Biomedical subjects

G P Purcell

Publications and source records attributed to G P Purcell.

7 recordsLinked to original sources

PubMed1999

Parathyroid localization with high-resolution ultrasound and technetium Tc 99m sestamibi.

HYPOTHESIS: High-resolution ultrasound and technetium Tc 99m sestamibi scanning can be used for preoperative localization of abnormal parathyroid glands in patients with hyperparathyroidism. DESIGN: Ultrasound and sestamibi scanning were performed in patients undergoing neck exploration for hyperparathyroidism. If the 2 scans agreed in identifying a single adenoma, and surgery confirmed the location of a single adenoma and an ipsilateral normal gland, a unilateral exploration was performed. SETTING: University tertiary care center. PATIENTS: Sixty-one consecutive patients undergoing surgery for hyperparathyroidism from September 1, 1994, through September 30, 1997. INTERVENTIONS: High-resolution ultrasound was performed in 59 patients and sestamibi scanning in 58 patients; all patients underwent neck exploration by a single surgeon. MAIN OUTCOME MEASURES: The results of preoperative ultrasound and sestamibi scanning were compared with operative and histological findings. RESULTS: All patients were cured of hypercalcemia. Specificity of ultrasound and sestamibi scanning was 98% and 99%, respectively; however, their sensitivity was only 57% and 54%, respectively. Both imaging modalities had lower sensitivities in the setting of multigland disease. If both imaging studies were considered as a single test, sensitivity for imaging in patients with primary hyperparathyroidism reached 78%. Our localization protocol allowed a unilateral approach in 43% of patients (23 of 53). CONCLUSIONS: These results confirm the value of preoperative localization in patients with hyperparathyroidism. A unilateral approach can be used with a high degree of success in cases when ultrasound and sestamibi scanning agree in the identification of a single adenoma confirmed by surgical exploration with the identification of a normal ipsilateral gland.

Adenoma↗

PubMed1999

Task-specific journal extracts for using the medical literature.

Clinicians and researchers use the medical literature in a variety of ways. The overwhelming volume of clinical journals necessitates tools to help healthcare professionals identify and employ relevant information. The structured abstract can facilitate browsing articles, but may not contain appropriate types of information or sufficient detail for all uses of the medical literature. We have created customized views of journal articles that provide information for specific research or clinical tasks, such as evaluating the scientific validity of a clinical trial. These summaries are called extracts because we literally extract information of a particular type from the full text of an article. We employ a context-based indexing scheme, previously designed for improving precision in literature searches, to automatically generate extracts from clinical research articles. In this paper, we present an evaluation of the content and utility of these task-specific extracts. Our results provide preliminary evidence that such extracts contain information that is relevant to clinical and research tasks and may facilitate understanding and use of the medical literature.

Abstracting and Indexing↗

PubMed1998

Changes to manuscripts during the editorial process: characterizing the evolution of a clinical paper.

CONTEXT: Biomedical manuscripts undergo substantive change as a result of the peer review and editorial revision processes. OBJECTIVE: To characterize quantitatively problems in manuscripts identified during peer review and changes made to address these problems. DESIGN AND SETTING: Descriptive analysis of manuscripts submitted to and articles published by the Annals of Internal Medicine. A taxonomy of problems that occur in reporting clinical research was developed from analysis of changes made to 7 manuscripts between submission and publication (published October 15, 1996, and November 1, 1996). The taxonomy was used to characterize changes to 12 additional manuscripts (published January 15, 1997, to April 1, 1997). MAIN OUTCOME MEASURE: Types of problems necessitating changes to manuscripts during peer review and revision. RESULTS: Changes occurred because of 5 types of problems: too much information, too little information, inaccurate information, misplaced information, and structural problems. Changes most often occurred because information was missing or extraneous. The distribution of changes seemed to be influenced by the type of information involved (such as background or conclusions). CONCLUSION: The proposed framework may be useful for characterizing quantitatively the effects of peer review and for comparing those effects across editors, journals, and specialties.

Peer Review↗

PubMed1995

Contextual models of clinical publications for enhancing retrieval from full-text databases.

Conventional methods for retrieving information from the medical literature are imprecise and inefficient. Information retrieval systems employ unmanageable indexing vocabularies or use full-text representations that overwhelm the user with irrelevant information. This paper describes a document representation designed to improve the precision of searching in textual databases without significantly compromising recall. The representation augments simple text word representations with contextual models that reflect recurring semantic themes in clinical publications. Using this representation, a searcher may indicate both the terms of interest and the contexts in which they should occur. The contexts limit the potential interpretations of text words, and thus form the basis for more precise searching. In this paper, we discuss the shortcomings of traditional retrieval systems and describe our context-based representation. Improved retrieval performance with contextual models is illustrated by example, and a more extensive study is proposed. We present an evaluation of the contextual models as an indexing scheme, using a variation of the traditional inter-indexer consistency experiments, and we demonstrate that contextual indexing is reproducible by minimally trained physicians and medical students.

Abstracting and Indexing↗

PubMed1992

An extended SQL for temporal data management in clinical decision-support systems.

We are developing a database implementation to support temporal data management for the T-HELPER physician workstation, an advice system for protocol-based care of patients who have HIV disease. To understand the requirements for the temporal database, we have analyzed the types of temporal predicates found in clinical-trial protocols. We extend the standard relational data model in three ways to support these querying requirements. First, we incorporate timestamps into the two-dimensional relational table to store the temporal dimension of both instant- and interval-based data. Second, we develop a set of operations on timepoints and intervals to manipulate timestamped data. Third, we modify the relational query language SQL so that its underlying algebra supports the specified operations on timestamps in relational tables. We show that our temporal extension to SQL meets the temporal data-management needs of protocol-directed decision support.

Acquired Immunodeficiency Syndrome↗

PubMed1992

SCOUT: information retrieval from full-text medical literature.

SCOUT is an expert system designed to aid the practicing physician in searching a full-text database of biomedical literature for information in the domain of oncology. In the clinical setting, an information-retrieval system needs to provide only a few relevant documents that address a patient-specific query. SCOUT models the expertise of an oncologist and an information specialist in a probabilistic inference system; it applies expert heuristics both in selecting the best source of information for a particular question and in developing an effective, source-specific search statement that retrieves only a minimal number of pertinent documents.

Expert Systems↗

PubMed1989

Validation of diagnostic procedures on stratified populations: application on the quantification of thallium myocardial perfusion scintigraphy.

In this report we present a method for the quantitative description of the degree of deviation from the norm of 201thallium single photon emission tomographic (SPECT) data. Validation is obtained from the frequency of "positive" outcomes in subgroup of patients in whom the prevalence of coronary artery disease, for the group as a whole, is known, even if individual patient outcomes are not verified. This approach overcomes the bias associated with nonrandomized clinical studies, in which the likelihood that a more invasive but definitive procedure (coronary arteriogram) will be performed is influenced by the result of the outcome of the procedure under study.

Coronary Disease↗