Ontology Usability Via a Visualization Tool for the Semantic Indexing of Medical Reports (DICOM SR)
One purpose of our research works is a contribution to a semantic indexing of structured reports in accordance with the DICOM SR standard and we propose to guide this process with an ontology. In this paper, we describe our motivations for building this o
- PDF / 602,843 Bytes
- 6 Pages / 430 x 660 pts Page_size
- 19 Downloads / 205 Views
The Computer Science Center of the Paris V University (CRIP5) 75006 Paris, France 2 The Computer Science Laboratory of the Paris XIII University (LIPN) 93430 Villetaneuse, France [email protected], [email protected]
Abstract. One purpose of our research works is a contribution to a semantic indexing of structured reports in accordance with the DICOM SR1 standard and we propose to guide this process with an ontology. In this paper, we describe our motivations for building this ontology according to a modularization approach assisted by the reuse of existing ontologies. Moreover, a prototype of a bilingual visualization tool is suggested. It allows specialists during their semantic indexing to load and visualize ontologies or modules from an ontology in a multi-axial way. Currently, six axes are planned: patient context, anatomy, pathology, visual descriptor, technique and recommendation. Keywords: DICOM SR, Visualization Tool, Semantic indexing, Ontologies.
1 Introduction For the medical imaging community taking advantage of the DICOM SR standard to improve medical image retrieval systems such as CBIR2 one become a challenging research issue. A CBIR system refers to the retrieval from image databases using information extracted from the content of images [10]. In this paper, a special emphasis is given to semantic CBIR systems and more particularly to the use of ontologies as a support of indexing [15]. Our initial aim is a contribution to an ontology-based semantic indexing of structured reports in accordance with the DICOM SR standard. DICOM3 is the only standard that can be used by imaging industry for the exchange and management of multimodalities images (radiology, MRI...). Since 1993, this standard was centered only on the image. In 2000, Structured Reporting (SR) was added to the DICOM standard to provide an efficient mechanism for the management of clinical reports [1][2][3][4]. The main advantage of SR is the ability to link clinical reports with the referenced images for simultaneous 1
Digital Imaging and Communications in Medicine for Structured Reporting. Content-Based Image Retrieval. 3 http://medical.nema.org 2
A. Holzinger (Ed.): USAB 2007, LNCS 4799, pp. 409–414, 2007. © Springer-Verlag Berlin Heidelberg 2007
410
S. Mhiri and S. Despres
retrieval and display. From the computerized systems perspective, SR has many potential advantages, such as the production of well-organized reports and the ability to communicate results promptly with more speed, reduced costs and fewer errors. The rest of this paper is organized as follows. In section 2, we discuss our motivations to build an ontology for the semantic indexing of DICOM SR documents according to a modularization approach. In section 3, a first solution is proposed and consists in a prototype of a bilingual visualization tool to help specialists in their semantic indexing of reports. In section 4, a brief overview on related works is given while in the last section, the conclusion of our work and a brief de
Data Loading...
