一体化医学语言系统中的循环等级关系

主要讨论了一体化医学语言系统(UMLS)超级叙词表中的循环等级关系和产生原因,及其识别和消除方法.     

Word-specific cortical activity as revealed by the mismatch negativity

Neurophysiological brain activity evoked by individual spoken words and pseudowords was recorded and the mismatch negativity (MMN), an automatic index of experience-dependent auditory memory traces, was calculated. Consistent with earlier reported results, the MMN response to word-final syllables was enhanced compared with that elicited by the same syllables placed in a pseudoword context. Here we now demonstrate that the enhancement of the MMN elicited by two individual words showed different scalp topographies. The early word-specific brain activity is consistent with the assumption that the memory traces activated by individual words are carried by large neuronal ensembles that differ in their distributions over the cortex. Current source estimates localized the between-word differences in the right hemisphere and in parieto-occipital left-hemispheric areas. The differential brain responses to individual words appeared as early as ∼100 ms after the recognition points of the words, suggesting that their specific memory traces become active almost immediately after the information in the acoustic input is sufficient for word identification.     

Oral language impairments in developmental disorders characterized by language strengths: A comparison of Asperger syndrome and nonverbal learning disabilities

Asperger syndrome (AS) and nonverbal learning disabilities (NLD) are developmental disorders in which linguistic ability is reported to be stronger than in disorders from which they must be distinguished for diagnosis. Children and adults with AS and NLD share pragmatic weaknesses, atypical social behaviours, and some cognitive features. To date, potential similarities between these disorders in oral language have not been directly examined in the literature. A review of the available research suggests that adequate structural language is another area of similarity for AS and NLD. However, systematic investigations of phonology, morphology, or syntax were not found; thus, the evidence for largely intact structural language in these disorders is indirect. The review also pointed to a common semantic profile across both disorders, characterized by strong vocabulary breadth in the face of limited depth and organization. These higher-order problems with semantics are proposed to be consistent with theoretical accounts of poor integrative abilities in AS and NLD, and to contribute to the well-documented pragmatic difficulties in these disorders.     

Utility-preserving privacy protection of textual healthcare documents

The adoption of ITs by medical organisations makes possible the compilation of large amounts of healthcare data, which are quite often needed to be released to third parties for research or business purposes. Many of this data are of sensitive nature, because they may include patient-related documents such as electronic healthcare records. In order to protect the privacy of individuals, several legislations on healthcare data management, which state the kind of information that should be protected, have been defined. Traditionally, to meet with current legislations, a manual redaction process is applied to patient-related documents in order to remove or black-out sensitive terms. This process is costly and time-consuming and has the undesired side effect of severely reducing the utility of the released content. Automatic methods available in the literature usually propose ad-hoc solutions that are limited to protect specific types of structured information (e.g. e-mail addresses, social security numbers, etc.); as a result, they are hardly applicable to the sensitive entities stated in current regulations that do not present those structural regularities (e.g. diseases, symptoms, treatments, etc.). To tackle these limitations, in this paper we propose an automatic sanitisation method for textual medical documents (e.g. electronic healthcare records) that is able to protect, regardless of their structure, sensitive entities (e.g. diseases) and also those semantically related terms (e.g. symptoms) that may disclose the former ones. Contrary to redaction schemes based on term removal, our approach improves the utility of the protected output by replacing sensitive terms with appropriate generalisations retrieved from several medical and general-purpose knowledge bases. Experiments conducted on highly sensitive documents and in coherency with current regulations on healthcare data privacy show promising results in terms of the practical privacy and utility of the protected output.     

Review article MURDOCH,B.E. (ed.), 1998, Dysarthria: A Physiological Approach to Assessment and Treatment (Cheltenham: Stanley Thornes,Ltd).[Pp. 431+xii.] ISBN 0-7487-3311-6

The naming and drawing responses of a child with specific language impairment (age 5.5 years) were used to test the hypothesis that deficient storage in the mental lexicon plays a role in the naming problems associated with SLI. In confrontation- and repeated naming, the child demonstrated frequent semantic substitutions and occasional phonologic substitutions. Stochastic modelling of his repeated naming revealed storage deficits to be a source of these errors. Comparative picture naming, picture drawing allowed exploration of this storage deficit and revealed that, for some semantic naming errors, sparse semantic representations were clearly at fault but for others, sparse phonological input and output representations played a role. Phonological naming errors, in contrast, were typically associated with strong semantic representations. Clinical, theoretical, and methodological contributions of this cognitive neuropsychological case study are discussed.     

Contribution of language studies to the understanding of cognitive impairment and its progression over time in Parkinson’s disease

Parkinson’s disease is a frequent neurodegenerative disease that is mostly known for its motor symptoms. However, cognitive impairment is now recognised as an important part of the disease. Studies of cognitive impairment in Parkinson’s disease reveal considerable heterogeneity in terms of which cognitive domains are impaired, and of how cognitive impairment progresses over time. In parallel, a growing body of research reports language difficulties in Parkinson’s disease, more specifically in the domains of sentence processing and lexical-semantic processing. In this review, the performance of patients with Parkinson’s disease in these domains of language will be reviewed with a focus on the links that they have with the rest of cognition and on how they could contribute to the earlier and more precise characterization and prediction of cognitive impairment in Parkinson’s disease. More specifically, the potential for modulation of complexity and sensitivity of language tasks to mild deficits and difficulties that are predictive of further decline will be emphasized. Other motivations for studying language difficulties in this disease will also be discussed.     

