Modeling medical content for automated summarization.

David B Johnson; Qinghua Zou; John D Dionisio; Victor Zhenyu Liu; Wesley W Chu

doi:10.1111/j.1749-6632.2002.tb04901.x

Home/Archive/PubMed

PubMed · 12594094

Modeling medical content for automated summarization.

David B Johnson·Qinghua Zou·John D Dionisio·Victor Zhenyu Liu·Wesley W Chu

View original source ↗Export metadata ↓

Abstract

Medical information is available from a variety of new online resources. Given the number and diversity of sources, methods must be found that will enable users to quickly assimilate and determine the content of a document. Summarization is one such tool that can help users to quickly determine the main points of a document. Previous methods to automatically summarize text documents typically do not attempt to infer or define the content of a document. Rather these systems rely on secondary features or clues that may point to content. This paper describes text summarization techniques that enable users to focus on the key content of a document. The techniques presented here analyze groups of similar documents in order to form a content model. The content model is used to select sentences forming the summary. The technique does not require additional knowledge sources; thus the method should be applicable to any set of text documents.

Explore related subjects

Automation Cluster Analysis Cues Documentation Humans Information Storage and Retrieval Medical Informatics Journal Article

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

David B Johnson, Qinghua Zou, John D Dionisio, Victor Zhenyu Liu, Wesley W Chu. 2002. Modeling medical content for automated summarization.. https://doi.org/10.1111/j.1749-6632.2002.tb04901.x

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

KEEP EXPLORING

Related citations

PubMed2007

Characteristics of the m2000 automated sample preparation and multiplex real-time PCR system for detection of Chlamydia trachomatis and Neisseria gonorrhoeae.

We evaluated a new real-time PCR-based prototype assay for the detection of Chlamydia trachomatis and Neisseria gonorrhoeae developed by Abbott Molecular Inc. This assay is designed to be performed on an Abbott m2000 real-time instrument system, which consists of an m2000sp instrument for sample preparation and an m2000rt instrument for real-time PCR amplification and detection. The limit of detection of this prototype assay was determined to be 20 copies of target DNA for both C. trachomatis and N. gonorrhoeae, using serially diluted linearized plasmids. No cross-reactivity could be detected when 55 nongonococcal Neisseria isolates and 3 non-C. trachomatis Chlamydia isolates were tested at 1 million genome equivalents per reaction. Concordance with the Roche Amplicor, BDProbeTec ET, and Gen-Probe APTIMA Combo 2 tests was assessed using unlinked/deidentified surplus clinical specimens previously analyzed with these tests. For C. trachomatis, concordance for positive results ranged from 93.7% to 100%, while concordance for negative results ranged from 98.2% to 100%. For N. gonorrhoeae, concordance for positive and negative results ranged from 91.4% to 100% and 99.3% to 100%, respectively. A workflow analysis of the prototype assay was conducted to obtain information on throughput under laboratory conditions. At 48 samples/run, the time to first result for both C. trachomatis and N. gonorrhoeae was 4.5 h. A total of 135 patient specimens could be analyzed in 8.9 h, with 75 min of hands-on time. This study demonstrated the technical and clinical feasibility of the new Abbott real-time PCR C. trachomatis/N. gonorrhoeae assay.

Automation↗

PubMed2007

Pronase-immobilized enzyme reactor: an approach for automation in glycoprotein analysis by LC/LC-ESI/MSn.

An automated analytical approach is proposed for simultaneous characterization of glycan and peptide moieties in pronase-generated glycopeptides. The proposed method is based on the use of a new pronase-immobilized enzyme reactor for the on-line rapid digestion of the target glycoprotein. By coupling the bioreactor to a Hypercarb chromatographic trap column, on-line selective glycopeptide enrichment prior to normal-phase liquid chromatography-mass spectrometry was obtained. A detailed study was carried out for integration and automation of each phase of the proposed analytical procedure. On-line digestion allowed extensive cleavage of the model protein (ribonuclease B), yielding to glycopeptides with peptide moieties up to eight amino acids, carrying the Man5-Man9 N-glycans each, selectively resolved on an Amide-80 column. The use of a linear ion trap instrument resulted in efficient ion capture and led to MS3 acquisition times and spectra quality similar to those for MS2, allowing the unambiguous identification of glycan (MS2) and peptide (MS3) sequences. The proposed procedure reduces the glycoprotein analysis time from approximately 3 days, as in most of the traditional off-line methods, to approximately 1 h.

Automation↗

PubMed2007

An improved high-throughput liquid chromatographic/tandem mass spectrometric method for terbinafine quantification in human plasma, using automated liquid-liquid extraction based on 96-well format plates.

A fully automated high-throughput liquid chromatography/tandem mass spectrometry (LC-MS/MS) method was developed for terbinafine quantification in human plasma. The plasma samples were treated by liquid-liquid extraction (LLE) in 2.2 mL 96-deepwell plates. Terbinafine and the internal standard (IS) N-methyl-1-naphthalenemethylamine were extracted from human plasma by LLE, using a mixture of methyl t-butyl ether (MTBE)-hexane (70:30, v/v) as the organic solvent. All liquid transfer steps, including preparation of calibration standards and quality control samples, as well as the addition of the IS, were performed automatically by using robotic liquid handling workstations. After vortexing, centrifugation and freezing, the supernatant organic solvent was evaporated and reconstituted in a small volume of a reconstitution solution. Sample analysis was performed by reversed-phase LC-MS/MS, with positive ion electrospray ionization, using multiple reaction monitoring (MRM). The method had a very short sample preparation time and a chromatographic run time of 2.2 min. It was proved to have excellent sensitivity, specificity, accuracy as well as inter- and intraday precision for the quantification of terbinafine in human plasma. The calibration curve was linear for the range of concentrations 5.0-2000.0 ng/mL. The proposed method was applied to the rapid and reliable determination of terbinafine in a bioequivalence study after per os administration of 250 mg tablet formulations of terbinafine.

Automation↗