The Image- and Data Quality Ontology for scientific data management

Transcription

1 The Image- and Data Quality Ontology for scientific data management Thomas Schrader, Yao Zhou, Sonja Niepage, Claudia Hahn, Sabine Hanss University of Applied Sciences Brandenburg, Department of Pathology Charité, Department of Medical Informatics Charité Outline 1 Background 2 Material & Methods 3 4 Discussion Schrader (FHB & Charité) Image- and Data Quality Ontology 2 / 25

2 Outline Background 1 Background 2 Material & Methods 3 4 Discussion Schrader (FHB & Charité) Image- and Data Quality Ontology 3 / 25 Background Secundary Use of Clinical Data Schrader (FHB & Charité) Image- and Data Quality Ontology 4 / 25

3 Background Discusion about data & image quality Usage of data in various use cases - study recruitment, pattern analysis, exploratory data analysis Images in different contexts - diagnostic process, second opinion, education, image analysis, follow up Different views of clinicians, scientists and technicians Different responsibilities in data life cycle Schrader (FHB & Charité) Image- and Data Quality Ontology 5 / 25 Outline Material & Methods 1 Background 2 Material & Methods 3 4 Discussion Schrader (FHB & Charité) Image- and Data Quality Ontology 6 / 25

4 Material & Methods Development Ontology Delphi-Method - Discusion and determination of quality criteria Analysis of quality criteria - calculation of numeric quality indicators Modeling the ontology via Protegé Schrader (FHB & Charité) Image- and Data Quality Ontology 7 / 25 Metadata Analysis Material & Methods 3461 DICOM-image from 3 Radiology Departments 822 DICOM-images from experiments with a phantom 2639 anonymized,randomly selected images 5 different Computer Tomographs SOMATON Sensation (Siemens) SOMATON Definition AS+ (Siemens) MX-16 (Philips) Brilliance (Philips) Brightspeed (GE) Schrader (FHB & Charité) Image- and Data Quality Ontology 8 / 25

5 Outline 1 Background 2 Material & Methods 3 4 Discussion Schrader (FHB & Charité) Image- and Data Quality Ontology 9 / 25 Data Life Cycle Schrader (FHB & Charité) Image- and Data Quality Ontology 10 / 25

6 Roles in Data Life Cycle Schrader (FHB & Charité) Image- and Data Quality Ontology 11 / / 25 Facets of data quality Schrader (FHB & Charité) Image- and Data Quality Ontology

7 Data types Schrader (FHB & Charité) Image- and Data Quality Ontology 13 / / 25 Methods of Quality Analsis of Images Schrader (FHB & Charité) Image- and Data Quality Ontology

8 Quality Dimensions Schrader (FHB & Charité) Image- and Data Quality Ontology 15 / / 25 Quality Dimension: Free of Error Schrader (FHB & Charité) Image- and Data Quality Ontology

9 Quality Dimension: Completeness.Resolution Schrader (FHB & Charité) Image- and Data Quality Ontology 17 / 25 Quality Dimension: Completeness Schrader (FHB & Charité) Image- and Data Quality Ontology 18 / 25

10 Completeness.Metadata: used metadata Figure: Number of shared metadata Schrader (FHB & Charité) Image- and Data Quality Ontology 19 / 25 Completeness.Metadata - shared metadata Schrader (FHB & Charité) Image- and Data Quality Ontology 20 / 25

11 Discussion Outline 1 Background 2 Material & Methods 3 4 Discussion Schrader (FHB & Charité) Image- and Data Quality Ontology 21 / 25 Thinking about Quality Discussion Quality is not equal quality - essential differentiation of quality dimension Quality & use case Who is responsible for what? - data life cycle & roles Specification of standardization aim - technolocial aspects vs. content vs. quality dimension Schrader (FHB & Charité) Image- and Data Quality Ontology 22 / 25

12 Discussion Framework for Analysis of Data Quality Y. Zhou, S. Hanß, M. Cornils, C. Hahn, S. Niepage, T. Schrader. A SOA-based Data Quality Assessment Framework in a Medical Science Center. ICIQ Potsdam, 2009 Schrader (FHB & Charité) Image- and Data Quality Ontology 23 / 25 Take Home Message Discussion OpEN.SC Ontology presents quality criteria and their relations Data Life Cycle Roles Use cases Ressources & devices Transparency Implementability Specific requirements Availability of an algorithms Allocation of quality indicators Schrader (FHB & Charité) Image- and Data Quality Ontology 24 / 25

13 Discussion Thank you very much for your attention Any questions? Contact Mail: Mobile: Tel.: +49 (0) Schrader (FHB & Charité) Image- and Data Quality Ontology 25 / 25