with elevated levels of a DNA binding activity, H-2 binding factor 1 (H-2 BF1), which recognizes the 5'-flanking sequences (5'-TGACGCG-3') of the H-2Dd gene. This target for transcription factor binding has been found to be identical in the 5'-regulatory region of 12 rodent class I genes, nine of which have been shown to be functional genes. Furthermore, this cis-element is found 5' of 20 primate class I genes (15 human genes), seven of which are known to be functional. Here, we demonstrate that activation transcription factor 1 (ATF-1) is one component of H-2 BF1. In addition, the levels of ATF-1 mRNA in uninfected and radiation leukemia virus-infected thymocytes parallel those of H-2Dd mRNA, and therefore, it is suggested that ATF-1 up-regulates the transcription of the H-2Dd gene after radiation leukemia virus infection of thymocytes. Transfection experiments also demonstrate that ATF-1 activates a reporter plasmid that contains the H-2 BF1 motif, but not a reporter lacking this motif. This is the first demonstration of the interaction of ATF-1 with 5'-regulatory sequences of major histocompatibility complex class I genes.

Habituation and iterative enhancement of multiple components of the Tritonia swim response

To understand the relationship between memory storage sites in the brain and learned changes in behavior, the learned behavior must be characterized. However, even simple types of learning may be quite complex. Repeated elicitation of the Tritonia swim produced multiple changes in the response. Several types of acquisition curves were observed in a single experiment depending on the response component measured. Habituation (response decrement) and iterative enhancement (response facilitation) occurred simultaneously in different swim components. The acquisition curve for one component could be modulated by stimulus strength. Because the Tritonia swim neural network is well studied, it may be possible to causally relate experience-dependent behavioral changes to the underlying memory trace in this marine mollusk.

Effect of physician profiling on utilization. Meta-analysis of randomized clinical trials

OBJECTIVES

An American Medical Association survey reported that more than half of physicians are subjects of either clinical or economic profiling. This multilevel meta-analysis was designed to assess the clinical effect of peer-comparison feedback intervention (profiles) in changing practice patterns.

METHODS

Systematic computerized and manual searches were combined to retrieve articles on randomized controlled clinical trials testing profiling reports. Eligible studies were randomized, controlled clinical trials that tested peer-comparison feedback intervention and measured utilization of clinical procedures. To use all available information, data were abstracted and analyzed on three levels: (1) direction of effects, (2) p value fro the statistical comparison, and (3) odds ratio (OR).

MAIN RESULTS

In the 12 eligible trials, 553 physicians were profiled. The test result was p < .05 for the vote-counting sign test of 12 studies (level 1) and p < .05 for the z-transformation test of 8 studies (level 2). There were 5 trials included in the OR analysis (level 3). The primary effect variable in two of the 5 trials had a nonsignificant OR. However, the overall OR calculated by the Mantel-Haenszel method was significant (1.091, confidence interval: 1.045 to 1.136).

CONCLUSIONS

Profiling has a statistically significant, but minimal effect on the utilization of clinical procedures. The results of this study indicate a need for controlled clinical evaluations before subjecting large numbers of physicians to utilization management interventions.

Habituation and iterative enhancement of multiple components of the Tritonia swim response

To understand the relationship between memory storage sites in the brain and learned changes in behavior, the learned behavior must be characterized. However, even simple types of learning may be quite complex. Repeated elicitation of the Tritonia swim produced multiple changes in the response. Several types of acquisition curves were observed in a single experiment depending on the response component measured. Habituation (response decrement) and iterative enhancement (response facilitation) occurred simultaneously in different swim components. The acquisition curve for one component could be modulated by stimulus strength. Because the Tritonia swim neural network is well studied, it may be possible to causally relate experience-dependent behavioral changes to the underlying memory trace in this marine mollusk.

