Limits...
Center of excellence in research reporting in neurosurgery--diagnostic ontology.

Zaveri A, Shah J, Pradhan S, Rodrigues C, Barros J, Ang BT, Pietrobon R - PLoS ONE (2012)

Bottom Line: Additionally, there is no central repository for storing and retrieving related articles.SR-MA are studies that aggregate several studies to come to one conclusion for a particular research question.We also report high percentage of agreement among five observers as a result of the interobserver agreement test that we conducted among them to annotate 13 articles using the diagnostic ontology.

View Article: PubMed Central - PubMed

Affiliation: Universität Leipzig, Institut für Informatik, Leipzig, Saxony, Germany.

ABSTRACT

Motivation: Evidence-based medicine (EBM), in the field of neurosurgery, relies on diagnostic studies since Randomized Controlled Trials (RCTs) are uncommon. However, diagnostic study reporting is less standardized which increases the difficulty in reliably aggregating results. Although there have been several initiatives to standardize reporting, they have shown to be sub-optimal. Additionally, there is no central repository for storing and retrieving related articles.

Results: In our approach we formulate a computational diagnostic ontology containing 91 elements, including classes and sub-classes, which are required to conduct Systematic Reviews-Meta Analysis (SR-MA) for diagnostic studies, which will assist in standardized reporting of diagnostic articles. SR-MA are studies that aggregate several studies to come to one conclusion for a particular research question. We also report high percentage of agreement among five observers as a result of the interobserver agreement test that we conducted among them to annotate 13 articles using the diagnostic ontology. Moreover, we extend our existing repository CERR-N to include diagnostic studies.

Availability: The ontology is available for download as an.owl file at: http://bioportal.bioontology.org/ontologies/3013.

Show MeSH
Steps involved in designing the diagnostic ontology.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3351462&req=5

pone-0036759-g001: Steps involved in designing the diagnostic ontology.

Mentions: The diagnostic ontology consists of 91 elements, including classes and sub-classes, which are required to conduct SR-MA for diagnostic studies. The hierarchy is displayed in Figure 1. Compared to the RCT ontology, there are 37 elements less as those concepts (or classes) are not considered while conducting diagnostic studies. There were a number of new classes that were added such as the sub-classes under AssesmentRiskBias to cover all the risks and biases that occur in such studies. The ontology contains a hierarchy representing class and sub-class relationships between the classes, which means that any instance of a sub-class is automatically an instance of the parent class. The classes were also specified to be disjoint and certain restrictions were specified for classes that take only certain values as input for their instances. The validation of the ontology was done by using the SPARQL query language, that would allows us to retrieve instances of the classes based on the information we require.


Center of excellence in research reporting in neurosurgery--diagnostic ontology.

Zaveri A, Shah J, Pradhan S, Rodrigues C, Barros J, Ang BT, Pietrobon R - PLoS ONE (2012)

Steps involved in designing the diagnostic ontology.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3351462&req=5

pone-0036759-g001: Steps involved in designing the diagnostic ontology.
Mentions: The diagnostic ontology consists of 91 elements, including classes and sub-classes, which are required to conduct SR-MA for diagnostic studies. The hierarchy is displayed in Figure 1. Compared to the RCT ontology, there are 37 elements less as those concepts (or classes) are not considered while conducting diagnostic studies. There were a number of new classes that were added such as the sub-classes under AssesmentRiskBias to cover all the risks and biases that occur in such studies. The ontology contains a hierarchy representing class and sub-class relationships between the classes, which means that any instance of a sub-class is automatically an instance of the parent class. The classes were also specified to be disjoint and certain restrictions were specified for classes that take only certain values as input for their instances. The validation of the ontology was done by using the SPARQL query language, that would allows us to retrieve instances of the classes based on the information we require.

Bottom Line: Additionally, there is no central repository for storing and retrieving related articles.SR-MA are studies that aggregate several studies to come to one conclusion for a particular research question.We also report high percentage of agreement among five observers as a result of the interobserver agreement test that we conducted among them to annotate 13 articles using the diagnostic ontology.

View Article: PubMed Central - PubMed

Affiliation: Universität Leipzig, Institut für Informatik, Leipzig, Saxony, Germany.

ABSTRACT

Motivation: Evidence-based medicine (EBM), in the field of neurosurgery, relies on diagnostic studies since Randomized Controlled Trials (RCTs) are uncommon. However, diagnostic study reporting is less standardized which increases the difficulty in reliably aggregating results. Although there have been several initiatives to standardize reporting, they have shown to be sub-optimal. Additionally, there is no central repository for storing and retrieving related articles.

Results: In our approach we formulate a computational diagnostic ontology containing 91 elements, including classes and sub-classes, which are required to conduct Systematic Reviews-Meta Analysis (SR-MA) for diagnostic studies, which will assist in standardized reporting of diagnostic articles. SR-MA are studies that aggregate several studies to come to one conclusion for a particular research question. We also report high percentage of agreement among five observers as a result of the interobserver agreement test that we conducted among them to annotate 13 articles using the diagnostic ontology. Moreover, we extend our existing repository CERR-N to include diagnostic studies.

Availability: The ontology is available for download as an.owl file at: http://bioportal.bioontology.org/ontologies/3013.

Show MeSH