 |
|
Posters
|
|
Abstract |
|
|
|
|
|
|
|
|
|
A Development of Metadata Framework to Describe Digitized Pathologic Images
|
|
|
|
|
|
|
|
Sujin Kim, Ph.D. and Robert Thomas, B.S. |
|
|
|
|
|
|
|
Visual findings in pathology are the core of its practice. With the advance of biomedical
imaging in pathology, well-described digital images are increasingly valuable resources
for pathologists as well as basic biomedical researchers. However, there is no such
standard for a "complete" set of metadata, for example, how an image capturing system
(such as digital microscope camera) should be described. To support complex biomedical
queries, researchers often need to integrate data from a number of independent but
related databases. One main challenge is to define metadata that maps the heterogeneously
structured data files into a commonly acceptable file structure such as eXtensible
Markup Language (XML) file to facilitate data interoperation. We developed a metadata
framework by merging two sets of essential microscopic data standards including
the Digital Imaging and Communications in Medicine and the Open Microscopy Environment.
These two sets of metadata (source projects) were individually described in an ontology
editing tool, Protégé 3.1.1 and then merged into a target project. The structured
metadata include data elements, attributes, and values. In addition, hierarchical
structures between data elements are also defined to improve semantic and syntactic
relationship. To provide seamless access to the microscopic image description, the
structured XML files were generated based on the merged metadata standard and then
used to create a searchable XML database.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