The spatial and temporal meanings of English prepositions can be independently impaired

English uses the same prepositions to describe both spatial and temporal relationships (e.g., at the corner, at 1:30), and other languages worldwide exhibit similar patterns. These space-time parallelisms have been explained by the Metaphoric Mapping Theory, which maintains that humans have a cognitive predisposition to structure temporal concepts in terms of spatial schemas through the application of a TIME IS SPACE metaphor. Evidence comes from (among other sources) historical investigations showing that languages consistently develop in such a way that expressions that originally have only spatial meanings are gradually extended to take on analogous temporal meanings. It is not clear, however, if the metaphor actively influences the way that modern adults process prepositional meanings during language use. To explore this question, a series of experiments was conducted with four brain-damaged subjects with left perisylvian lesions. Two subjects exhibited the following dissociation: they failed a test that assesses knowledge of the spatial meanings of prepositions, but passed a test that assesses knowledge of the corresponding temporal meanings of the same prepositions. This result suggests that understanding the temporal meanings of prepositions does not necessarily require establishing structural alignments with their spatial correlates. Two other subjects exhibited the opposite dissociation: they performed better on the spatial test than on the temporal test. Overall, these findings support the view that although the spatial and temporal meanings of prepositions are historically linked by virtue of the TIME IS SPACE metaphor, they can be (and may normally be) represented and processed independently of each other in the brains of modern adults.     

Category specific dysnomia after thalamic infarction: a case-control study

Category specific naming impairment was described mainly after cortical lesions. It is thought to result from a lesion in a specific network, reflecting the organization of our semantic knowledge. The deficit usually involves multiple semantic categories whose profile of naming deficit generally obeys the animate/inanimate dichotomy. Thalamic lesions cause general semantic naming deficit, and only rarely a category specific semantic deficit for very limited and highly specific categories. We performed a case-control study on a 56-year-old right-handed man who presented with language impairment following a left anterior thalamic infarction. His naming ability and semantic knowledge were evaluated in the visual, tactile and auditory modalities for stimuli from 11 different categories, and compared to that of five controls. In naming to visual stimuli the patient performed poorly (error rate>50%) in four categories: vegetables, toys, animals and body parts (average 70.31+/-15%). In each category there was a different dominating error type. He performed better in the other seven categories (tools, clothes, transportation, fruits, electric, furniture, kitchen utensils), averaging 14.28+/-9% errors. Further analysis revealed a dichotomy between naming in animate and inanimate categories in the visual and tactile modalities but not in response to auditory stimuli. Thus, a unique category specific profile of response and naming errors to visual and tactile, but not auditory stimuli was found after a left anterior thalamic infarction. This might reflect the role of the thalamus not only as a relay station but further as a central integrator of different stages of perceptual and semantic processing.     

A meta-analytic review of category naming in Alzheimer's disease

Patients with Alzheimer's disease (AD) experience word-finding difficulties that become increasingly pronounced as pathological changes accrue in the brain. One question that has received increasing attention over the last two decades concerns whether the anomia in AD is category-specific, i.e. differentially affects the ability to name living things (LT) and non-living things (NLT). The current meta-analysis systematically reviewed the effect sizes for naming pictures of LT and NLT in comparisons of AD patients and healthy controls in 21 studies with over 1000 participants (557 patients and 509 healthy controls). A random effects model analysis revealed no significant difference in the large weighted effect sizes for naming pictures of LT and NLT (d=1.76 and 1.49, respectively). Moderator variable analyses revealed a significant impact of stimulus colour on the effect size for LT, indicating that using colour stimuli significantly increases the impairment of naming LT in AD patients. Additionally, we found that LT and the NLT effect sizes were larger for samples with proportionally more female patients; smaller samples produced larger LT effect sizes. In contrast, effect sizes were not significantly related to dementia severity, patient age, the number of stimuli, years of education, or the number of matching variables controlled.     

Rightward bisection errors for letter lines: the role of semantic information

When bisecting words in their middle, people reveal leftward bisection errors. This tendency might emerge from an attentional bias towards the beginning of the word. However, when longer meaningless letter strings are presented, people reveal a rightward bisection bias. To test the role of semantic information on leftward or rightward bisection biases, we tested letter line bisection performance in healthy right-handed students in four independent experiments. A third of the letter lines contained an embedded four-letter word to the left of true centre, another third contained an embedded four-letter word to the right of true centre, while the remaining lines contained no words. Half of these words were emotional words, the other half were neutral words. Results across experiments revealed a stronger rightward bisection bias: (i) for letter lines containing emotional as compared to neutral words, (ii) for letter lines containing words in the left as compared to right half of the lines, and (iii) for those experiments in which the spatial position of letter lines remained within a narrow body-centred space. Findings from this study suggest that letter line bisection performance might be only minimally determined by visuo-spatial attention. Rather, letter line perception might activate the left hemisphere more than the right hemisphere, shifting the subjective midpoint to the right of true centre. Leftward bisection biases for words only, as had been described in the literature, may thus have resulted from automated reading strategies rather than from attentional biases towards the left hemispace.