Search

The German Research Foundation (DFG) recommends preserving research data for at least ten years. The DFG funded project LABIMI/F establishes an infrastructure for preservation, retrieval and reuse of biomedical research data based on grid/cloud computing technology for two applications a) genome and b) imaging data. The requirements for this infrastructure were determined during workshops with relevant stakeholders. Afterwards product evaluations were conducted and the relevant products were integrated into the infrastructure concept. In this paper, we address the suitability of our solution concerning the fulfillment of the requirements. It is shown that the solution satisfies five of the eight requirement categories completely and the other three categories partly. Furthermore, in order to prove the adherence to the widely accepted Open Archival Information System (OAIS) standard, we successfully performed a mapping of our technical components to the functional entities of the OAIS.

The current situation at many university medical centers regarding the management of biomedical research imaging data leaves much to be desired. In contrast to the recommendations of the German Research Foundation (DFG) and the German Council of Sciences and Humanities regarding the professional management of research data, there are commonly many individual data pools for research data in each institute and the management remains the responsibility of the researcher. A possible solution for this situation would be to install local central repositories for biomedical research imaging data. In this paper, we developed a scenario based on abstracted use-cases for institutional research undertakings as well as collaborative biomedical research projects and analyzed the functional requirements that a local repository would have to fulfill. We determined eight generic categories of functional requirements, which can be viewed as a basic guideline for the minimum functionality of a central repository for biomedical research imaging data.

Research data management (RDM) is an important topic for biomedical research due to the issue of "bit rot". RDM aims to implement access to reliable digital data for local and distributed research groups. A key aspect for the understanding of data is the use of metadata. This understanding has been investigated on the basis of two use cases of the DFG project LABIMI/F: RDM for genome data and biomedical image data. The results show that metadata can improve research not only for others but also for the researcher himself. However, RDM is still far from integrating all biomedical data. In addition, RDM is not (yet) a valid approach for clinical trial data management.

Natural scientists such as physicists pioneered the sharing of computing resources, which led to the creation of the Grid. The inter domain transfer process of this technology has hitherto been an intuitive process without in depth analysis. Some difficulties facing the life science community in this transfer can be understood using the Bozeman's "Effectiveness Model of Technology Transfer". Bozeman's and classical technology transfer approaches deal with technologies which have achieved certain stability. Grid and Cloud solutions are technologies, which are still in flux. We show how Grid computing creates new difficulties in the transfer process that are not considered in Bozeman's model. We show why the success of healthgrids should be measured by the qualified scientific human capital and the opportunities created, and not primarily by the market impact. We conclude with recommendations that can help improve the adoption of Grid and Cloud solutions into the biomedical community. These results give a more concise explanation of the difficulties many life science IT projects are facing in the late funding periods, and show leveraging steps that can help overcoming the "vale of tears".