Infection of human cells by murine ecotropic viruses: retroviral vectors carrying the hygromycin resistance-encoding gene

The construction of a new retroviral vector, pSKV, is described. This vector carries two unique cloning sites, located between two Moloney leukemia virus-derived LTR, into which genes of interest may be introduced. The gene encoding hygromycin resistance (HyR) was subsequently introduced into one of the two sites, producing a second vector (pSKV/HyR) containing a unique SfiI site for the introduction of cDNA clones under the control of the cytomegalovirus (CMV) promoter (P-CMV). The cDNA (mH13), encoding a protein that has been shown to serve as a murine ecotropic retroviral receptor in transient assays, was cloned into the SfiI site (pSKV/HyR/mH13). Both constructs can be packaged into retroviral particles following transfection into an appropriate packaging cell line. Stable transfectants of the human glioblastoma cell line (U118MG) carrying each of these two constructs were generated by transfection and subsequent Hy selection. Clones expressing both the selectable marker and the mH13 gene, but not those expressing only the selectable marker, are shown to be susceptible to infection with murine ecotropic retroviral particles. These cells (HyR and mH13 positive) were then exposed to CRE/Xtk culture supernatant, a packaging cell line producing ecotropic retroviral particles carrying the HSV-TK (Herpes simplex virus-thymidine kinase) and neoR (neomycin-resistance) genes. Selection was in the presence of G418. In vitro growth of the U118MG/HyR/mH13/TK cells, but not that of the U118MG/HyR/mH13 cells, was inhibited by ganciclovir (GCV), indicating the successful transfer of HSV-TK by infection of human cells with murine retroviruses via the mH13 product.

Modelling human memory: connectionism and convolution

The mathematical operation of convolution is used as an associative mechanism by several recent influential models of human memory. Convolution can be used to associate two vectors (representing items to be remembered) into a memory trace vector in one operation. An approximation to either of the input vectors can then be retrieved, using the other vector as a probe. Recent convolution-based memory models have accounted for a wide range of data. Connectionist models may have greater potential for providing developmental accounts, but the architectures that have been most widely used to account for developmental phenomena cannot perform one-trial learning and this has limited their use as models of human memory. We show that a connectionist-like architecture can learn, using a gradient-descent algorithm, to perform single-trial learning in a similar manner to convolution. The solution that the network finds leads to less variable retrieval than does convolution. Furthermore, the network can learn to carry out the convolution operation itself. This provides a link between connectionist and convolution approaches, and a basis for models with many of the attractions of both connectionist and convolution approaches.

The clinical value of computerized information services. A review of 98 randomized clinical trials

OBJECTIVE

To review all randomized clinical trials addressing the efficacy of clinical information systems and to determine the clinical settings, types of interventions, and effects studied.

DATA SOURCES

Extensive and systematic MEDLINE searches were conducted using a combination of medical subject headings (MeSH) and textword terms to collect trial reports. Manual searches of books and monographs as well as informal contacts were also used.

STUDY SELECTION

The eligibility criteria were (1) randomized controlled clinical trial, (2) computerized information intervention in the study group, and (3) effect measured on the process or outcome of care.

DATA EXTRACTION

Two research assistants independently abstracted from the selected reports the following structured information: trial sites, computerized interventions, effect variables, and outcomes. Three investigators evaluated the combined list of trial features for setting, intervention, and effect. The statistical analysis included an evaluation of agreement in developing classifications and an analysis of the ratio of positive trial outcomes.

DATA SYNTHESIS

Most information services were tested in outpatient care (82%), particularly in primary care (66%). The information intervention targeted the provider in 64% of the trials. The effect was primarily measured for the process of care (76%). Provider prompt/reminder, computer-assisted treatment planner, interactive patient education/therapy, and patient prompt/reminder were significantly successful interventions (sign test, P < .05).

CONCLUSIONS

Randomized clinical trials confirm that four generic information interventions are active ingredients of computer systems and can make a significant difference in family medicine (physician and patient reminders, treatment planner, and patient education). To manage care and improve quality, primary care computer systems should incorporate these effective information services.

New curriculum challenges in health administration education

Tradition curriculum content and structure in health administration education programs have focused on the managing of individual institutions, the business function, and the hospital as the center of the health system. These assumptions are challenged by managed care, with its orientation on integrating the business and clinical functions; prevention and primary care; and the provision of services through integrated service networks. The design and process of curriculum change in response to new organization structures and management challenges are explored.

