APIII - Advancing Practice, Instruction & Innovation Through Informatics

The Creation of Multiple Standards with Common Data-Types

http://www.newportinstruments.com/cytometryml

Robert C. Leif, PhD (rleif@rleif.com) XML_Med, a Division of Newport Instruments, San Diego, CA
Content: The process of creation of cytology-pathology standards-specifications that can interoperate has been greatly confounded by the differences in the requirements, needs, and past histories of the creating organizations. Presently, efforts to create these standards-specifications include: the Laboratory Digital Imaging Project (LDIP) Data Exchange Specification; DICOM WG 26; Health Level 7 (HL 7), Open Microscopy Environment (OME); Flowcyt, and CytometryML.

Technology: Except for DICOM, the standardization efforts are all based on XML.  The interests of the groups differ. LDIP and DICOM WG 26 are primarily interested in standards for pathology imaging and those standards created by members of the International Society for Analytical Society (ISAC) are primarily addressed towards the data generated from images.  At this point, LDIP's main interest is in the relationships between different objects (elements); ISAC and OME share a common interest in the transfer of data; and the DICOM Working Group 26 plans to evaluate and extend the current DICOM standard as it relates to newer microscopic techniques including whole slide imaging.

The heterogeneity of the imaging modalities: digital microscopy and flow cytometry, as well as the differences in data obtained from single cells, as opposed to tissues, provide significant complications to the creation of interoperable data standards. The problem data exchange between the standards has been exacerbated by the inability of the XML ontology languages (RDF and OWL) to directly reference and thus reuse the complexTypes, which are the XML schema construct for describing classes.

Design: The achievement of interoperability can be facilitated by employing data-types with common semantics and limiting the difference between standards-specifications to syntax.

Results: A collection of schemas written in CytometryML has demonstrated the feasibility of reusing the semantics of data-types from DICOM and those from ISAC’s FCS standard, as well as reusing their documentation. The CytometryML schemas contain attributes that link to the DICOM standard and the legacy ISAC Flow Cytometry Standard.  The essential unity of flow cytometers and digital microscopes has been demonstrated by deriving both data-types from a common cytometer data-type.  CytometryML will be extended to include many of the new DICOM data-types that describe patients, specimens, etc. being prepared by DICOM WG 26 and by LDIP.

Conclusion: In situations where multiple standards are being created, interoperability can be facilitated by employing data-types based on a common set of semantics.
Keywords: CytometryML, DICOM, FCS, LDIP, OME, flow cytometry, pathology informatics, specification, standard, XML schema, RDF