Biomedical subjects

George Thoma

Publications and source records attributed to George Thoma.

4 recordsLinked to original sources

PubMed2005

Automatically identifying health outcome information in MEDLINE records.

OBJECTIVE: Understanding the effect of a given intervention on the patient's health outcome is one of the key elements in providing optimal patient care. This study presents a methodology for automatic identification of outcomes-related information in medical text and evaluates its potential in satisfying clinical information needs related to health care outcomes. DESIGN: An annotation scheme based on an evidence-based medicine model for critical appraisal of evidence was developed and used to annotate 633 MEDLINE citations. Textual, structural, and meta-information features essential to outcome identification were learned from the created collection and used to develop an automatic system. Accuracy of automatic outcome identification was assessed in an intrinsic evaluation and in an extrinsic evaluation, in which ranking of MEDLINE search results obtained using PubMed Clinical Queries relied on identified outcome statements. MEASUREMENTS: The accuracy and positive predictive value of outcome identification were calculated. Effectiveness of the outcome-based ranking was measured using mean average precision and precision at rank 10. RESULTS: Automatic outcome identification achieved 88% to 93% accuracy. The positive predictive value of individual sentences identified as outcomes ranged from 30% to 37%. Outcome-based ranking improved retrieval accuracy, tripling mean average precision and achieving 389% improvement in precision at rank 10. CONCLUSION: Preliminary results in outcome-based document ranking show potential validity of the evidence-based medicine-model approach in timely delivery of information critical to clinical decision support at the point of service.

Artificial Intelligence↗

PubMed2005

The role of title, metadata and abstract in identifying clinically relevant journal articles.

Access to current clinical information involves searches of bibliographic databases, such as MEDLINE, and subsequent evaluation of retrieval results for relevance to a specific clinical situation and quality of the reported research. We establish the amount of information that needs to be provided by an information retrieval system to assist healthcare practitioners in identifying clinically relevant information and evaluating its potential strength of evidence. We find 92% of titles informative enough for a practitioner to correctly classify publications as clinical, but not sufficient for classification of research quality. We suggest automatic organization of retrieval results into strength of evidence categories to supplement title-based judgments and provide quick access to the abstracts of the most promising articles. We find information in the abstracts sufficient to identify articles potentially immediately useful for clinical decision support. These findings are important to the design of information retrieval systems supporting small, low-bandwidth handheld computers.

Abstracting and Indexing↗

PubMedDate not supplied

Anterior osteophyte discrimination in lumbar vertebrae using size-invariant features.

Radiologists often examine X-rays of cervical, thoracic and lumbar vertebrae for determining the presence of osteoarthritis and osteoporosis. For individual vertebra assessment, the boundary increasingly digresses from the general rectangular shape as the vertebra becomes less normal in appearance. For an abnormal vertebra, bony growths ('osteophytes') may appear at the vertebral comers, resulting in a change in the vertebra's shape. Image processing techniques are presented for computing size-invariant, convex hull-based features to highlight anterior osteophytes. Feature evaluation of 714 lumbar spine vertebrae using a multi-layer perceptron yielded normal and abnormal average correct discrimination of 90.5 and 86.6%, respectively.

Algorithms↗

PubMedDate not supplied

Image analysis techniques for characterizing disc space narrowing in cervical vertebrae interfaces.

Image analysis techniques are introduced for evaluating disc space narrowing of cervical vertebrae interfaces from X-ray images. Four scale-invariant, distance transform-based features are presented for characterizing the spacing between adjacent vertebrae. K-means and self-organizing map clustering techniques are applied to estimate the degree of disc space narrowing using a four grade (0-3) scoring system, where 0 and 3 represent normal spacing and significant narrowing, respectively. For a data set of 294 vertebrae interfaces, experimental results yield average correct grade assignment of greater than 82.10% for each of the four grades using a one grade window around the correct grade.

Algorithms↗