Parametric features of habituation of swim cycle number in the marine mollusc tritonia diomedea

When repeatedly elicited, the oscillatory escape swim of the marine mollusc Tritonia diomedea undergoes habituation of the number of cycles per swim. Because the neural circuit for this behavior is reasonably well understood, a cellular analyses of habituation in Tritonia is feasible. Since such a study must ultimately relate cellular correlates to behavioral modifications, we have sought to increase our understanding of the parametric features of cycle number habituation in Tritonia. Habituation was compared when using different intertrial intervals, repeated training sessions, and different stimulus locations. Stimulus site generalization of habituation was demonstrated, suggesting that at least one site of plasticity underlying habitation is located postsynaptic to the sensory neurons for the response. Dishabituation from an above-zero baseline response level was not obtained. An isolated brain preparation was tested as a potential simplified system for cellular studies of habituation mechanisms. Repeated stimulation of a nerve containing sensory afferent processes resulted in a progressive reduction of swim motor program cycle number, with a rate similar to that seen in the behavior. Together, these findings: (1) establish a set of parametric features of cycle numbers habituation to be explained by physiological studies; (2) suggest that at least one circuit modification underlying the habituation is located among the circuit interneurons; and (3) indicate that the isolated brain preparation may serve as a useful neural analogue for studies of the cellular mechanisms of cycle number habituation in Tritonia.

New disability rights require organizational change and leadership

On July 26, 1990, President George Bush signed the Americans with Disabilities Act (ADA). This legislation has been referred to as the "emancipation proclamation for the disabled" because of its great importance to persons with disabilities in our society. The ADA will not be implemented in a substantive way, however, without truly transformative leadership. Because of their traditional role in this area, health care professionals are in a unique position to provide that leadership and to help effect the necessary organizational changes. Passage of the ADA is only the beginning; the true test of its success will be in its implementation.

Cell targeting for gene delivery: use of fusion protein containing the modified human receptor for ecotropic murine leukemia virus

We have previously cloned a human gene (H13) homologous to the murine ecotropic retrovirus (E-MuLV) receptor, which, however, does not confer susceptibility to E-MuLV infection. The extracellular domain 3 (ECD3) of H13 contains amino acid residues critical for E-MuLV binding in that the modified H13 gene (mH13), substituted with amino acids from the actual receptor, has the ability to bind E-MuLV. Here we have expressed a fusion protein consisting of mH13/ECD3 and transforming growth factor-alpha in Escherichia coli and demonstrated its binding activity to both ecotropic AKR virus and the epidermal growth factor receptor expressed on the cell surface. Fusion proteins of mH13/ECD3 and ligands to cell surface molecules might be useful for specific cell targeting in E-MuLV-based gene delivery systems.

Apoptosis of neurons in the vestibular nuclei of adult mice results from prolonged change in the external environment

Pharmacological manipulations which result in abnormal levels of excitatory amino acid (EAA) mediated neurotransmission can result in neuronal apoptosis. We accordingly hypothesised that manipulations of the external environment which induce prolonged EAA-mediated transmission in sensory neurons may also induce apoptosis. This hypothesis was tested by placing groups of adult mice, housed in their home cage, on a turntable which slowly rotated (0.8 rev./min). This non-invasive manipulation will have resulted in abnormal discharge patterns in the vestibular nuclei. Significantly greater levels of neuronal apoptosis were seen in the vestibular complex after rotation for 48 h compared with non-rotated controls. This finding was also predicted independently from a computational approach.

Methods of randomized controlled clinical trials in health services research

The randomized controlled clinical trial is an increasingly used method in health services research. Analysis of methodology is needed to accelerate practical implementation of trial results, select trials for meta-analysis, and improve trial quality in health services research. The objectives of this study are to explore the methodology of health services research trials, create and validate a streamlined quality evaluation tool, and identify frequent quality defects and confounding effects on quality. The authors developed a quality questionnaire that contained 20 evaluation criteria for health services research trials. One hundred one trials from the Columbia Registry of Controlled Clinical Trials were evaluated using the new quality tool. The overall agreement between independent reviewers, Cohen's kappa, was 0.94 (+/- 0.01). Of a possible score of 100, the trials received an average score of 54.8 (+/- 12.5). Five evaluation criteria indicated significant quality deficiencies (sample size, description of case selection, data on possible adverse effects, analysis of secondary effect variables, and retrospective analysis). The quality of study characteristics was significantly weaker than the quality of reporting characteristics (P < 0.001). The total average scores of Medline-indexed journals were better than the non-Medline-indexed journals (P < 0.001). There was a positive correlation between the overall quality and year of publication (R = 0.21, P < 0.05). The authors conclude that the new quality evaluation tool leads to replicable results and there is an urgent need to improve several study characteristics of clinical trials. In comparison to drug trials, site selection, randomization, and blinding often require different approaches in health services research.

