Search PubMed⌕ Search
Home/Archive/PubMed

PubMed · 16360342

Terminology model discovery using natural language processing and visualization techniques.

Li Zhou·Ying Tao·James J Cimino·Elizabeth S Chen·Hongfang Liu·Yves A Lussier·George Hripcsak·Carol Friedman
View original source ↗Export metadata ↓

Abstract

Medical terminologies are important for unambiguous encoding and exchange of clinical information. The traditional manual method of developing terminology models is time-consuming and limited in the number of phrases that a human developer can examine. In this paper, we present an automated method for developing medical terminology models based on natural language processing (NLP) and information visualization techniques. Surgical pathology reports were selected as the testing corpus for developing a pathology procedure terminology model. The use of a general NLP processor for the medical domain, MedLEE, provides an automated method for acquiring semantic structures from a free text corpus and sheds light on a new high-throughput method of medical terminology model development. The use of an information visualization technique supports the summarization and visualization of the large quantity of semantic structures generated from medical documents. We believe that a general method based on NLP and information visualization will facilitate the modeling of medical terminologies.

Explore related subjects

AutomationBiotechnologyComputational BiologyDatabase Management SystemsHumansInformation Storage and RetrievalMedical RecordsNatural Language ProcessingProgramming LanguagesTerminology as TopicUser-Computer InterfaceVocabulary, ControlledJournal ArticleResearch Support, N.I.H., Extramural

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Li Zhou, Ying Tao, James J Cimino, Elizabeth S Chen, Hongfang Liu, Yves A Lussier, George Hripcsak, Carol Friedman. 2006. Terminology model discovery using natural language processing and visualization techniques.. https://doi.org/10.1016/j.jbi.2005.10.006

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2006

On-line estimation of dissolved methane concentration during methanotrophic fermentations.

Knowledge of the aqueous phase methane concentration is critical to understanding and controlling process kinetics in methanotrophic bioreactors. Unfortunately since no dissolved methane probe is commercially available, this data must be obtained off-line by the time-consuming gas-liquid partition method. In this study we demonstrate how knowledge of the reactor's k(L)a for oxygen combined with gas phase methane analysis can be used to continuously estimate the aqueous phase concentration of dissolved methane. The on-line estimation was verified in two reactor systems with greatly different values of k (L)a. In both systems the measured and calculated dissolved methane concentrations were in good agreement although dissolved methane was underestimated in both cases. The utility of this methodology was demonstrated by revealing a possible metabolic bottleneck in the model system.

Automation↗
PubMed2006

Purification of an elastin-like fusion protein by microfiltration.

This article describes a simple and potentially scalable microfiltration method for purification of recombinant proteins. This method is based on the fact that when an elastin-like polypeptide (ELP) is fused to a target protein, the inverse phase transition behavior of the ELP tag is imparted to the fusion protein. Triggering the phase transition of a solution of the ELP fusion protein by an increase in temperature, or isothermally by an increase in salt concentration, results in the formation of micron-sized aggregates of the ELP fusion protein. In this article, it is shown that these aggregates are efficiently retained by a microfiltration membrane, while contaminating E. coli proteins passed through the membrane upon washing. Upon reversing the phase transition by flow of Milli-Q water, soluble, pure, and functionally active protein is eluted from the membrane. Proof-of principle of this approach was demonstrated by purifying a fusion of thioredoxin with ELP (Trx-ELP) with greater than 95% recovery of protein and with greater than 95% purity (as estimated from SDS-PAGE gels). The simplicity of this method is demonstrated for laboratory scale purification by purifying Trx-ELP from cell lysate using a syringe and a disposable microfiltration cartridge. The potential scalability of this purification as an automated, continuous industrial-scale process is also demonstrated using a continuous stirred cell equipped with a microfiltration membrane.

Automation↗
PubMed2006

Measurement of human urinary organophosphate pesticide metabolites by automated solid-phase extraction derivation and gas chromatography-tandem mass spectromy.

Organophosphorus (OP) pesticides are among the most widely used pesticides in the United States. Human exposure to these pesticides may occur from their use on crops in agriculture and for pest control in residential settings. Most of the OP pesticides used in the United States are metabolized to up to three of six common urinary dialkyl phosphate metabolites. Quantification of these metabolites provides information on cumulative exposure to most OP pesticides. To accurately quantify OP pesticide metabolites in human urine, we developed a simple, highly sensitive, analytic method involving automated solid-phase extraction (SPE) of human urine, followed by post-extraction derivatization of the organophosphorus metabolites with 1-chloro-3-iodopropane, and analysis by isotope dilution gas-chromatography-tandem mass spectrometry. The styrene-divinyl benzene polymer-based SPE cartridges yielded good SPE recoveries of the metabolites because of their enhanced non-polar interactions. This method is less labor-intensive, more time-efficient, and reproducible than previously reported methods. Automation of the SPE allowed unattended extraction of urine samples, and hence, increased the sample throughput and reduced the inter- and intra-day variations. The method limits of detection were excellent for all analytes ranging from 50 pg/ml to 170 pg/ml. Relative standard deviations ranged from 2% to 12%.

Automation↗