Search tips
Search criteria

Results 1-5 (5)

Clipboard (0)

Select a Filter Below

more »
Year of Publication
Document Types
1.  Automatic discourse connective detection in biomedical text 
Relation extraction in biomedical text mining systems has largely focused on identifying clause-level relations, but increasing sophistication demands the recognition of relations at discourse level. A first step in identifying discourse relations involves the detection of discourse connectives: words or phrases used in text to express discourse relations. In this study supervised machine-learning approaches were developed and evaluated for automatically identifying discourse connectives in biomedical text.
Materials and Methods
Two supervised machine-learning models (support vector machines and conditional random fields) were explored for identifying discourse connectives in biomedical literature. In-domain supervised machine-learning classifiers were trained on the Biomedical Discourse Relation Bank, an annotated corpus of discourse relations over 24 full-text biomedical articles (∼112 000 word tokens), a subset of the GENIA corpus. Novel domain adaptation techniques were also explored to leverage the larger open-domain Penn Discourse Treebank (∼1 million word tokens). The models were evaluated using the standard evaluation metrics of precision, recall and F1 scores.
Results and Conclusion
Supervised machine-learning approaches can automatically identify discourse connectives in biomedical text, and the novel domain adaptation techniques yielded the best performance: 0.761 F1 score. A demonstration version of the fully implemented classifier BioConn is available at:
PMCID: PMC3422833  PMID: 22744958
Analysis; automated learning; controlled terminologies and vocabularies; discovery; display; image representation; knowledge acquisition and knowledge management; knowledge bases; knowledge representations; machine learning; natural language processing; NLP; ontologies; processing; text and data mining methods
2.  An Investigation into the Feasibility of Spoken Clinical Question Answering 
Spoken question answering for clinical decision support is a potentially revolutionary technology for improving the efficiency and quality of health care delivery. This application involves many technologies currently being researched, including automatic speech recognition (ASR), information retrieval (IR), and summarization, all in the biomedical domain. In certain domains, the problem of spoken document retrieval has been declared solved because of the robustness of IR to ASR errors. This study investigates the extent to which spoken medical question answering benefits from that same robustness. We used the best results from previous speech recognition experiments as inputs to a clinical question answering system, and had physicians perform blind evaluations of results generated both by ASR transcripts of questions and gold standard transcripts of the same questions. Our results suggest that the medical domain differs enough from the open domain to require additional work in automatic speech recognition adapted for the biomedical domain.
PMCID: PMC3243288  PMID: 22195154
3.  “Racial and social class gradients in life expectancy in contemporary California” 
Social science & medicine (1982)  2010;70(9):1373-1380.
Life expectancy, or the estimated average age of death, is among the most basic measures of a population's health. However, monitoring differences in life expectancy among sociodemographically defined populations has been challenging, at least in the United States (US), because death certification does not include collection of markers of socioeconomic status (SES). In order to understand how SES and race/ethnicity independently and jointly affected overall health in a contemporary US population, we assigned a small area-based measure of SES to all 689,036 deaths occurring in California during a three-year period (1999-2001) overlapping the most recent US census. Residence at death was geocoded to the smallest census area available (block group) and assigned to a quintile of a multifactorial SES index. We constructed life tables using mortality rates calculated by age, sex, race/ethnicity and neighborhood SES quintile, and produced corresponding life expectancy estimates. We found a 19.6 (±0.6) year gap in life expectancy between the sociodemographic groups with the longest life expectancy (highest SES quintile of Asian females; 84.9 years) and the shortest (lowest SES quintile of African-American males; 65.3 years). A positive SES gradient in life expectancy was observed among whites and African-Americans but not Hispanics or Asians. Age-specific mortality disparities varied among groups. Race/ethnicity and neighborhood SES had substantial and independent influences on life expectancy, underscoring the importance of monitoring health outcomes simultaneously by these factors. African-American males living in the poorest 20% of California neighborhoods had life expectancy comparable to that reported for males living in developing countries. Neighborhood SES represents a readily available metric for ongoing surveillance of health disparities in the US.
PMCID: PMC2849870  PMID: 20171001
racial disparities; social class disparities; life expectancy; California; population-based; USA; socioeconomic status (SES)
4.  Who wins and who loses? Public transfer accounts for US generations born 1850 to 2090 
Public transfer programs in industrial nations are thought to benefit the elderly through pension and health care programs at the expense of the young and future generations. However, this intergenerational picture changes if public education is also considered as a transfer program. We calculate the net present value (NPV) of benefits received minus taxes paid for US generations born 1850 to 2090. Surprisingly, all generations 1950 to 2050 are net gainers, while many current elderly are losers. Windfall gains from starting Social Security and Medicare partially offset windfall losses from starting public education, roughly consistent with the Becker-Murphy theory.
PMCID: PMC2840408  PMID: 20300431
5.  Cholangiocyte proliferation and liver fibrosis 
Cholangiocyte proliferation is triggered during extrahepatic bile duct obstruction induced by bile duct ligation, which is a common in vivo model used for the study of cholangiocyte proliferation and liver fibrosis. The proliferative response of cholangiocytes during cholestasis is regulated by the complex interaction of several factors, including gastrointestinal hormones, neuroendocrine hormones and autocrine or paracrine signalling mechanisms. Activation of biliary proliferation (ductular reaction) is thought to have a key role in the initiation and progression of liver fibrosis. The first part of this review provides an overview of the primary functions of cholangiocytes in terms of secretin-stimulated bicarbonate secretion – a functional index of cholangiocyte growth. In the second section, we explore the important regulators, both inhibitory and stimulatory, that regulate the cholangiocyte proliferative response during cholestasis. We discuss the role of proliferating cholangiocytes in the induction of fibrosis either directly via epithelial mesenchymal transition or indirectly via the activation of other liver cell types. The possibility of targeting cholangiocyte proliferation as potential therapy for reducing and/or preventing liver fibrosis, and future avenues for research into how cholangiocytes participate in the process of liver fibrogenesis are described.
PMCID: PMC2675635  PMID: 19239726

Results 1-5 (5)