Mathematical and connectionist models of human memory: a comparison

Recent convolution-based models of human memory (e.g. Lewandowsky & Murdock, 1989), have accounted for a wide range of data. However such models require the relevant mathematical operations to be provided to the network. Connectionist models, in contrast, have generally addressed different data, and not all architectures are appropriate for modelling single-trial learning. Furthermore, they tend to exhibit catastrophic interference in multiple list learning. In this paper we compare the ability of convolution-based models and DARNET (Developmental Associative Recall NETwork), to account for human memory data. DARNET is a connectionist approach to human memory in which the system gradually learns to associate vectors, in one trial, into a memory trace vector. Either of the vectors can than be retrieved. It is shown that the new associative mechanism can be used to account for a wide range of relevant experimental data as successfully as can convolution-based models with the same higher-level architectures. Limitations of the models are also addressed.

Understanding barriers to basing nursing practice upon research: a communication model approach

It is suggested in this paper that the potential barriers to facilitating research-based practice can be comprehensively understood using a communication model approach. Barriers can thus be explored related to the research findings' message, to the ways in which this message is encoded and transmitted by nurse researchers and to the ways in which the message is received, interpreted and accepted by nurse practitioners. The role of research-based practice as a communication feedback mechanism is also described. The paper concludes by proposing that strategies for increasing the research base of our profession are more likely to be successful if all of these potential barriers are taken into account.

Amalgamating colleges of nursing and midwifery: staff perceptions of issues and their importance

This research aimed to identify the issues considered important by primary stakeholders in the amalgamation of two college of nursing and midwifery and to explore differences in issue perceptions between stakeholder groups. The purpose was to provide data which would facilitate the amalgamation process. The research was carried out in two stages, the first of which was issue identification, whereby a qualitative approach to exploration was employed. This was followed by the stage of issue validation, carried out using a quantitatively oriented questionnaire. Stage 1 identified 38 specific issues relating to 13 separate issue topics. These could be further categorized into structural, teaching/learning/curricular and affective themes. Stage 2 demonstrated that these 38 issues were all considered important by all staff groups, but highlighted important, significant differences in the perceptions of how well these issues were perceived to be addressed.

Effects of word frequency and age of acquisition on short-term memory span

The experiments reported examine the effects of two highly related variables, word frequency and age of acquisition, on short-term memory span. Short-term memory span and speech rate were measured for sets of words which independently manipulated frequency and age of acquisition. It was found that frequency had a considerable effect on short-term memory span, which was not mediated by speech rate differences--although frequency did affect speech rate in one experiment. For age of acquisition, this situation was reversed; there was a small but significant effect of age of acquisition on speech rate, but no effect on memory span. This occurred despite results confirming that the stimuli used in the experiments produce an effect of age of acquisition on word naming. The results are discussed in terms of a two-component view of performance on short-term memory tasks.

Enrolled nurses: where do they go from here?

This paper reviews the circumstances which have affected the development of and change in the role and status of enrolled nurses. Factors which have contributed to the drive to provide conversion pathways for enrolled nurses are examined and current opportunities are described. The paper highlights prospects for newly converted enrolled nurses and for those who do not wish to convert and argues that both groups continue to have educational and professional status needs which must be recognized and taken account of.

Cell suicide in the developing nervous system: a functional neural network model

A computational model of programmed cell death (PCD) in the nervous system is described. A neurobiologically realisable method for identifying and removing the least useful cells from a network is developed, and it is shown by simulation that an artificial neural network can solve difficult problems efficiently if it is given more neurons initially than it needs subsequently. The least useful neurons die off gradually after learning is complete, and the learned solution can then be maintained with a smaller number of units than were needed for initial learning. The research suggests a functional role for PCD, and how self-limiting PCD could be achieved in real neural systems.