English
Digital pathology (DP) provides powerful tools for the acquisition, management, storage, viewing, and analysis of high-resolution digital microscope images. During the last decade, DP tools were increasingly used for a number of applications including biobank, research, education, remote consultations, etc. However, full implementation of the Digital Pathology System (DPS) into the laboratory workflow has remained still challenging in terms of regulatory, technical, financial, and interoperability obstacles. This thesis presents a tightly integrated image handling system for the biobank application within the complex IT infrastructure of the German Competence Network Multiple Sclerosis (KKNMS) research network. The developed concept addresses all steps started from a very early phase of planning and ends with the evaluation, operation, and maintenance of the image handling system for biorepository within the University Medical Center Göttingen (UMG).
In the beginning, the thesis represents the requirements engineering process to address the needs for setting up the complex IT infrastructure for biobank purposes. The requirements were mostly elicited from regular meetings with the direct stakeholders of the KKNMS comprising the main pathologist of the brain bank, research associates, and IT professionals. Additional features were also collected with regard to the scientific literature analysis in already existing DPS established by different pathology laboratories and the market analysis for available image handling systems. Multiple requirements were gathered, determined, and categorized into the functional, technical, quality, and environmental requirements that had been prioritized as essential, preferred, and desirable features of the system.
Once the requirements were determined, a digital workflow for the laboratory processes has been developed and an interoperable image handling system has been set up according to needs. Prior to the digitization of the glass slides, labeling and identification mechanism for more than 4,000 glass slides of the biorepository were determined. For the digitization of the glass slides, hardware and software components of the DPS were configured and customized in such a way that provides an effective approach for acquisition, management, and storage of digital images. Using the automated tools, the scanning procedures were accelerated and the quality of digital workflow was enhanced by avoiding manual steps.
Finally, to increase the efficiency of the separate systems within the biobank infrastructure, an interfacing solution was developed between the DPS and the Laboratory Information Management System (STARLIMS) where the specimens are registered. By using the automated and manual tools, multiple imaging data stored on the DPS can be linked to the corresponding specimen in STARLIMS.
Thus, the thesis provides an effective solution for the interoperable image handling system applied to the biobank research environment, which can be improved further in response to the needs of stakeholders and researchers of the KKNMS. Additionally, it can be adapted to different research needs within the UMG in the future.
Keywords: Digital Pathology; Whole Slide Imaging; Virtual Microscopy
Schlagwörter: Digital Pathology; Whole Slide Imaging; Virtual Microscopy
Persistent Address:
http://dx.doi.org/10.53846/goediss-7367
